Interior Lighting: Fundamentals, Technology and Application 3030171949, 9783030171940

Table of contents :
Preface......Page 6
Contents......Page 8
About the Author......Page 16
Abbreviations......Page 17
Part I: Fundamentals......Page 19
Chapter 1: Visual Mechanism......Page 20
1.1 Visual Sensation......Page 21
1.2 Optics of the Eye......Page 22
1.3 Retina and Photoreceptors......Page 23
1.4.1 Cones and Rods......Page 26
1.4.2 Photopic and Scotopic Vision......Page 28
1.5 Receptive Fields......Page 30
1.6 Colour Vision......Page 34
1.7 Pupillary Reflex......Page 36
1.8.1 Perceptual Constancy......Page 37
1.8.2 Maintaining Constancy......Page 38
References......Page 39
2.1 Perceived Colour......Page 41
2.2.1.1 Chromaticity Coordinates......Page 42
2.2.1.2 Standard Colorimetric Observer......Page 44
2.2.1.3 From Three-Dimensional Space to Two-Dimensional Plane......Page 47
2.2.2 CIE u′-v′ Chromaticity Diagram......Page 48
2.2.3 Colour Spaces......Page 49
2.3.1 Correlated Colour Temperature CCT......Page 51
2.4 Dominant Wavelength and Excitation Purity......Page 53
2.5 Colour Rendering......Page 54
2.5.1.1 Colour Samples......Page 55
2.5.1.2 Reference Light Sources......Page 56
2.5.1.4 Use of General Colour Rendering Index Ra......Page 57
2.5.2 General Colour Fidelity Index, Rf......Page 58
2.5.2.1 Colour Samples......Page 59
2.5.2.4 Use of General Fidelity Index, Rf......Page 60
2.5.3.1 Gamut Area......Page 62
2.5.3.2 IES Gamut Index Rg......Page 63
2.5.4.1 IES Colour Vector Graphic......Page 64
2.5.4.2 CIE Colour Vector Graphic......Page 65
2.5.5 Colour Discrimination......Page 66
2.5.6 Surface-Colour Metamerism......Page 67
2.6 Summary of Colour Metrics......Page 68
References......Page 69
Chapter 3: Visual Performance......Page 72
3.1 Visual Task......Page 73
3.2 Threshold Visibility......Page 75
3.2.1 Visual Acuity......Page 76
3.2.2 Threshold Contrast......Page 78
3.3 Suprathreshold Visibility......Page 80
3.3.1 Landolt-Ring Task......Page 81
3.3.2 Search Task......Page 86
3.3.3 Verification Task......Page 87
3.3.4 Rea´s Visual Performance Model......Page 89
3.4 Disability Glare......Page 91
3.5.1 Non-self-luminous Tasks......Page 94
3.5.2 Self-Luminous Devices......Page 96
3.6.1 Pupil Size......Page 97
3.6.2 Visual Acuity......Page 98
3.6.3 Equivalent Visual Efficiency Method......Page 99
References......Page 101
Chapter 4: Visual Satisfaction......Page 103
4.1.1 Brightness-Luminance Relation......Page 104
4.1.2 Influence of Spectrum......Page 107
4.2.1.1 Task Surface......Page 109
4.2.1.2 Room Boundary Surfaces......Page 110
4.2.1.3 Metric for Subjective Lighting Quality: B40 Band Based......Page 115
4.2.2.1 Indirect Illuminance at the Eye......Page 116
4.2.2.2 Formulas for the Calculation of MRSE......Page 119
4.2.2.3 MRSE-Based Metric for Subjective Lighting Quality......Page 120
4.2.3 Items for Further Study......Page 121
4.3 Directionality and Modelling......Page 122
4.3.1 Flow of Lighting......Page 123
4.3.2 Light Tubes......Page 126
4.4.1 Fundamental Approach......Page 128
4.4.2 Unified Glare Rating, UGR......Page 129
4.4.2.1 UGR for a Specific Location and Viewing Direction......Page 131
4.4.2.2 Unified Glare Rating for a Lighting Installation, UGRL......Page 133
4.4.2.3 Non-uniform Glare Sources......Page 135
4.4.2.4 Influence of Spectrum......Page 139
4.4.3 Overhead Glare......Page 140
4.5 Light Colour Preference......Page 141
References......Page 144
Chapter 5: Non-visual Biological Mechanism......Page 151
5.1.1 The Principle......Page 152
5.1.2 Bodily Rhythms......Page 155
5.1.3 Biological Clock......Page 156
5.1.4 Chronotypes......Page 157
5.1.5.2 Entrainment by Colour Transitions......Page 158
5.1.6 Phase Shifting......Page 159
5.2.1 Photosensitive Retinal Ganglion Cells (pRGCs)......Page 161
5.2.3 Spatial Distribution of pRGCs......Page 162
5.3 Pineal Gland and Adrenal Cortex......Page 163
5.4 Direct Photobiological Effects......Page 165
5.5.1.1 Melatonin Suppression by Monochromatic Light......Page 166
5.5.1.2 Melatonin Suppression by Polychromatic Light......Page 168
5.5.1.3 Single Non-visual Biological Action Spectrum?......Page 170
5.5.2 Spectral Sensitivity of Photopigments......Page 171
5.5.3 Spectral Characterisation of Lighting Installations......Page 176
References......Page 177
6.1 Sleep......Page 183
6.1.1 Sleep Mechanism......Page 184
6.1.3 Daytime Light......Page 185
6.1.3.1 Light Level......Page 186
6.1.3.2 Light Spectrum......Page 188
6.2.1 Alertness and Performance Measures......Page 190
6.2.2 Daytime Light......Page 191
6.2.2.1 Light Level......Page 192
6.2.2.2 Light Spectrum......Page 194
6.3 Dynamic Daytime Lighting Scenario......Page 195
References......Page 197
Chapter 7: Shift Work, Light, Sleep and Performance......Page 201
7.1 Circadian Misalignment......Page 202
7.2.1 Sleep......Page 203
7.2.2 Alertness and Performance......Page 204
7.3.1 Bright Light......Page 206
7.3.2 Intermittent Bright Light......Page 209
7.3.3 Short-Wavelength Depleted White Light......Page 211
7.4 Dynamic Lighting Scenarios......Page 213
References......Page 217
8.1 Changes in the Visual and Circadian System......Page 221
8.2.1 Loss of Lens Elasticity......Page 222
8.2.2 Reduced Pathway Signal......Page 223
8.2.3 Increased Glare......Page 226
8.2.5 Reduced Visual Performance......Page 227
8.3.2 Melatonin Concentration......Page 229
8.3.4 Sleep and Alertness......Page 230
References......Page 231
9.1 Chronotherapy......Page 235
9.2 Disrupted Circadian Rhythms......Page 236
9.3.1.2 Non-seasonal Affective Disorders......Page 237
9.3.3 Alzheimer´s Disease......Page 238
9.3.4 Parkinson´s Disease......Page 239
9.3.5 ADHD Disorder......Page 240
9.3.7.1 Patient Rooms......Page 241
References......Page 243
10.1 Adverse Effects of Lamp Flicker......Page 247
10.1.1.1 Cause of Lamp Flicker......Page 248
10.1.1.2 Light Modulation Metrics......Page 250
10.1.2.1 Visible Flicker......Page 251
10.1.2.2 Stroboscopic Effect......Page 252
10.1.3.1 Epileptic Seizures......Page 256
10.1.3.2 Migraine, Headache, Eye Strain and Malaise......Page 257
10.1.4 One Tool to Specify the Risk of Flicker?......Page 258
10.2.1 Photochemical and Thermal Damage......Page 260
10.2.2.1 The Mechanism......Page 261
10.2.2.2 Grouping Lamps in Risk Groups......Page 263
10.2.2.3 Evaluating in the Lighting Installation......Page 265
10.3.1 Circadian Disruption, Melatonin and Medical Disorders......Page 267
10.3.2.1 Animal Studies......Page 268
10.3.2.2 Epidemiological Studies with Humans......Page 269
References......Page 271
Part II: Technology......Page 275
11.1 Introduction......Page 276
11.2.1 System Efficacy......Page 278
11.2.2.1 Incandescent Lamps......Page 279
11.2.2.2 Gas-Discharge Lamps......Page 280
11.2.2.3 Solid-State Lamps......Page 281
11.3 Incandescent Lamps: GLS......Page 285
11.3.2 Construction......Page 286
11.3.3 Lamp Properties......Page 287
11.4.2 Construction......Page 288
11.4.3 Lamp Properties......Page 289
11.5.1.1 Gas Discharge......Page 290
11.5.1.2 Fluorescence......Page 292
11.5.3 Lamp Properties......Page 293
11.6 Gas-Discharge Lamps: Compact Fluorescent......Page 295
11.6.2 Construction......Page 296
11.6.3 Lamp Properties......Page 297
11.7.1 Working Principle......Page 298
11.7.3 Lamp Properties......Page 299
11.8.2 Construction......Page 300
11.8.3 Lamp Properties......Page 301
11.9 Control Gear for Gas-Discharge Lamps......Page 302
11.9.2 Ballasts......Page 303
11.10 Solid-State Lamps: LED......Page 305
11.10.1.1 Light Emission......Page 306
11.10.1.2 Efficiency Droop......Page 308
11.10.1.3 White Light......Page 309
11.10.2 Construction and Manufacturing......Page 311
11.10.4 Lamp Properties......Page 317
11.11 Solid-State Lamps: OLED......Page 319
11.11.1 Working Principle......Page 320
11.11.2 Construction......Page 321
11.12.1 Drivers......Page 322
11.12.2 Dimmers......Page 324
References......Page 325
12.1 Introduction......Page 328
12.2 Origin of Daylight......Page 329
12.3.1 Variability of Daylight......Page 330
12.3.2 Daylight Levels......Page 331
12.3.3 Spectrum and Colour......Page 332
12.3.4 Sky Luminance......Page 334
12.4.1 Interior Daylight Levels......Page 335
12.4.1.1 Average Daylight Factor......Page 336
12.4.1.2 Daylight Factor at a Point......Page 338
12.4.2 Flow of Daylight......Page 339
References......Page 340
Chapter 13: Luminaires......Page 341
13.1.1 Light Distribution......Page 342
13.1.2.1 Light Output Ratio......Page 345
13.1.2.2 Utilisation Factors......Page 346
13.1.4 UGR Table......Page 349
13.1.5 Shielding Angle......Page 353
13.2.1 Mirrors......Page 354
13.2.2 Prisms and Lenses......Page 356
13.2.3 Total Internal Reflection (TIR) Optics......Page 357
13.3 Thermal Characteristics......Page 359
13.4 Lifetime and Lumen Maintenance......Page 360
References......Page 361
Chapter 14: Connected Smart Lighting......Page 363
14.1.1 Switching and Dimming......Page 364
14.1.3.1 Automatic Control Systems......Page 365
14.1.3.2 Personal Control Systems......Page 367
14.2.2 Network Topologies......Page 369
14.2.3 Wired and Wireless Transmission......Page 370
14.2.4.1 Analogue Protocols......Page 371
14.2.4.2 Digital Wired Protocols and Technologies......Page 372
14.2.4.3 Digital Wireless Protocols and Technologies......Page 375
14.3 Power over Ethernet (PoE)......Page 379
References......Page 380
15.1.1 Internet of Things (IoT)......Page 383
15.1.2.2 IoT Lighting System Technology......Page 385
15.1.3 Services of IoT Lighting Systems......Page 386
15.2.1 Principle......Page 387
15.2.2 Data Push Applications......Page 391
15.2.3 Indoor Navigation......Page 392
15.2.4 Light as Sensor......Page 393
15.2.5 Li-Fi......Page 394
References......Page 395
Part III: Application......Page 398
16.1 Lighting Quality Parameters......Page 399
16.1.1.1 Horizontal Illuminance......Page 400
16.1.1.4 Mean Room Surface Exitance......Page 401
16.1.1.6 Circadian Stimulus......Page 402
16.1.3.2 Cylindrical-to-Horizontal Illuminance Ratio......Page 403
16.1.4.2 Chromaticity Distance from Blackbody Locus......Page 404
16.2 Standards and Recommendations......Page 405
16.2.2 European Standard......Page 406
16.2.2.1 Illuminance of the Working Plane and Its Surroundings......Page 407
16.2.2.4 Discomfort Glare of the Installation......Page 410
16.2.2.6 Indirect Glare due to Reflections in Display Screens......Page 411
16.2.3 North American Standard......Page 412
16.2.3.1 Illuminance Target Value System......Page 413
16.2.3.4 Discomfort Glare of the Installation......Page 414
16.2.3.7 Colour......Page 415
References......Page 416
Chapter 17: Design Aspects......Page 417
17.1.1 Analysis of the Lighting Function......Page 418
17.1.2 Determination of Lighting Quality......Page 419
17.1.5 Determination of Number and Positions of Luminaires......Page 420
17.2.1.1 Functional Lighting System......Page 421
17.2.1.2 Ambient Lighting System......Page 424
17.2.2.1 Work Area Lighting......Page 425
17.2.2.2 Task-Related Dedicated Lighting Effects......Page 426
17.2.3.1 Lighting Goals......Page 427
17.2.4.1 Wardrooms......Page 430
17.2.4.3 Nursing Homes......Page 432
17.2.5 Emergency Lighting......Page 433
17.2.5.1 Categories of Emergency Lighting......Page 434
17.2.5.2 Emergency Lighting Installation......Page 437
References......Page 438
18.1 Calculations......Page 440
18.1.1 The Lumen Method......Page 441
18.2.1.1 Types......Page 442
18.2.1.2 V(λ) Correction......Page 443
18.2.1.4 Pulsed Light Measurement......Page 444
18.2.1.7 Accuracy......Page 445
18.2.2.1 Luminous Flux......Page 446
18.2.2.2 Spectral Data......Page 447
18.2.3.1 Light Distribution......Page 448
18.2.3.2 Light Output......Page 450
18.2.4.1 Illuminance and Luminance Measurements......Page 451
18.2.4.2 Glare Measurements......Page 452
18.2.5 Light-Logging Devices......Page 453
References......Page 454
Appendix A: Standardised Relative Spectral Sensitivity Values V(λ)......Page 456
Appendix B: Calculation of x-y Chromaticity Coordinates......Page 457
CIE Colour-Matching Functions......Page 458
Appendix C: RVP Model of Weston......Page 460
Appendix D: RVP Model of Rea......Page 461
Appendix E: Evector/Escalar Ratio......Page 463
Appendix F: Position Indices, p......Page 465
Appendix G: Groningen Sleep Quality Scale......Page 468
Appendix H: Normalised Formula for SVM according to CIE (2016)......Page 469
References......Page 470
Author Index......Page 471
Subject Index......Page 479

Citation preview

Wout van Bommel

Interior Lighting

Fundamentals, Technology and Application

Interior Lighting

Wout van Bommel

Interior Lighting Fundamentals, Technology and Application

Wout van Bommel Van Bommel Lighting Consultant Nuenen, Noord-Brabant, The Netherlands

ISBN 978-3-030-17194-0 ISBN 978-3-030-17195-7 https://doi.org/10.1007/978-3-030-17195-7

(eBook)

© Springer Nature Switzerland AG 2019 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Dedicated to my former lighting teachers, Prof. Nito de Boer ({), Prof. Jochen SchmidtClausen and Prof. Dietert Fischer ({), who gave me lots of trust, motivation and opportunities and to Hari Mamak ({) and Prof. Tai-Ming Zhou who were instrumental in creating challenging teaching opportunities for me in Asia.

Preface

This book outlines the underlying principles on which interior lighting should be based for arriving at good visual performance, comfort, alertness and health for the users of the lit space. It is the first interior lighting book in which the visual and non-visual biological effects of lighting are dealt with on an equal footing. Around the same time that solid-state lighting (LED lighting) was introduced, some 20 years ago, we began to learn that lighting has apart from visual effects also far-reaching non-visual biological effects. These effects influence the way our body “operates” and, therefore, influence our health, well-being and alertness. Interior lighting installations today have to be designed so that they can provide both suitable visual and non-visual biological effects. Lighting that indeed does this is referred to as human-centric lighting. This book gives all the fundamental and practical information needed to design human-centric lighting installations. The introduction of solid-state light sources has provided the possibility to design innovative, truly sustainable lighting installations which are adaptable to changing circumstances. The design of such solid-state lighting installations is more difficult than it was with gas discharge lamps. To avoid disappointments with LED lighting installations, detailed knowledge of the typical characteristics of the many different solid-state light sources is essential. Already long-available information on vision and colour seeing has to be combined with entirely new fundamental research on the relationship between lighting on the one hand and vision, performance, comfort, health and well-being on the other hand. LEDs offer the possibility to use them not only for lighting but also for data transmission. LEDs can, therefore, be used as the heart of the Internet of Things (IoT). Here, data communication is combined with microsensor technology to create connected smart environments. The use of LED lighting as a means for data communication is referred to as “light beyond illumination”. Visible light communication (VLC), LiFi and light itself used as sensor are part of this subject. The modern lighting professional has to get familiarised with these new technologies and applications.

vii

viii

Preface

The book is divided into three parts. Part I discusses the fundamentals of the visual and non-visual mechanisms and the practical consequences for visual performance and comfort, for sleep, for health and for alertness. It includes chapters on shift work, on therapeutic and hazardous effects of lighting and on the effect of the ageing eye. Part II, Technology, deals with the lighting hardware, lamps (with emphasis on LEDs), gear, drivers and luminaires, including chapters about smart connected lighting and light used for data communication (VLC and Li-Fi) and as sensor. It also has a chapter about daylight. Part III, the application part, provides the link between theory and practice and supplies the reader with the knowledge needed for lighting design. It describes the relevant lighting criteria for efficient interior lighting and discusses the international, European and North American standards for interior lighting. This part concludes with a chapter about interior lighting calculations and measurements. I wish to record my gratitude to Prof. Steve Fotios for reviewing the entire manuscript and to Wang Shen for reviewing Chaps. 14 and 15. For the lighting professional or student being active in both indoor and outdoor lighting, I refer to my book Road Lighting: Fundamentals, Technology and Application that was published in 2015 with Springer. Nuenen, Noord-Brabant, The Netherlands

Wout van Bommel

Contents

Part I

Fundamentals

1

Visual Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 Visual Sensation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 Optics of the Eye . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3 Retina and Photoreceptors . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.4 Spectral Sensitivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.4.1 Cones and Rods . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.4.2 Photopic and Scotopic Vision . . . . . . . . . . . . . . . . . 1.5 Receptive Fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.6 Colour Vision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.7 Pupillary Reflex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.8 High-Level Vision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.8.1 Perceptual Constancy . . . . . . . . . . . . . . . . . . . . . . . 1.8.2 Maintaining Constancy . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3 4 5 6 9 9 11 13 17 19 20 20 21 22

2

Colour . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Perceived Colour . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Colour Specification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.1 CIE x–y Chromaticity Diagram . . . . . . . . . . . . . . . . 2.2.2 CIE u0 –v0 Chromaticity Diagram . . . . . . . . . . . . . . . 2.2.3 Colour Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 Colour Appearance of White Light . . . . . . . . . . . . . . . . . . . . . 2.3.1 Correlated Colour Temperature CCT . . . . . . . . . . . . 2.3.2 Distance from Blackbody Locus, Duv . . . . . . . . . . . . 2.4 Dominant Wavelength and Excitation Purity . . . . . . . . . . . . . . 2.5 Colour Rendering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5.1 General Colour Rendering Index Ra . . . . . . . . . . . . . 2.5.2 General Colour Fidelity Index, Rf . . . . . . . . . . . . . . 2.5.3 Gamut Index Rg . . . . . . . . . . . . . . . . . . . . . . . . . . .

25 25 26 26 32 33 35 35 37 37 38 39 42 46 ix

x

Contents

2.5.4 Colour Vector Graphics . . . . . . . . . . . . . . . . . . . . 2.5.5 Colour Discrimination . . . . . . . . . . . . . . . . . . . . . . 2.5.6 Surface-Colour Metamerism . . . . . . . . . . . . . . . . . 2.6 Summary of Colour Metrics . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . .

48 50 51 52 53

3

Visual Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Visual Task . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Threshold Visibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.1 Visual Acuity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.2 Threshold Contrast . . . . . . . . . . . . . . . . . . . . . . . . . 3.3 Suprathreshold Visibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.1 Landolt-Ring Task . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.2 Search Task . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.3 Verification Task . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.4 Rea’s Visual Performance Model . . . . . . . . . . . . . . . 3.4 Disability Glare . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5 Indirect Glare . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5.1 Non-self-luminous Tasks . . . . . . . . . . . . . . . . . . . . . 3.5.2 Self-Luminous Devices . . . . . . . . . . . . . . . . . . . . . . 3.6 Influence of Spectrum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6.1 Pupil Size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6.2 Visual Acuity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6.3 Equivalent Visual Efficiency Method . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

57 58 60 61 63 65 66 71 72 74 76 79 79 81 82 82 83 84 86

4

Visual Satisfaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 Spatial Brightness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.1 Brightness-Luminance Relation . . . . . . . . . . . . . . . 4.1.2 Influence of Spectrum . . . . . . . . . . . . . . . . . . . . . . 4.2 Room Appearance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.1 Room Surface Illuminance and Luminance . . . . . . . 4.2.2 Mean Room Surface Exitance, MRSE . . . . . . . . . . 4.2.3 Items for Further Study . . . . . . . . . . . . . . . . . . . . . 4.3 Directionality and Modelling . . . . . . . . . . . . . . . . . . . . . . . . 4.3.1 Flow of Lighting . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.2 Light Tubes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4 Discomfort Glare . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4.1 Fundamental Approach . . . . . . . . . . . . . . . . . . . . . 4.4.2 Unified Glare Rating, UGR . . . . . . . . . . . . . . . . . . 4.4.3 Overhead Glare . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4.4 Indirect Glare . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5 Light Colour Preference . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

89 90 90 93 95 95 102 107 108 109 112 113 113 116 126 127 127 131

. . . . . . . . . . . . . . . . . .

Contents

xi

Non-visual Biological Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . 5.1 Circadian Rhythms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1.1 The Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1.2 Bodily Rhythms . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1.3 Biological Clock . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1.4 Chronotypes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1.5 Entrainment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1.6 Phase Shifting . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Third Type of Photoreceptor . . . . . . . . . . . . . . . . . . . . . . . . 5.2.1 Photosensitive Retinal Ganglion Cells (pRGCs) . . . 5.2.2 Retinal Neural Wiring . . . . . . . . . . . . . . . . . . . . . . 5.2.3 Spatial Distribution of pRGCs . . . . . . . . . . . . . . . . 5.2.4 Field of View . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3 Pineal Gland and Adrenal Cortex . . . . . . . . . . . . . . . . . . . . . 5.4 Direct Photobiological Effects . . . . . . . . . . . . . . . . . . . . . . . 5.5 Spectral Sensitivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5.1 Action Spectra . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5.2 Spectral Sensitivity of Photopigments . . . . . . . . . . 5.5.3 Spectral Characterisation of Lighting Installations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

6

Light, Sleep, Alertness and Performance . . . . . . . . . . . . . . . . . . . . 6.1 Sleep . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.1 Sleep Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.2 Sleep Measures . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.3 Daytime Light . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2 Alertness and Performance . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.1 Alertness and Performance Measures . . . . . . . . . . . 6.2.2 Daytime Light . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3 Dynamic Daytime Lighting Scenario . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . .

169 169 170 171 171 176 176 177 181 183

7

Shift Work, Light, Sleep and Performance . . . . . . . . . . . . . . . . . . 7.1 Circadian Misalignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2 Sleep, Alertness and Performance . . . . . . . . . . . . . . . . . . . . . 7.2.1 Sleep . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2.2 Alertness and Performance . . . . . . . . . . . . . . . . . . 7.3 Night-Time Lighting Strategies . . . . . . . . . . . . . . . . . . . . . . 7.3.1 Bright Light . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.3.2 Intermittent Bright Light . . . . . . . . . . . . . . . . . . . . 7.3.3 Short-Wavelength Depleted White Light . . . . . . . . 7.4 Dynamic Lighting Scenarios . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . .

187 188 189 189 190 192 192 195 197 199 203

5

137 138 138 141 142 143 144 145 147 147 148 148 149 149 151 152 152 157

. 162 . 163

xii

Contents

8

Age Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.1 Changes in the Visual and Circadian System . . . . . . . . . . . . . . 8.2 Age and Visual Effects of Light . . . . . . . . . . . . . . . . . . . . . . . 8.2.1 Loss of Lens Elasticity . . . . . . . . . . . . . . . . . . . . . . 8.2.2 Reduced Pathway Signal . . . . . . . . . . . . . . . . . . . . . 8.2.3 Increased Glare . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.2.4 Slower Dark Adaptation . . . . . . . . . . . . . . . . . . . . . 8.2.5 Reduced Visual Performance . . . . . . . . . . . . . . . . . . 8.3 Age and Non-visual Biological Effects of Light . . . . . . . . . . . 8.3.1 Reduced Pathway Signal . . . . . . . . . . . . . . . . . . . . . 8.3.2 Melatonin Concentration . . . . . . . . . . . . . . . . . . . . . 8.3.3 Circadian Rhythm . . . . . . . . . . . . . . . . . . . . . . . . . . 8.3.4 Sleep and Alertness . . . . . . . . . . . . . . . . . . . . . . . . 8.3.5 Light-Dark Scenario . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

207 207 208 208 209 212 213 213 215 215 215 216 216 217 217

9

Therapeutic Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.1 Chronotherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.2 Disrupted Circadian Rhythms . . . . . . . . . . . . . . . . . . . . . . . 9.3 Light Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.3.1 Depressions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.3.2 Sleep Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . 9.3.3 Alzheimer’s Disease . . . . . . . . . . . . . . . . . . . . . . . 9.3.4 Parkinson’s Disease . . . . . . . . . . . . . . . . . . . . . . . 9.3.5 ADHD Disorder . . . . . . . . . . . . . . . . . . . . . . . . . . 9.3.6 Eating Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . 9.3.7 Healing in the Hospital . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . .

221 221 222 223 223 224 224 225 226 227 227 229

10

Hazardous Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.1 Adverse Effects of Lamp Flicker . . . . . . . . . . . . . . . . . . . . . 10.1.1 Lamp Flicker . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.1.2 Visual Adverse Effects . . . . . . . . . . . . . . . . . . . . . 10.1.3 Neurophysiological Adverse Effects . . . . . . . . . . . . 10.1.4 One Tool to Specify the Risk of Flicker? . . . . . . . . 10.2 Optical Radiation Hazards . . . . . . . . . . . . . . . . . . . . . . . . . . 10.2.1 Photochemical and Thermal Damage . . . . . . . . . . . 10.2.2 Blue Light Hazard . . . . . . . . . . . . . . . . . . . . . . . . 10.3 Adverse Effects of Light at Night . . . . . . . . . . . . . . . . . . . . . 10.3.1 Circadian Disruption, Melatonin and Medical Disorders . . . . . . . . . . . . . . . . . . . . . 10.3.2 Breast Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . .

233 233 234 237 242 244 246 246 247 253

. 253 . 254 . 257

Contents

Part II 11

xiii

Technology

Lamps, Gear and Drivers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.2 Performance Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . 11.2.1 System Efficacy . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.2.2 Lifetime and Lumen Depreciation . . . . . . . . . . . . . . 11.3 Incandescent Lamps: GLS . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.3.1 Working Principle . . . . . . . . . . . . . . . . . . . . . . . . . 11.3.2 Construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.3.3 Lamp Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.4 Incandescent Lamps: Halogen . . . . . . . . . . . . . . . . . . . . . . . . 11.4.1 Working Principle . . . . . . . . . . . . . . . . . . . . . . . . . 11.4.2 Construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.4.3 Lamp Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.5 Gas-Discharge Lamps: Tubular Fluorescent . . . . . . . . . . . . . . 11.5.1 Working Principle . . . . . . . . . . . . . . . . . . . . . . . . . 11.5.2 Construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.5.3 Lamp Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.6 Gas-Discharge Lamps: Compact Fluorescent . . . . . . . . . . . . . . 11.6.1 Working Principle . . . . . . . . . . . . . . . . . . . . . . . . . 11.6.2 Construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.6.3 Lamp Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.7 Gas-Discharge Lamps: Induction . . . . . . . . . . . . . . . . . . . . . . 11.7.1 Working Principle . . . . . . . . . . . . . . . . . . . . . . . . . 11.7.2 Construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.7.3 Lamp Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.8 Gas-Discharge Lamps: Compact Metal-Halide . . . . . . . . . . . . 11.8.1 Working Principle . . . . . . . . . . . . . . . . . . . . . . . . . 11.8.2 Construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.8.3 Lamp Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.9 Control Gear for Gas-Discharge Lamps . . . . . . . . . . . . . . . . . 11.9.1 Igniters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.9.2 Ballasts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.9.3 Dimmers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.10 Solid-State Lamps: LED . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.10.1 Working Principle . . . . . . . . . . . . . . . . . . . . . . . . . 11.10.2 Construction and Manufacturing . . . . . . . . . . . . . . . 11.10.3 Binning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.10.4 Lamp Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.11 Solid-State Lamps: OLED . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.11.1 Working Principle . . . . . . . . . . . . . . . . . . . . . . . . . 11.11.2 Construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.11.3 Lamp Properties . . . . . . . . . . . . . . . . . . . . . . . . . . .

263 263 265 265 266 272 273 273 274 275 275 275 276 277 277 280 280 282 283 283 284 285 285 286 286 287 287 287 288 289 290 290 292 292 293 298 304 304 306 307 308 309

xiv

Contents

11.12

Drivers and Dimmers for LEDs and OLEDs . . . . . . . . . . . . . 11.12.1 Drivers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.12.2 Dimmers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . .

309 309 311 312

12

Daylight . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12.2 Origin of Daylight . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12.3 Properties of Exterior Daylight . . . . . . . . . . . . . . . . . . . . . . . 12.3.1 Variability of Daylight . . . . . . . . . . . . . . . . . . . . . 12.3.2 Daylight Levels . . . . . . . . . . . . . . . . . . . . . . . . . . 12.3.3 Spectrum and Colour . . . . . . . . . . . . . . . . . . . . . . 12.3.4 Sky Luminance . . . . . . . . . . . . . . . . . . . . . . . . . . 12.4 Properties of Daylight in Interiors . . . . . . . . . . . . . . . . . . . . . 12.4.1 Interior Daylight Levels . . . . . . . . . . . . . . . . . . . . 12.4.2 Flow of Daylight . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . .

315 315 316 317 317 318 319 321 322 322 326 327

13

Luminaires . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.1 Photometrical Characteristics . . . . . . . . . . . . . . . . . . . . . . . . 13.1.1 Light Distribution . . . . . . . . . . . . . . . . . . . . . . . . . 13.1.2 Luminaire Efficiency and Utilisation Factors . . . . . 13.1.3 Luminance Distribution . . . . . . . . . . . . . . . . . . . . . 13.1.4 UGR Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.1.5 Shielding Angle . . . . . . . . . . . . . . . . . . . . . . . . . . 13.2 The Optical System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.2.1 Mirrors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.2.2 Prisms and Lenses . . . . . . . . . . . . . . . . . . . . . . . . 13.2.3 Total Internal Reflection (TIR) Optics . . . . . . . . . . 13.2.4 Diffusers, Louvres and Baffles . . . . . . . . . . . . . . . . 13.3 Thermal Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.4 Lifetime and Lumen Maintenance . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

329 330 330 333 337 337 341 342 342 344 345 347 347 348 349

14

Connected Smart Lighting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.1 Lighting Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.1.1 Switching and Dimming . . . . . . . . . . . . . . . . . . . . . 14.1.2 Lighting Control Strategies . . . . . . . . . . . . . . . . . . . 14.1.3 User Satisfaction . . . . . . . . . . . . . . . . . . . . . . . . . . 14.2 Smart Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.2.1 Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.2.2 Network Topologies . . . . . . . . . . . . . . . . . . . . . . . . 14.2.3 Wired and Wireless Transmission . . . . . . . . . . . . . . 14.2.4 Communication Protocols . . . . . . . . . . . . . . . . . . . . 14.3 Power over Ethernet (PoE) . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

351 352 352 353 353 357 357 357 358 359 367 368

Contents

15

xv

Light Beyond Illumination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15.1 Smart Lights and Internet of Things . . . . . . . . . . . . . . . . . . . . 15.1.1 Internet of Things (IoT) . . . . . . . . . . . . . . . . . . . . . 15.1.2 IoT Lighting System . . . . . . . . . . . . . . . . . . . . . . . . 15.1.3 Services of IoT Lighting Systems . . . . . . . . . . . . . . 15.2 Visible Light Communication, VLC . . . . . . . . . . . . . . . . . . . . 15.2.1 Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15.2.2 Data Push Applications . . . . . . . . . . . . . . . . . . . . . . 15.2.3 Indoor Navigation . . . . . . . . . . . . . . . . . . . . . . . . . . 15.2.4 Light as Sensor . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15.2.5 Li-Fi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Part III

371 371 371 373 374 375 375 379 380 381 382 383

Application

16

Lighting Quality and Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.1 Lighting Quality Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . 16.1.1 Lighting Level and Uniformity . . . . . . . . . . . . . . . . 16.1.2 Glare Restriction . . . . . . . . . . . . . . . . . . . . . . . . . . 16.1.3 Face Recognition and Modelling . . . . . . . . . . . . . . . 16.1.4 Colour Appearance . . . . . . . . . . . . . . . . . . . . . . . . . 16.1.5 Colour Rendering . . . . . . . . . . . . . . . . . . . . . . . . . . 16.2 Standards and Recommendations . . . . . . . . . . . . . . . . . . . . . . 16.2.1 ISO-CIE Standard . . . . . . . . . . . . . . . . . . . . . . . . . . 16.2.2 European Standard . . . . . . . . . . . . . . . . . . . . . . . . . 16.2.3 North American Standard . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

389 389 390 393 393 394 395 395 396 396 402 406

17

Design Aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17.1 The Design Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17.1.1 Analysis of the Lighting Function . . . . . . . . . . . . . . 17.1.2 Determination of Lighting Quality . . . . . . . . . . . . . . 17.1.3 Choice of Lighting and Control System . . . . . . . . . . 17.1.4 Choice of Lamp, Luminaire and Control Type . . . . . . . . . . . . . . . . . . . . . . . . . . 17.1.5 Determination of Number and Positions of Luminaires . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17.2 Application-Specific Aspects . . . . . . . . . . . . . . . . . . . . . . . . . 17.2.1 Office Lighting . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17.2.2 Industrial Lighting . . . . . . . . . . . . . . . . . . . . . . . . . 17.2.3 Classroom Lighting . . . . . . . . . . . . . . . . . . . . . . . . 17.2.4 Lighting for Healthcare Institutions . . . . . . . . . . . . . 17.2.5 Emergency Lighting . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

407 408 408 409 410 410 410 411 411 415 417 420 423 428

xvi

18

Contents

Calculations and Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . 18.1 Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18.1.1 The Lumen Method . . . . . . . . . . . . . . . . . . . . . . . . 18.1.2 Computerised Calculations . . . . . . . . . . . . . . . . . . . 18.2 Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18.2.1 Light Detectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18.2.2 Measuring Lamps . . . . . . . . . . . . . . . . . . . . . . . . . . 18.2.3 Measuring Luminaires . . . . . . . . . . . . . . . . . . . . . . 18.2.4 Measuring Lighting Installations . . . . . . . . . . . . . . . 18.2.5 Light-Logging Devices . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

431 431 432 433 433 433 437 439 442 444 445

Appendix A: Standardised Relative Spectral Sensitivity Values V(λ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 447 Appendix B: Calculation of x–y Chromaticity Coordinates . . . . . . . . . . 449 Appendix C: RVP Model of Weston . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453 Appendix D: RVP Model of Rea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455 Appendix E: Evector/Escalar Ratio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457 Appendix F: Position Indices, p . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 459 Appendix G: Groningen Sleep Quality Scale . . . . . . . . . . . . . . . . . . . . . 463 Appendix H: Normalised Formula for SVM according to CIE (2016) . . . . . . . . . . . . . . . . . . . . . . . . . . 465 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467 Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 469 Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 477

About the Author

Wout van Bommel worked for more than 35 years with the head office of Philips Lighting in the Netherlands in different lighting application functions. He became responsible for the company’s international lighting application know-how centre. He has carried out research into many different lighting subjects. Some concepts now used in international standards for lighting are based on his research work. For the period 2003–2007, Wout van Bommel has been president of the International Lighting Commission (CIE). He was president and is now an honorary member of the Dutch “Light and Health Research Foundation” (SOLG). Wout van Bommel was appointed consulting professor at the Fudan University of Shanghai in 2004 and external examiner of the Master Course “Light and Lighting” for the period 2008–2012 at the University College London (UCL-Bartlett Institute). He has published many papers in national and international lighting journals in different languages. All over the world, he has presented papers, has taught at universities and schools and has given many invited lectures at conferences. After his retirement from Philips Lighting, he advises lighting designers, researchers, companies and governmental bodies as an independent lighting consultant, this in addition to his lecturing and writing activities. In 2019, he was the first recipient of an award of the Dutch Lighting Society (NSVV) which got his name: the “Wout van Bommel Award”.

xvii

Abbreviations

ANSI CEN CFL CIE COB COG CRM CS DALI DLMO DMX DOE EC EEG EOG fMRI GLA GLF GLS GPS GSQS HVAC IARC ICNIRP IEC IEEE IESNA IoT ipRGC IPS

American National Standards Institute European Normalization Commission Compact Fluorescent Lamp International Lighting Commission Chip on Board Chip on Glass Customer Relation Management Circadian Stimulus Digital Addressable Lighting Interface Dim Light Melatonin Onset Digital MultipleXed Department of Energy (USA) European Commission Electroencephalography (brain wave activity) Electrooculogram (eye blinking pattern) Functional Magnetic Resonance Imaging Global Lighting Association Global Lighting Forum General Lighting Service (Lamp) Global Positioning System Groningen Sleep Quality Scale Heating, Ventilation and Air Conditioning International Agency for Research on Cancer of the WHO International Commission on Nonionizing Radiation Protection International Electrotechnical Commission Institute of Electrical and Electronics Engineers Illuminating Engineering Society of North America Internet of Things Intrinsic Photosensitive Retinal Ganglion Cell Indoor Positioning System xix

xx

ISO KSS LED Li-Fi LoS MICI MRSE OLED OOK PAI PCB PLC PoE PRC pRGC PSE PSG PSQI PVT PWM RDM RF SCENIHR SCN SMD SSL SVS SWSD TAIR TCP/IP TIR TLA ToF VAS-e VDU VLC VLP WHO Zhaga

Abbreviations

International Organization for Standardization Karolinska Sleepiness Scale Light-Emitting Diode Light Fidelity Length of Stay (in a hospital) Mean Indirect Cubic Illuminance Mean Room Surface Exitance Organic Light-Emitting Diode On-Off Keying (including: none-return-to-zero) Perceived Adequacy of Illumination Printed Circuit Board Powerline Communication Power Over Ethernet Phase Response Curve Photosensitive Retinal Ganglion Cell Power Supply Equipment Polysomnography (recording of sleep parameters) Pittsburgh Sleep Quality Index Psychomotor Vigilance Test Pulse-Width Modulation Remote Device Management Radio Frequency Scientific Committee on Emerging and Newly Identified Health Risks of the European Union Suprachiasmatic Nucleus or Nuclei (biological clock) Surface-Mounted Device Solid-State Lighting Subjective Vitality Scale Shift Work Sleep Disorder Target Ambient Illuminance Ratio Transmission Control Protocol/Internet Protocol Total Internal Reflection Temporal Light Artefacts (caused by lamp flicker) Time of Flight (of light) Visual Assessment Scale for Energy (Lee) Visual Display Unit Visible Light Communication Visible Light Positioning World Health Organization International Consortium developing interface specifications that enable interchangeability of LED light sources

Part I

Fundamentals

Chapter 1

Visual Mechanism

Abstract A visual sensation is the result of processes in the eye and brain. Light entering the eye is projected on the back of the inner part of the eye, the retina. The retina contains photoreceptor cells: cones and rods. Photopigments in these receptor cells absorb light, resulting in a chemical-electrical signal which travels down a nerve into the visual cortex part of the brain where the visual sensation is invoked. A small area of the retina around the axis of the eye, the fovea, only contains cone cells. The other, peripheral, areas have few cone and many rod cells. The cone cells in the fovea have a one-to-one nerve connection to the brain. Rod photoreceptor cells are located in the periphery of the retina. Many of them converge on a single ganglion cell. Consequently, foveal vision is sharp and peripheral vision is not sharp. The set of rods converging on the same ganglion (the receptive field of that cell) are processed through an opponent mechanism. Colour vision is possible because there are three types of cones, one with sensitivity for reddish, one for greenish and one for bluish light. A colour opponent mechanism processes their signals. Since we have just one type of rod cell, colour vision with rods is impossible. Cones are mainly active at lighting levels larger than some 5 cd/m2. Vision is then referred to as photopic. The spectral eye sensitivity curve V(λ) defined for photopic vision is the basis for all photometric units.

This chapter explains how the various components of the visual system function to produce visual perception under widely different lighting circumstances. The system uses photomechanical, photochemical and photoelectrical processes. The expression “photo” refers to the fact that light controls these processes. Photomechanical processes take place in the eye itself. Think of pupil and eye lens changes. Photochemical and photoelectrical processes take place in the photoreceptor cells located in the retina of the eye. These processes are essential for relaying messages from these photoreceptors to the area in the brain where the visual sensation is evoked. The visual lighting effects are a direct consequence of these processes. The lighting professional needs to have a basic knowledge of these processes to understand the relations between lighting and visual performance and comfort. Chapter 2 will discuss how the spectrum of light influences the perception of colours. © Springer Nature Switzerland AG 2019 W. van Bommel, Interior Lighting, https://doi.org/10.1007/978-3-030-17195-7_1

3

4

1 Visual Mechanism

The visual system also evokes non-visual biological effects. This chapter only discusses this subject there where there is a direct relationship with vision. Part II of this book (Chap. 5) will deal with the non-visual biological system.

1.1

Visual Sensation

Already around 1490 Leonardo da Vinci showed in an anatomical drawing that vision is the result of combined processes taking place in the eye and the brain (Fig. 1.1). Da Vinci’s picture represents Greek, Arab and Medieval views on these processes (Gross 1997). Da Vinci added his views based on dissections he carried out on human cadavers (McMurrich 1930). The lens, drawn by da Vinci circularly and centrally located in the eye, is not considered by him as having an optical function. Only in 1602 Johannes Kepler described the paths of light rays in the eye.

Fig. 1.1 Anatomical study by Leonardo da Vinci, approx. 1490 (Windsor Castle, Royal Library)

1.2 Optics of the Eye

5

He defined the eye as an optical device creating an inverted image at the back of the eye where light receptors (photoreceptors) are located. In da Vinci’s drawing, the middle part of the brain processes the light collected by the eye. He considered that part of the brain as the site that receives input from all five sense organs: “sensus communis” (common sense). Today we know that different locations in the brain process the different senses. In these locations, the effect of the senses is produced: the sensation. In the case of the sense of sight, this is the visual sensation representing the scene in front of the eye. The area in the brain where the final processing for vision takes place lies not in the middle of the brain, as da Vinci thought, but in the lower rear end of it. It is called the visual cortex. Recent research indicates that there exists some communication between the different sense areas of the brain: multisensory integration (Macaluso and Driver 2005; Witten and Knudsen 2005; Marrelec et al. 2008).

1.2

Optics of the Eye

Figure 1.2 shows a cross section of the eye. The outer surface of the eye, the sclera also called the white of the eye, consists of hard white tissue giving rigidity to the eye ball. It bulges out at the front where it is translucent. This part is named the cornea. Light enters the eye through the cornea and travels through a circular diaphragm formed by the iris. The colour of the iris tissue determines “the colour of the eye”. The “owner” of the eye depicted in Fig. 1.2 has brown eyes. The opening of the iris is called the pupil. A circularly shaped muscle running through the iris can change the size of this opening. The change is controlled by the amount of light entering the eye and is one of the mechanisms of adaptation to different light levels. Just as with a diaphragm of a camera the actual size of the pupil also influences the size of the area in front of the eye that is seen sharp: a smaller pupil size results in a larger field of depth. The eye lens projects an inverted image from the scene in front of the eye on the back of the inner part of the eye, the retina. An object on the line of sight, and thus on the axis of the eye, is projected at the position of the retina where the fovea is located. The shape of the lens is changed, dependant on where the eye focuses on, from flat to more spherical by a system of muscles. In this way, the eye can project a sharp image on the retina from a distant part of a scene (flat lens shape) or a more nearby part (more spherical shaped lens). This process of adjusting the lens in dependence of the distance viewed is called accommodation. The inner part of the lens contains transparent fibres with proteins of the crystalline type. This is the reason that the eye lens often is referred to as the crystalline lens. The large area of the eye behind the lens, called the vitreous body, is filled with a vitreous, transparent, colourless gel existing for some 99% of water. It presses the retina against the back of the eye so that it stays in place. The retina at the back of the eye can be compared with the light-sensitive film in conventional cameras or with CCD cells in electronic cameras which transform light into a chemical action (paper or plastic film) or an electric action (CCD cell). In the

6

1 Visual Mechanism

muscle

iris

cornea

blind spot

lens

pupil

optic nerve

retina

axis of eye

fovea

vitreous body circular muscle

retina

choroid sclera (white of eye)

Fig. 1.2 Cross section of the eye

retina of the eye, the projected image is transformed into neural activity. The actual transformation takes place in some 100–130 million photoreceptor cells located in the retina. The dark-brown layer behind the retina, called the choroid, absorbs the light which is not transformed into a neural action. In this way, disturbing internal reflection of light within the eye, stray light, is prevented. The choroid layer has the same function as the black interior of a camera. Another function of the choroid is nourishing the internal parts of the eye, in particular, essential for the retina. For this purpose, it contains many blood vessels.

1.3

Retina and Photoreceptors

The retina is a very thin tissue with a thickness between 0.1 and 0.3 mm (about the thickness of two sheets of paper). It consists of many different layers. Figure 1.3 shows the most important ones. Surprisingly the layer of the photoreceptor cells, the cones and rods, is not located at the front but at the back of the retina. It means that light must pass through different layers of neuron cells (ganglion and collector cells) before reaching the photoreceptor cells. These layers also contain many blood vessels, in Fig. 1.3 indicated as pink shaded.

1.3 Retina and Photoreceptors

7

optic nerve into brain

light pRGC

fovea

ganglion cells

collector cells

cones and rods pigment epithelium

Fig. 1.3 Part of the retina with photoreceptor cells and their connections. Collector cells are bipolar, horizontal and amacrine cell types. The pink-shaded area represents blood vessels

The human photoreceptor cells for vision can be distinguished, according to function and geometry, into two categories of cells: cones and rods. The eye has some 5–6 million cones and 100–120 million rods. The cone and rod-shaped outer part of the photoreceptor cells have hundreds of thin membrane plates that contain photopigment molecules (Fig. 1.4). They are called opsins. The cones and rods have a different type of opsin called, respectively, photopsin and rhodopsin. The tips of the photoreceptors are in contact with the pigment epithelium layer at the back of the retina (Fig. 1.3). This layer provides vitamin A to the photoreceptors which chemically binds with the opsin molecules to make them

8

1 Visual Mechanism

Fig. 1.4 Cones and rod photoreceptor cells. From left to right: synapse through which electric-chemical signals pass, the nucleus and the lamellae membrane containing photopigment molecules, making the cone cells colour sensitive (Note that Fig. 1.3 shows the cells completely coloured to make them easier to distinguish)

photosensitive. Vitamin A is converted from beta-carotene, available from food as, for example, carrots. The opsin molecules can now absorb a photon of light and trigger a cascade of chemical reactions, ultimately changing the electrical state of the photoreceptor cell. This process is called phototransduction. The resulting electricchemical signal is transmitted through a series of different types of collector cells towards the ganglion cells. Here the signals undergo a first processing, and the result is forwarded through the optic nerve into the brain towards an area called visual cortex (Fig. 1.3). Here the signal is processed into a light and colour sensation (Weston 1949; Tovée 1996; Kolb 2003). While a photopigment absorbs a photon, the colour of the photoreceptor bleaches and the photoreceptor is temporarily depleted (exhausted). As a consequence, when, because of high lighting levels, a large part of the photopigment is bleached, the probability of photon absorption decreases and the cell is less active. In the case of rods, the probability tends to be zero at lighting levels more than some 5 cd/m2: the rods are saturated and no longer active as light receptors (Walraven et al. 1990). This is important because rods have a much larger sensitivity to light than cones. If rods remained active at high lighting levels, tremendous glare would be the unwanted result. When the light level decreases, a reverse chemical reaction reactivates them. This reactivating process is very slow. It is the reason that adaptation from high to low lighting levels may take minutes and in the case of adaptation to very dim light even half an hour. In this way, rods provide for vision only at low to extremely low lighting levels. Vision with rods is of low acuity and monochrome. Cones provide for high-acuity colour vision at higher lighting levels. Colour vision with cones is possible because there are three types of cones each with a different type of photopigment, one with maximum sensitivity for blueish, one for greenish and one for reddish light. The small thin and pit-shaped area of the retina around the axis of the eye with a diameter of around 1.5 mm is called the fovea (see again Fig. 1.3). It receives light from within a cone of approximately 2
centred on the axis of the eye (CIE 2010). The fovea only contains cone cells. The other, peripheral, areas of the retina have few cone cells and many rod cells. Cone cells in the fovea are much thinner and

1.4 Spectral Sensitivity

9

Table 1.1 Comparison of the foveal and peripheral retina (adapted from Mann 2016) Receptor type Convergence of cells Blood vessels and ganglion cells Sensitivity Lighting required to be active Function

Foveal retina Cones only None Not in front of the fovea Moderate Moderate–high Central vision Colour vision Detail vision

Peripheral retina Few cones and many rods Considerable In front of the peripheral area High Very low–low Peripheral vision Achromatic vision Poor detail vision

closer packed than cone cells outside the fovea. Figure 1.3 illustrates that every cone in the fovea has a one-to-one direct connection with the optic nerve and the brain. Outside the fovea, the area of peripheral vision, there is no such one-to-one direct connection with the brain. Here, many cells converge on a single collector cell, and different collector cells converge in turn on a single ganglion cell before the combined signal travels to the brain. Some 120 million photoreceptors converge on approximately 1 million ganglion cells. The convergence increases with distance from the fovea (Watson 2014). On the area outside the fovea, an average of some 100 rods and cones converge on the same ganglion cell. This combination of signals increases the sensitivity of rod vision considerably, making vision possible under very dim lighting conditions. On the other hand, the message sent to the brain loses information about the exact location from where the light originates. Consequently, peripheral vision with rods is highly sensitive at low light levels but results in blurred images and thus low-acuity vision. Foveal vision, on the other hand, results in sharp, high-acuity vision but needs higher lighting levels. This is because of the thinner, densely packed, cones and the one-to-one connection with the brain. High-acuity vision is here also helped by the fact that the fovea area is free from blood vessels and the ganglion cells are displaced to leave a clear area for the light to pass through (Fig. 1.3). This is the reason that the fovea has the shape of a pit with a rim and is the thinnest area of the retina. Table 1.1 summarises the most significant differences between the foveal and peripheral parts of the retina

1.4 1.4.1

Spectral Sensitivity Cones and Rods

As described in the previous section, the photopigments in the cones and rods (photopsin and rhodopsin) absorb incident light and transmit as a result of this an electric signal to the brain where then the visual sensation is evoked. The rods and three types of cones contain a different kind of photopigment. They each have a

10

1 Visual Mechanism

different absorption spectrum and thus different wavelength sensitivity. The photopigments of “blue” cones absorb light with relatively short wavelengths in the blue range of the spectrum. These cones are referred to as S-cones, where S stands for short wavelength. The S-cone type of opsin is called photopsin of the type cyanolabe, usually, more simply, referred to as the S-cone opsin. The spectral sensitivity of S-cone opsins has a pronounced peak in the short part of the spectrum (blue light) as shown in Fig. 1.5. Spectral sensitivity curves are sometimes shown on a linear scale (Fig. 1.5, top) and sometimes on a logarithmic scale with better information about the lower sensitivities (Fig. 1.5, bottom). The photopigments of the “green” and “red” cones provide a broader sensitivity spectrum (Fig. 1.5). The green one peaks in the middle-wavelength part of the spectrum and the red one in the long-wavelength part. Consequently, these opsins are called M-cone and L-cone opsins, formally called photopsins of the type chlorolabe and erythrolabe, respectively. The spectral sensitivities, shown in Fig. 1.5, are measured relative to light entering the cornea. They thus also take filtering by eye media in front of the retina, such as the eye lens, into account. The cone sensitivities are formally referred to as “cone fundamentals”. The sensitivity spectrum of rods coincides with the green-bluish part of the spectrum (Fig. 1.5). Their opsins are called rod-opsins (or rhodopsins). The sensitivity curves shown in Fig. 1.5 are valid for a 32-year-old observer. Fig. 1.5 Relative spectral sensitivities for light at the outer surface of the eye for the three types of cones and for the rods. Basis: 32-yearold standard observer; rod curve: CIE (2015); cone curves: CIE (2006). Bottom figure in logarithmic units

Relative Sensitivity S -cone

1.0

M-cone

0.8

L -cone 0.6

rod

0.4 0.2 0 400

450

500

550

600

650

450

500

550

600

650

700

λ (nm)

1

0.1

0.01 400

700

λ (nm)

1.4 Spectral Sensitivity

11

The spectral sensitivities of the three different cone types form the basis for colour vision. Since we only have one type of rod cell, colour vision with only rods is impossible: achromatic vision.

1.4.2

Photopic and Scotopic Vision

It has already been mentioned that successful transformation of light into signals to the brain is very much dependent upon the light level to which the eyes are adapted. At adaptation levels larger than some 5 cd/m2 the photopigments of the rods are completely inactive. What remains is pure cone vision. Vision under these conditions is called photopic vision. Lighting levels of most indoor lighting installations are in the photopic vision range. At lower lighting levels the activity of rods gradually increases until, at some 0.005 cd/m2, they have reached their maximum sensitivity. Therefore, at adaptation levels between 5 and 0.005 cd/m2 vision is a combination of rod and cone vision. Vision under these conditions is called mesopic vision. Most road lighting applications result in mesopic vision. At lighting levels lower than some 0.005 cd/m2 the sensitivity of the cones is far too small to play a role in vision so that only the rods determine vision. Vision under these conditions is monochrome and of low visual acuity. This type of vision is called scotopic vision. Only at night in areas without any artificially light and with no full moon vision is scotopic. The overall spectral sensitivity under photopic vision, important for interior lighting applications, is not a simple summation of the sensitivities of the three cone types as given in Fig. 1.5. Cones are not evenly distributed over the retina (Fig. 1.6). Blue S-cones are located more at the outside border of the fovea. The number of the three cone types is also different: there are substantially less blue S-cones (only 5–10% of all cones). The absolute sensitivity of blue S-cones is considerably lower than that of the other cone types. All this means that the overall spectral sensitivity is dependent upon the type of sensation or “action” the light is expected to deliver. Examples are creating brightness, enabling detection and enabling small detail vision. The size of the visual field plays a role: for a small field (often relevant for seeing small details) the fovea (with cones only) plays a bigger role than in the case of a larger visual field (often relevant in creating a bright environment). In 1924 CIE, as international standardisation body for light and lighting, defined a spectral sensitivity curve on the basis of the brightness of a 2
(foveal) visual field. The curve was published in 1926 as the CIE spectral luminous efficiency function V(λ) for photopic vision (CIE 1926). Figure 1.7 shows the V(λ) curve. The peak is at 555 nm in the yellow area of the wavelength range. Appendix A gives the values of this curve per 5 nm. For purposes where a larger visual field is relevant, CIE defined a spectral sensitivity curve for a 10
visual field (CIE 2005). In 1951, CIE published a spectral sensitivity curve for scotopic vision, again based on brightness, but here for a 20
visual field (CIE 1951).

12

1 Visual Mechanism

Fig. 1.6 Distribution of the S-cones (blue), M-cones (green) and L-cones (red) in the fovea

Fig. 1.7 Standardized relative spectral sensitivity curves V(λ) and V0 (λ) for photopic and scotopic vision, respectively (CIE 1926, 1951). Bottom figure in logarithmic units

Relative Sensitivity V(λ) V’(λ)

1.0

photopic scotopic

0.8 0.6 0.4 0.2 0 400

450

500

550

600

650

450

500

550

600

650

700

λ (nm)

1

0,1

0,01 400

700

λ (nm)

1.5 Receptive Fields

13

It is called the spectral luminous efficiency function V0 (λ) for scotopic vision. It peaks at a shorter wavelength of 507 nm (Fig. 1.7). The photopic spectral sensitivity curve V(λ) is used as a weighting function to convert radiometric units (such as radiant power expressed in watt) into photometric units (such as luminous flux expressed in lumen). The lumen is thus the radiant energy of visible radiation as “sensed” (weighted) by the eye under photopic vision conditions. The other light units, candela and lux, are also based on the same V(λ) curve.

1.5

Receptive Fields

Ganglion cells do not simply transmit the signals from the photoreceptors towards the brain. In the ganglion cells, the input signals from the photoreceptors undergo the first processing. The output of a ganglion cell is a series of voltage spikes (neural action potentials) that travel down the optic nerve towards the brain where further processing finally leads to the visual sensation. The output signal of the ganglion cells is determined by the rate of voltage spikes and not by the voltage value. Firing voltage spikes at different rates is the means with which communication, in general, takes place in the brain. Especially outside the fovea, many photoreceptor cells converge on a single ganglion cell. This is because the horizontal cell type of collector cells connects with several photoreceptors (Fig. 1.8). The total area of photoreceptor cells converging on a particular ganglion cell is called the receptive field of that cell. Kuffler (1953) was the first to map receptive fields by radiating tiny spots of light on the retina and measuring with miniature electrodes electrical activity in ganglion cells. Figure 1.8 shows a sketch of receptive fields of two different ganglion cells. Further down in the periphery of the retina, the receptive fields become gradually larger, because there more photoreceptors converge on a single ganglion cell. The sketch of Fig. 1.8 may suggest that the dimensions of a receptive field are large, but they are very small. Remember that the retina contains more than 100 million photoreceptors and 1 million ganglion cells. A so-called bipolar cell connects the centre part of the receptive field directly with the ganglion cell (see again Fig. 1.8). The other photoreceptors of the same receptive field have a more indirect connection via a horizontal cell which interconnects the photoreceptors of the receptive field. This area of the receptive field is referred to as the surround area. Signals from the receptive field are not just passed along but are processed by the ganglion cell to facilitate interpretation of a scene. For this interpretation, light-dark patterns or contrasts, particularly at the edges of objects and surfaces, are important. Ganglion cells compare signals arriving from an inner circular area of the receptive field with signals arriving from the outer circular area (the surrounds) of the same receptive field. One type of ganglion cell increases its output (“excites” is the term normally used) when the centre circle of its receptive field is illuminated but decreases (“inhibits”) its output when the surrounds are illuminated. This type of cell

14 Fig. 1.8 Receptive fields of two ganglion cells

1 Visual Mechanism Optic nerve

Ganglion cells

Receptive field of ganglion cell 1 Bipolar cells

Horizontal cells

1

Voltage spikes

2

Photoreceptors centre: surround:

is called an ON-centre ganglion cell. Figure 1.9b shows the situation of such a cell when the light completely overlaps with both the centre and the surrounding part of the receptive field. The ganglion cell shows only spontaneous activity, just as in the case with no light at all (Fig. 1.9a). Figure 1.9c shows the situation of an ON-centre ganglion cell when only the centre part of the receptive field is illuminated. The ganglion cell fires a series of voltage spikes of high rate that travel to the brain (an ON response). Contrary to this, when the light illuminates only the surround of the centre of the receptive field (Fig. 1.9d), the firing of spikes stops completely (an OFF response). When both the centre and the surrounds are partly illuminated (Fig. 1.9e), spikes fire at a rate higher than the spontaneous rate, but at a lower rate than in the situation with all light in the centre. When the centre and part of the surround are illuminated (Fig. 1.9f), again the cell fires at a rate higher than the spontaneous rate. There is another type of cell, called OFF-centre ganglion cell, that processes signals in the opposite way: light in the centre of the receptive field of an OFF-centre cell decreases or stops the firing of spikes while light in the surround increases the firing rate (Fig. 1.9g–i). The number of ON-centre and OFF-centre ganglion cells is more or less the same. The centre-surround ON-OFF processing by the retinal ganglion cells as described here enables detection of light-dark transitions and thus edge detection of bright objects or light sources. Figure 1.10 illustrates this by showing how a bright

1.5 Receptive Fields Fig. 1.9 Voltage firing rate of ON-centre and OFF-centre types of ganglion cells in dependence of a light spot (yellow circle) at different areas of the receptive field of the cell. The voltage-spike scale indicates with a yellow bar when the light spot is on. (a) Rest situation (no light) with spontaneous activity; (b–f) ON-centre type of ganglion cells; (g–i) OFF-centre type of ganglion cells

15

Firing rate

Ganglion cell

a

ON-centre cell spontaneous activity

+ time

Voltage spikes rate

b

ON-centre cell

spontaneous activity only

+

off

on

off ON-centre cell

c

+

off

on

off ON-centre cell

d

+

off

on

off ON-centre cell

e

+

off

on

off ON-centre cell

f

+

off

on

off OFF-centre cell

g

+ +

off

on

+ +

off

OFF-centre cell

h

+ +

off

on

+ +

off

OFF-centre cell

i

+ +

off

on

off

+ +

Receptive field

16 Fig. 1.10 Edge detection of an object or light source of uniform luminance (yellow circle) through the processing of receptive fields by retinal ganglion cells

1 Visual Mechanism +

+

+

+

+

+

+

+ + +

+

+

+

+

+

+ +

+

+

+

+

+

+

+

+

+

+

+

+ +

+

+

+

+

+

+ +

+

+

+

+

+

+ +

+

+ +

+ +

+

+

+ +

+ +

+ +

+

+ +

+

+

+

+ +

+

+

+ +

+

+

+

+

+ +

+ +

+ +

+

+

+ +

+

+

+ +

+

+

+

+ +

+

+

+

+

+

+ +

+

+

+

+

+

+ +

+ +

+

+

+

+

+ +

+

+

+ +

+

+

+

+

+

+

+

+ +

+

+

+

+ +

+

+ +

+

+

+

+ +

+ +

+

+

+

+ +

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+ +

+ +

+

+

+ +

+ +

+

+

+

+

+ +

+

+

+

+ +

+

+

+ +

+

+

+

+ +

+

+

+

+

+

+

+

+ +

+

+

+ +

+ +

+

+

+

+

+ +

+

+

+ +

+

+ +

+ +

+

+ +

+

+

+

+ +

+

+ +

+

+ +

+

+ +

+

+

+ +

+

+ +

+

+

+

+

+

increased firing rate activity suppressed spontaneous activity only

circular object (or light source) of uniform luminance interacts with a group of receptive fields of neighbouring ganglion cells. The bright object (light source) overlaps entirely with the inner ganglion cells. The output activity of these inner cells is consequently suppressed to spontaneous activity only, as is shown in Fig. 1.9b. Many ganglion cells that overlap with the edges of the object overlap partly with both the centres and surrounds. Their output activity is, therefore, higher, as shown in Fig. 1.9e, f, h, i. By processing the information from the receptive fields in this way, the edge of the object is detected. Only that information is forwarded to the visual cortex. Since the retina contains more than a million ganglion cells, the edge of the bright object overlaps with a much larger number of ganglion cells than sketched in Fig. 1.10. The ganglion cells are also much more densely packed. Therefore, the edges of the objects or light sources are in reality very sharply marked by the ganglion cells. From the uniform parts of the object or light source, less information is forwarded by the ganglion cells to the brain so that the brain is less engaged. A multitude of smaller bright objects of light sources (as matrix LED luminaires) excite more ganglion cells by the larger number of edges and consequently engage the brain more. Image processing computer software for image compression and automatic object recognition use a similar process.

1.6 Colour Vision

17

Glare has much to do with light-dark transitions. Chapter 4 (Sect. 4.4) will discuss recent studies that take the centre-surround processing of receptive fields by ganglion cells, as a fundamental-physiological basis for the development of glare models. The opponent mechanism of ON-OFF signal processing in ganglion cells is also typical for colour processing of visual information further down the visual pathway in the brain itself. It is discussed in the next section.

1.6

Colour Vision

The fact that the fovea of the retina contains three different types of cones, each with a different sensitivity in the short (S), medium (M) and long (L) wavelength range, makes that we can see colours. The spectral sensitivities have been shown in Fig. 1.5. The three-dimensional character of colour vision is called the trichromatic theory. It is also referred to as the “Young-Helmholtz theory” after the nineteenth-century developers of the theory. Young and Helmholtz demonstrated that each colour can be produced by mixing different amounts of light of the three primary colours: red, green and blue (RGB). With two colours this is impossible, while more than three colours are not needed. Mixing different colours of light is called additive colour mixing, this in contrast with mixing paint that is called subtractive colour mixing. With additive mixing, the result is brighter than the individual components and white can be obtained (Fig. 1.11, left). With subtractive mixing, the result is darker than the individual components (the paints absorb light) and eventually black is obtained (Fig. 1.11, right). As has been discussed, each cone in the fovea connects to its own ganglion cell in the retina. Thus, contrary to rods, cones do not converge in the retina. They do, however, converge on other type of ganglion cells further down the visual pathway at a kind of substation located in the central part of the brain (the thalamus). This substation is called lateral geniculate nucleus (LGN). Like the ganglion cells in the retina, these so-called LGN-ganglion cells have a receptive field in the retina. This field consists of cone cells only. De Valois et al. (1966) were the first to map, with microelectrodes, the connection between LGN-ganglion cells in the brain and cones in the fovea. An LGN-ganglion cell processes the signals from the cones with a similar opponent ON-OFF mechanism as the retinal ganglion cells discussed in the Fig. 1.11 Left: Additive mixing of primary colours (RGB) gives all the complementary colours and white. Right: Subtractive mixing of the complementary colours gives the primary colours (but darker) and black light beams

paints

18

1 Visual Mechanism

previous section. Retinal ganglion cells process opposing brightness signals arriving from the rods in the receptive field (light “on” and light “off”, respectively). LGN-ganglion cells process opposing colour signals arriving from the cones in the receptive field. Already in the late nineteenth century, Hering (1878) suggested that colour vision is based on two pairs of opposing colours: blue and yellow on the one side and red and green on the other side. This suggestion came from the observation that the colours bluish-yellow and reddish-green do not exist, while bluish-green (purple) and reddish-yellow (orange) do exist. The fact that a person who is colourblind for blue is also colour-blind for yellow and that the same holds for colourblindness to red and green also supported Hering’s suggestion. The LGN-ganglion cells indeed process the signals from the cones by comparing the opposing colours blue and yellow and red and green, respectively (Hubel 1995). Figure 1.12 illustrates the mechanism. Yellow is available for processing by adding the signals from red-sensitive cones (L-type) and green-sensitive cones (M-type). A third class exists, namely that of opposing black and white in which the three signals of the red, green and blue sensitive cones are added. As in the case of the ON-centre and OFF-centre retinal ganglion cells, each opposing colour class also has two types of LGN-ganglion cells. Light of the right colour in the centre of the receptive field (+) increases the output of the cell (increases the rate of voltage spikes). Light of the right colour in the surrounds of the receptive field () decreases (inhibits) the output signal (reduces the rate of spikes). Quite some physiological aspects of the colour opponent mechanism are not yet known (Conway 2014; Wuerger and Xiao 2016). Important discoveries are being made on further colour processing mechanisms in the visual cortex itself (Johnson et al. 2008; Harada et al. 2009; Wandell and Chichilnisky 2012; Mély and Serre 2017). Fig. 1.12 The opponent colour processing mechanism of LGN-ganglion cells receiving signals from cones. Light of the right colour in the centre (+) of the receptive field increases the voltage firing rate of the ganglion cell, while light of the right colour in the surrounds () decreases the voltage firing rate

B + Y cell

Y + B cell

Cones L

+

Blue

+

Yellow

M S

R + G cell

G + R cell

+

+

L

Red

Green

Black + White cell

M S

White + Black cell L

+

Black

+

White

M S

1.7 Pupillary Reflex

1.7

19

Pupillary Reflex

The section about photoreceptors mentioned that the photochemical bleaching and regeneration process taking place in the pigments of cones and rods is responsible for the possibility of the eye to adapt to a wide range of dark and light. Because of adaptation, vision is possible over a gigantic range, from a moonlit scene (with less than 0.1 lux) to a sunlit one (with even more than 100,000 lux). The relatively slow photochemical change in photoreceptors is not the only process that is responsible for adaptation. Fast neural changes account for the first few seconds of adaptation. Pupil size change in dependence on the amount of light (the pupillary reflex) also plays a role in the adaptation process, although only for a relatively small amount. The changing size of the pupil has, however, a significant influence on the quality of the retinal image, especially regarding the depth of field. Until recently, it was thought that it is the rods that control the pupil size. One of the reasons for this speculation was that the pupil does not change in a restricted field of view situation when rods, being located in the periphery of the retina, cannot play a role. As recent as 2002, a novel photoreceptor type has been discovered in the retina of the eye (Berson et al. 2002). These cell types appear to have a more dominant role in changing the pupil size than rods. Surprisingly, these cells are ganglion cells. Ganglion cells are, just like rods, located in the periphery of the retina, outside the fovea. Of the circa 1 million ganglion cells of the eye, some 1–5% appears to be photosensitive (Lucas et al. 2013). These photosensitive ganglion cells have, contrary to cones and rods, no direct contact with the pigment epithelium layer of the retina. They contain their own opsin photopigment type, melanopsin, which makes them intrinsically photosensitive (Provencio et al. 2000; Lucas et al. 2003). They are called photosensitive retinal ganglion cells, pRGCs. These novel cells have no direct function for visual image forming. They have pathways to areas of the brain, different from the visual cortex. They are essential for synchronising circadian rhythms of the body and therefore of great significance for lighting effects on health. Part II of this book will discuss in detail about the subject light and health and the role of pRGCs. It is now known that a particular subtype of pRGC cells have a pathway towards an area in the brain (called OPN) known to play a role in driving the pupillary reflex (Lucas et al. 2001; Berson 2003; Chen et al. 2011; Güler et al. 2008; Takahashi et al. 2011). Rods, cones and pRGCs all contribute to controlling the pupil size (McDougal and Gamlin 2010; CIE 2015). Their relative contributions are not constant but change with, among other things, light level, light duration and light spectrum. Research on the interplay of cones, rods and pRGCs in driving the pupillary reflex is ongoing. Of course, the change of pupil size affects the amount of light reaching the retina and the field of depth of observers. Therefore, pRGCs play, apart from their essential role in the relationship between light and health, also a role in the relationship between light and vision.

20

1.8

1 Visual Mechanism

High-Level Vision

How we experience a scene is not solely dependent on the physical retinal image. Memory and experience also play a role. The photograph of a canyon, shown in Fig. 1.13, illustrates this. Because we have learned, unconsciously, that outdoors shadows are always cast by light coming from above, we see a river in the bottom of a canyon. After turning the photograph upside down, the brain explains the bright and dark patterns so that it seems that shadows are cast again by light from above. The upside-down photograph, wrongly, is seen as that of a mountain ridge. Objects or complete scenes have cues that, through experience, help us not only to detect but also to recognise them properly. Vision which includes cognitive processes that incorporate experience with objects, materials and scenes is called “high-level vision”. This is in contrast with vision that only relates to the retinal image (Kosslyn 1987; Adelson 2000; Cox 2014).

1.8.1

Perceptual Constancy

High-level vision is responsible for the phenomenon of perceptual constancy. This term is used for the ability of the visual system to perceive stable lightness and colour of objects and scenes under widely different lighting conditions. The most Fig. 1.13 Canyon changes into mountain ridge by turning photograph upside down

1.8 High-Level Vision

21

Fig. 1.14 Crumpled paper experienced as having the same lightness, in spite of the shadows with lower illuminance and luminance values

commonly used example is that of black coal, which continues to be experienced as black when the light level changes from very low to extremely high. We experience a piece of white paper that receives so little light that its luminance is lower than that of brightly lit black coal adjacent to it, still as white. A painted wall with uneven lighting is experienced as a wall of uniform colour, not as a wall that is irregularly painted with different tints of paint. The lightness property of the wall, like the blackness property of the coal, remains constant. High-level vision neglects here, for an important part, the illumination, but appraises the reflective property of the wall. The same holds for the piece of crumpled paper shown in Fig. 1.14 experienced as the same white, notwithstanding the dark shadows on the paper. Lightness is one of the fundamental attributes of objects, which possess perceptual constancy. Colour is another object attribute that knows perceptual constancy. Here chromatic adaptation plays a major role in how we perceive colours. Chromatic adaptation keeps colours of a scene partially constant under changing colour and amount of the illumination. Experience may help, additionally, with colour constancy. A banana lit with a bad colour-rendering lamp may not look appetising, but we easily recognise it as a yellow banana.

1.8.2

Maintaining Constancy

Lighting can help secure visual constancy or can break it (Lynes 1971; Coaton and Marsden 1997; Cuttle 2008; Boyce 2014). The example of the canyon picture of Fig. 1.13 has already shown that unusual sharp shadows can break perceptual constancy. Many optical illusions are based on breaking perceptual constancy. Stage lighting is sometimes designed on purpose to break perceptual constancy to create drama. The “bad man effect” obtained with strong lighting from below the actor’s face is an example of this. Similarly, in display lighting, breaking perceptual

22

1 Visual Mechanism

constancy may be used to attract attention. Of course, in daily life, it is important to maintain perceptual constancy. Cuttle (2008) describes this importance aptly: “Our lives would become chaotic if objects changed from black to grey to white when carried from shade to full light”. Lynes (1971, 1994) gives recommendations for lighting a scene to help maintain perceptual constancy: • Adequate illuminance, also on the surroundings of the object or scene being viewed • Good colour rendering • Limit glare • Obvious, but not necessarily visible, light sources • Avoid sharp shadows • Reveal surface textures

References Adelson EH (2000) Lightness perception and lightness illusions. In: Gazzaniga M (ed) The new cognitive neurosciences, 2nd edn. The MIT Press, Cambridge, MA, pp 339–351 Berson DM (2003) Strange vision: ganglion cells as circadian photoreceptors. Trends Neurosci 26(6):314–320 Berson DM, Dunn FA, Takao M (2002) Phototransduction by retinal ganglion cells that set the circadian clock. Science 295(5557):1070–1073 Boyce PR (2014) Human factors in lighting, 3rd edn. CRC Press, Boca Raton, FL Chen SK, Badea TC, Hattar S (2011) Photoentrainment and pupillary light reflex are mediated by distinct populations of ipRGCs. Nature 476:92–95 CIE (1926) Proceedings CIE 6th Session 1924, Geneva. Recueil des Travaux et Compte Rendu de Séances. University Press, Cambridge, pp 67–69 CIE (1951) Proceedings CIE 12th Session, Stockholm, vol 3, pp 37–39 CIE (2005) CIE Publication 165:2005, CIE 10 degree photopic photometric observer CIE (2006) CIE Publication 170-1:2006, Fundamental chromaticity diagram with physiological axes—part 1 CIE (2010) CIE Publication 191:2010, Recommended system for visual performance based on mesopic photometry CIE (2015) CIE Technical Note 003:2015, Report on the first international workshop on circadian and neurophysiological photometry, 2013 Coaton JR, Marsden AM (1997) Lamps and lighting, 4th edn. Arnold, London Conway BR (2014) Color signals through dorsal and ventral visual pathways. Vis Neurosci 31:197–209 Cox DD (2014) Do we understand high-level vision? Curr Opin Neurobiol 25:187–193 Cuttle C (2008) Lighting by design, 2nd edn. Architectural Press, Oxford De Valois RL, Abramov I, Jacobs GH (1966) Analysis of response patterns of LGN cells. J Opt Soc Am 56:966–977 Gross CH (1997) Leonardo da Vinci on the brain and eye. Neuroscientist 3:347–354 Güler AD, Ecker JL, Lall GS, Haq S, Altimus CM, Liao HW, Barnard AR, Cahill H, Badea TC, Zhao H, Hankins MW, Berson DM, Lucas RJ, Yau KW, Hattar S (2008) Melanopsin cells are the principal conduits for rod-cone input to non-image-forming vision. Nature 453 (7191):102–105

References

23

Harada T, Goda N, Ogawa T, Ito M, Yoyoda H, Sadato N, Komatsu H (2009) Distribution of colour-selective activity in the monkey inferior temporal cortex revealed by functional magnetic resonance imaging. Eur J Neurosci 30(10):1960–1970 Hering E (1878) Zur Lehre vom Lichtsinn. Gerold, Vienna Hubel DH (1995) Eye, brain and vision, 2nd edn. Scientific American Library Series, no. 22, New York Johnson EN, Hawkes MJ, Shapley R (2008) The orientation selectivity of color-responsive neurons in macaque V1. J Neurosci 28:8096–8106 Kolb H (2003) How the retina works. Am Sci 91:28–35 Kosslyn SM (1987) Seeing and imagining in the cerebral hemispheres: a computational approach. Psychol Rev 94(2):148–175 Kuffler SW (1953) Discharge patterns and functional organization of mammalian retina. J Neurophysiol 16:37–68 Lucas RJ, Douglas RH, Foster RG (2001) Characterization of an ocular photopigment capable of driving pupillary constriction in mice. Nat Neurosci 4(6):621–626 Lucas RJ, Hattar S, Takao M (2003) Diminished pupillary light reflex at high irradiances in melanopsin-knockout mice. Science 299(5604):245–247 Lucas RJ, Peirson SN, Berson DM, Brown TM, Cooper HM, Czeisler CA, Figueiro MG, Gamlin PD, Lockley SW, O’Hagan JB, Price LA, Provencio I, Skene DJ, Brainard GC (2013) Measuring and using light in the melanopsin age. Trends Neurosci 37(1):1–9 Lynes JA (1971) Lightness, colour constancy in lighting design. Lighting Res Technol 3:24–42 Lynes JA (1994) Daylight and the appearance of indoor surfaces. In: Proceedings CIBSE National lighting conference, Cambridge, England, pp 98–110 Macaluso E, Driver J (2005) Multisensory spatial interactions: a window onto functional integration in the human brain. Trends Neurosci 28(5):264–271 Mann MD (2016) The nervous system in action. Webbook Marrelec G, Bellec P, Krainik A, Duffau H, Pelegrini-Isaac M (2008) Multisensory regions, systems, and the brain: hierarchical measures of functional integration in fMRI. Med Image Anal 12(4):484–496 McDougal DH, Gamlin PD (2010) The influence of intrinsically-photosensitive retinal ganglion cells on spectral sensitivity and response dynamics of the human pupillary light reflex. Vis Res 50:72–87 McMurrich JP (1930) Leonardo da Vinci the anatomist. Williams & Wilkins, Baltimore, MD Mély DA, Serre T (2017) Towards a theory of computation in the visual cortex. In: Zhao Q (ed) Computational and cognitive neuroscience of vision. Springer, Singapore Provencio I, Rodriguez IR, Jiang G, Hayes WP, Moreira EF, Rollag MD (2000) A novel human opsin in the inner retina. J Neurosci 20(2):600–605 Takahashi Y, Katsuura T, Shimomura Y, Iwanago K (2011) Prediction model of light-induced melatonin suppression. J Light Vis Environ 35(/2):123–135 Tovée MJ (1996) An introduction to the visual system. Cambridge University Press, Cambridge Walraven J, Enroth-Cugell C, Hood DC, MacLeod DIA, Schnapf JL (1990) The control of visual sensitivity: receptoral and postreceptoral processes. In: Spillman L, Werner JS (eds) Visual perception: the neurophysiological foundations. Academic Press, New York Wandell BA, Chichilnisky EJ (2012) Squaring cortex with color. Nat Neurosci 15:809–810 Watson AB (2014) A formula for human retinal ganglion cell receptive field density as a function of visual field location. J Vis 15:1–17 Weston HC (1949) Sight light and efficiency. Lewis & Co, London Witten IB, Knudsen EI (2005) Why seeing is believing: merging auditory and visual worlds. Neuron 48(3):489–496 Wuerger S, Xiao K (2016) Color vision, opponent theory. In: Luo MR (ed) Encyclopedia of color science and technology. Springer, New York

Chapter 2

Colour

Abstract Solid-state light sources offer far more possibilities to engineer lamp spectra to suit different colour quality requirements than gas discharge lamps did. Accurate lamp colour specification based on perceived colour has therefore received renewed attention. This concerns in the first place the specification of different types of white light sources. Coloured LEDs are more and more used in interior spaces so that also an accurate specification of coloured light sources is needed. For the specification of chromaticity coordinates of light sources, the CIE x–y chromaticity diagram (CIE colour triangle) is the basis. It is based on the standard CIE 1931 colorimetric observer, defined with colour-matching functions. Correlated colour temperatures of light sources, as a characterisation of the tint of whiteness, are easily obtained from the x–y chromaticity coordinates. MacAdam ellipses, in a more uniform u0 –v0 chromaticity diagram, are the basis for the binning process in the LED manufacturing process. A wealth of new research on colour science is available as a basis to replace some colour concepts that have been developed between the 1930s and 1960s. New uniform three-dimensional colour spaces have been introduced. The CIECAM02UCS colour space is proposed as a basis for a novel two-metric colour-rendering system with a fidelity index Rf and a gamut index Rg. Here, Rg is a measure of colour saturation. Vector graphics visualise the colour properties of light sources. They represent an indispensable new tool for the lighting designer in the LED era.

2.1

Perceived Colour

Perceived colour is a subjective result of a physical colour stimulus in the form of direct light or light reflected from a surface. The stimulus and the observer condition together determine how colour is perceived. The stimulus is characterised by the spectrum and quantity of light and the size, shape and surroundings of the stimulus. The observer condition concerns especially the adaptation state of the observer. The colour stimulus of surfaces of objects is determined by the light reflected from its surface which, in turn, depends on the spectral composition of the surface and the © Springer Nature Switzerland AG 2019 W. van Bommel, Interior Lighting, https://doi.org/10.1007/978-3-030-17195-7_2

25

26

2 Colour R) (5Y e Hu

Hue (5G)

Hue (5R)

Lightness

Fig. 2.1 Visualisation of the three attributes of perceived colour: hue, saturation and lightness. Adapted from the Munsell Book of Colour (Munsell 1929). Munsell used the term “chroma” for saturation and “value” for lightness

Saturation

Saturation

spectral properties of the light incident on the surface. Perceived colour thus changes with the light incident on the surface. As a consequence, an object surface does not have one “real” colour. Three attributes can describe perceived colour: hue, saturation (or chroma) and lightness (or brightness in the case of light source colours). Figure 2.1 visualises these attributes for the colours green and red arranged according to the Munsell system for specifying surface colours (Munsell 1929). Munsell arranged the different hues as the pages of a book (e.g. 5G green, 5R red and 5YR yellow-reddish). Hue refers to the monochromatic spectral colours, red, yellow, green, blue and combination of adjacent pairs of these colours (including blue-red, i.e. purple). Saturation or chroma refers to the colourfulness or vividness of the actual hue. Lightness or brightness relates to the total amount of light of the colour stimulus. All colours meet at the axis of the Munsell book where they are perceived as tints ranging from black to grey to white. By gradually reducing the brightness of a colour stimulus, finally all colours appear black: a red colour via reddish brown, an orange colour via brown, yellow via yellowish brown, green via olive green and white via grey. The influence of brightness on perceived colour is substantial when the brightness of the colour stimulus changes while the remaining parts of the visual scene do not change. When the brightness of the whole scene together with the colour stimulus changes, the effect on the perceived colour is smaller. This is because of perceptualcolour constancy as discussed in Sect. 1.8.1 of Chap. 1.

2.2

Colour Specification

2.2.1

CIE x–y Chromaticity Diagram

2.2.1.1

Chromaticity Coordinates

In 1931 CIE defined a numerical system for the definition of colours (CIE 1932; Schanda 2007; ISO/CIE 2019b). The system allows for the calculation of colour

2.2 Colour Specification

27

0.9 λ (nm)

y

520 530

0.8

D

540

510

G

0.7

550 560

A

0.6

P

C

570

F

500

0.5

580

3000K 4000K 5000K 6000K 7000K

0.4

2000K

590 600

D50

10,000K

610 620

D65

20,000K

0.3

E

490

R

640 700

0.2

480

0.1

470 460

0 0

0.1

B 380 0.2

0.3

0.4

0.5

0.6

0.7

x

0.8

Fig. 2.2 CIE x–y chromaticity diagram with the spectral wavelengths on the spectral locus. Correlated colour temperature lines are indicated on the blackbody locus. E: equal-energy white point; D50 and D65: daylight chromaticity points; R, G and B: CIE colour-matching stimuli. Chromaticity coordinates of a point on a line (example: P) can be obtained by a mixture of two different spectra with chromaticity points lying on the line on either side of that point (e.g. points A and C or A and F). Point D indicates the dominant wavelength of point P

points, called chromaticity coordinates x and y, from the spectrum of a colour stimulus (a light source or light reflected from a coloured surface). The x–y chromaticity coordinates define the position of the colour stimulus in a rectangular colour diagram or colour triangle (Fig. 2.2). The x–y coordinates correspond to the values of the horizontal and vertical axis of the triangle, respectively.

28

2 Colour

The monochromatic spectral colours (indicated with the corresponding wavelengths) are located on the outer border of the colour triangle. It is called the spectrum locus. Here the colours are most saturated. Purples are no spectral colours but mixtures of red and blue or violet. Chromaticity coordinates falling in this area of the diagram can therefore not be characterised by a spectral wavelength. By moving inwards to the centre of the diagram, the colours become less saturated until in the centre of the triangle white is obtained. It is the point where all spectral colours, each with the same energy, are mixed. It is called the equal-energy white point (point E in the diagram). The chromaticity points of two standardised types of daylight sky, D50 and D65, are also shown. The CIE x–y colour triangle has been developed so that the different hues occupy large enough areas and the white point is located rather centrally in the diagram. The x–y chromaticity point of a mixture of light of two different spectra lies on the line connecting the chromaticity points of those two spectra. In the example shown in Fig. 2.2, the spectrum corresponding to colour point P can be obtained by a mixture of spectra corresponding to chromaticity points A and C. The relative quantities (luminances) of the two components A and C determine the exact location of point P on the line. The same colour point P can also be obtained by mixing, in the appropriate quantities, the spectra corresponding to other points on the line, for example, points A and F. Different lines can be drawn through point P giving more mixtures resulting in the same colour point. Thus, many different mixtures of spectra can give the same chromatic coordinates. The curve through the centre of the colour triangle is the blackbody locus (also referred to as Planckian locus). Section 2.3 discusses the blackbody locus and its practical meaning. Section 2.4 deals with the dominant wavelength, represented by point D.

2.2.1.2

Standard Colorimetric Observer

The chromaticity diagram is based on the trichromatic theory, the basic law of colour vision already dealt with in Chap. 1, Sect. 1.6. This law states that each colour can be produced by mixing different amounts of light of three primary colours: red (R), green (G) and blue (B). The CIE system is based on colour-matching tests done by observers in the late 20s of the last century (Wright 1928–29; Guild 1931). From these tests, CIE defined a so-called standard colorimetric observer. In the tests, observers were asked to match the colour of two test stimuli. Figure 2.3 shows the principle for the example of a light-blue test stimulus. The test field is a monochromatic spectral colour defined by its wavelength, λ. It has to be matched with the lower field by adjusting the amounts of three primary colours consisting of red (R), green (G) and blue (B) light. Primary colours used in these tests were the monochromatic wavelengths 700.0, 546.1 and 435.8 nm for R, G and B, respectively. Of course, for each different wavelength of the test field, different amounts of R, G and B are required. As a result of these tests, each monochromatic wavelength between 380 and 780 nm could be characterised by

2.2 Colour Specification

29

Fig. 2.3 Principle of colour-matching tests with red (R), green (G) and blue (B) primaries. For some test colours, a match is only possible by adding a little of the R primary to the test field (dashed lines)

Test field (λ)

Matching field

700.0 nm 546.1 nm 435.8 nm

Fig. 2.4 Colour-matching functions of the CIE 1931 RGB system expressed as relative sensitivity. Primary colours R, G and B defined by wavelengths 700.0, 546.1 and 435.8 nm, respectively

Relative Sensitivity 0.4 0.3

g(λ)

b(λ)

r(λ)

0.2 0.1 0 - 0.1 350

400

450

500

550

600

650

700 750

λ (nm)

the required quantities of R, G and B components. Figure 2.4 gives the required relative amounts of the R, G and B colours with the red, green and blue curves, respectively. The curves are referred to as colour-matching functions R, G and B. The fact that the red curve has a part with negative values is because some test colours can only be matched by adding a certain amount of the red primary to the test colour itself (indicated with the dashed red lines in Fig. 2.3). That some colours cannot be matched is a consequence of the actual R, G and B primaries chosen for the matching tests. Figure 2.2 shows these R, G and B primaries with the letters R, G and B. Colours located outside the triangle RGB cannot be created by a mixture of these primaries only. The thin grey lines in Fig. 2.4 give an example of how to match the light-blue colour (485 nm) of the test field shown in Fig. 2.3. To match that colour, the relative amounts for the matching field are 0.10 for blue and 0.05 for green; a relative amount of 0.06 of red has to be added to the test field itself.

30

2 Colour

In the final 1931 CIE system the R, G and B colours used in the matching tests have been converted, mathematically, into hypothetical colours referred to as X, Y and Z primary colours. They are formally called the X, Y and Z tristimulus values. These, so to say, imaginary, bluer than blue, greener than green and redder than red, X, Y and Z primary colours, are chosen so that their matching functions are always positive and that the value of Y is proportional to the luminance of the colour stimulus. The conversion formulas are X ¼ 0:49R þ 0:31G þ 0:20B Y ¼ 0:17697R þ 0:81240G þ 0:01063B Z ¼ 0:01G þ 0:99B The colour-matching functions corresponding to the hypothetical primary colours X, Y and Z can be determined by converting Fig. 2.4 with these formulas. These converted matching functions are shown in Fig. 2.5. They represent the colourmatching functions of the CIE standard colorimetric observer and are referred to as xðλÞ, yðλÞ and zðλÞ colour-matching functions. As the value of Y is proportional to the luminance of the stimulus, the yðλÞ colour-matching function equals the standardized spectral sensitivity curve for photopic vision, V(λ), given in Fig. 1.7. In 2007 the colour-matching functions of the CIE standard colorimetric observer of Fig. 2.5 are standardized in an ISO-CIE standard (ISO/CIE 2019a). As colour vision is only possible under photopic conditions, the standard colorimetric observer, of course, is a photopic observer. The 1931 CIE standard colorimetric observer has been standardised by a colour stimulus with a test field diameter of 2 . In 1964 CIE defined an additional set of colour-matching functions for a 10 visual field, referred to as x10 ðλÞ, y10 ðλÞ and z10 ðλÞ (CIE 1964). These functions are indicated in Fig. 2.5 with the dashed curves. They are also standardised in the ISO-CIE standard (ISO/CIE 2019a). Fig. 2.5 Colour-matching functions xðλÞ, yðλÞ and zðλÞ of the CIE 1931 standard colorimetric observer (2 field), expressed as relative sensitivity. Maximum value of yðλÞ set as 1. Dashed curves: 10 field

Relative Sensitivity 2.0 1.5 1.0

y(λ)

z(λ)

x(λ)

0.5 0 350

400

450

500

550

600

650

700 750

λ (nm)

2.2 Colour Specification

31

The physiological fundamentals behind the three-colour-matching functions are the spectral sensitivities of the three different types of cones. In Chap. 1 these so-called cone fundamentals are given in Fig. 1.5. In the time of the development of the colorimetric system, these cone fundamentals were not yet accurately known and could therefore not be used as a basis for defining the colour-matching functions. Stockman and Sharpe (1999) demonstrated much later that it is possible to derive colour-matching functions directly from the cone fundamentals. On this basis, CIE has published a report with colour-matching functions and chromaticity diagrams calculated from the cone fundamentals for field sizes between 2 and 10 (CIE 2015). This is important for colour research, but the 2 and 10 colour-matching functions described earlier remain the standard of the colorimetric observer.

2.2.1.3

From Three-Dimensional Space to Two-Dimensional Plane

All colours can be defined by its X, Y and Z components and presented in a threedimensional XYZ space. Figure 2.6 shows an example of such a three-dimensional presentation. Here each colour is defined by its X, Y and Z components. Principally, all colours that can be produced by three primary colours can only be presented in a three-dimensional space. However, a simplification towards a two-dimensional plane presentation is possible by neglecting the effect of differences in brightness of the colour stimulus and concentrating on hue and saturation of the colour sensation only. A two-dimensional plane presentation is, in fact, one cross section of the three-dimensional space of Fig. 2.6. For defining colours in the two-dimensional plane, it needs to have its own coordinates that relate to the primary colours X, Y and Z of the three-dimensional space. These coordinates are defined by the relative amounts of the tristimulus values X, Y and Z:

Fig. 2.6 Three-dimensional colour space XYZ. Adapted from Bouma (1971)

Y

X

Z

32

2 Colour

x¼

X XþY þZ

y¼

Y XþY þZ

With the two coordinates x and y, used in a rectangular coordinating system, the CIE 1931 x–y chromaticity diagram, already shown in Fig. 2.2, is constructed. With the definitions given above, the x–y coordinates can be calculated for each given spectrum. Appendix B summarises the calculation procedure and gives the CIE colour-matching functions in tabular form. Figure 2.6 illustrates that just one single two-dimensional plane does not provide all information on how colour is perceived. With lower Y values (i.e. with a lower brightness) the colours gradually become darker until they all appear as the same black. The importance of the two-dimensional plane presentation is that if the x–y chromaticity coordinates of different colour stimuli (let it be coloured surfaces or light sources) are the same, their perceived colours match. They do that at each brightness. It means that the chromaticity coordinates are suitable for the specification of the colour of light sources.

2.2.2

CIE u0 –v0 Chromaticity Diagram

The x–y chromaticity diagram is developed for the numerical description of colour stimuli, and it serves this purpose well. However, for the evaluation of colour differences, it is less suited. This is because the x–y diagram is not uniform as far as colour differences are concerned. An equal distance between two chromaticity points in the diagram does not represent at all locations in the diagram an equal perceived colour difference. Figure 2.7 in which the ellipses represent areas of just noticeable colour difference illustrates this. The ellipses are drawn on a ten times enlarged scale for ease of illustration of the effect. In the greenish part of the diagram, the colour coordinates can be varied much more than in the reddish and, especially, in the bluish part, before a colour difference is noticed. These ellipses have become known as the “MacAdam ellipses”. MacAdam (1942) measured, with many observers, the size and form of the ellipses and proposed a more uniform chromaticity diagram by a simple mathematical transformation of the x–y chromaticity coordinates. Largely based on MacAdam’s proposal, CIE defined in 1976 the “uniform chromaticity scale diagram (UCS)”. It uses u0 –v0 coordinates which are transformed by simple formulas from the x–y coordinates (CIE 1975; ISO/CIE 2016): u0 ¼

4x 2x þ 12y þ 3

2.2 Colour Specification Fig. 2.7 CIE x–y chromaticity diagram with MacAdam ellipses (representing a just noticeable colour difference). Dashed curve: blackbody locus

33

y 0.8

0.6

0.4

0.2

Ellipses 10 times enlarged

0 0

v0 ¼

0.2

0.4

0.6

x

0.8

9y 2x þ 12y þ 3

The symbol 0 is used in the u0 –v0 coordinates to differentiate from an earlier u–v chromaticity diagram with slightly different transformation formulas. Because of the transformation both the colour plane and the ellipses change from form and size as shown in Fig. 2.8. In the ideal situation, all ellipses should have the shape of perfect circles of the same size. From Fig. 2.8 it is clear that the ideal situation is not completely obtained. However, the ellipses are considerably more equal in size than those in the x–y diagram of Fig. 2.7. For evaluating colour differences, the u0 –v0 diagram is substantially better than the x–y diagram. One of the following sections of this chapter shows that the CIE colour rendering index Ra of light sources is based on the u0 –v0 chromaticity diagram. The section about LED light sources in Chap. 11 will describe the so-called binning process used in LED manufacturing. Binning is done to ensure that batches of LEDs have unnoticeable colour differences. This binning process also uses the u0 –v0 chromaticity diagram as a basis. Thanks to this LED binning process the expression “MacAdam ellipses” has become quite common again.

2.2.3

Colour Spaces

Two-dimensional chromaticity diagrams are used for many purposes in many different professions. For lighting, they are particularly important for the

34 Fig. 2.8 CIE uniform chromaticity scale diagram (u0 –v0 diagram) with MacAdam ellipses (representing a just noticeable colour difference). Dashed curve: blackbody locus

2 Colour 0.6

v'

0.4

0.2

Ellipses 10 times enlarged

0 0

0.2

0.4

0.6

u'

specification of the colour of light sources including the tint of whiteness of light sources. In some disciplines, the effect of lightness on the colour of surface colours is of great importance. A three-dimensional colour space is then needed to characterise colour surfaces with sufficient accuracy. Think of the colourant industry (producing pigments or dyes used to colour paints, textile, plastics, ceramics, food) but also the photocopier and display screen industry. There exist fundamental differences between colours of lights and colours of objects. Object colours include the colours of brown and grey, but colours of lights cannot produce these colours. Object colours have the attribute of lightness, while colours of lights have the attribute of brightness. Since coloured object surfaces are used as a basis for the determination of the colour-rendering quality of light sources, the effect of lightness on colour samples is also important for the lighting discipline. To serve the different needs of all these disciplines, CIE has defined some different three-dimensional colour spaces. These spaces have exotic names based on their chromaticity coordinates: WUV uniform colour space, CIELUV uniform colour space and CIELAB uniform colour space. The most recent and most advanced uniform colour space is called CIECAM02-UCS. Its name stands for CIE Colour Appearance Model. All these spaces are uniform colour spaces (UCS) in the sense that equal distances in the space represent approximately equal perceived colour differences. The WUV space is based on the u–v chromaticity coordinates (predecessors of the u0 –v0 coordinates). It was developed specifically for colours of lights. In non-lighting disciplines, it is little used today because the other colour spaces are more uniform. However, in the lighting profession, WUV is used since 1964 for the determination of the CIE colour rendering index Ra (Sect. 2.5 deals with this subject). CIELUV (specifically for colours of lights) and CIELAB (specifically for object colours) are about equally good as far as uniformity is concerned (Hunt and Pointer 2011). CIELUV is based on u0 –v0 chromaticity coordinates

2.3 Colour Appearance of White Light Fig. 2.9 Three-dimensional CIECAM02-UCS colour space

35 J plane view

b’

b’

a’ a’

(ISO/CIE 2016) while CIELAB uses L, a and b coordinates defined by a nonlinear mathematical transformation of the X, Y and Z primaries (ISO/CIE 2019c). Since a novel colour-rendering system for light sources, to be discussed later, uses the CIECAM02-UCS as a basis, its uniform space will be described here in some more detail. It is the most uniform model to date (Fairchild 2005; Luo et al. 2006; Schanda 2007; CIE 2019). Figure 2.9 shows an illustration of this space. The three chromaticity coordinates are plotted on three mutually perpendicular axes. Vertical axis J represents lightness. The horizontal planes with the a0 and b0 axes give the different hues arranged according to the opponent colour-processing mechanism in colour vision, discussed earlier in Chap. 1 (Sect. 1.6). The positive a0 axis represents reddish colours and the negative a0 axis the opponent greenish colours, while the positive and negative b0 axis represents yellowish and opponent bluish colours, respectively. Larger values of a0 and b0 , both in positive and negative directions, correspond to more saturated colours. CIELUV and CIELAB spaces use the same type of presentation, but of course with their own chromaticity coordinates along the axes. Colour spaces are particularly important for the evaluation of colour difference of different surface samples for quality control of products. As has already been mentioned, they are also used for the determination of colour-rendering properties of light sources. CIECAM02-UCS has a complex mathematical calculation model that allows for accurate calculation of colour differences. It is the only model that takes different viewing conditions (background and surroundings of the colour sample) and chromatic adaptation into account.

2.3 2.3.1

Colour Appearance of White Light Correlated Colour Temperature CCT

Colour temperature is often used to describe the warmer or cooler appearance of white light. The colour of the light of the whitish light source is compared with the colour of the light radiated by a so-called blackbody radiator. A blackbody radiator is

36

2 Colour

an ideal thermal radiator. Thermal radiators are light sources that emit light when brought to high temperature. Incandescent lamps are thermal radiators. The spectrum of a blackbody type of thermal radiator can exactly be calculated from its temperature (Planck’s law). Therefore, also the chromaticity coordinates x and y can exactly be calculated for each temperature. The curve through the centre of the colour triangle (Fig. 2.2) is the blackbody locus also referred to as Planckian locus. It passes through the chromaticity coordinates of a blackbody radiator of different temperatures. Note that the temperature of thermal radiators is expressed in Kelvin (equal to degree Celsius + 273). As to be expected from the behaviour of practical thermal radiators such as incandescent lamps, the blackbody locus runs from dark red at low temperatures to bluish-white at high temperatures. Light looks “white” if its colour is similar to the colour of a blackbody radiator between 2700 and 10,000 K. More precise: the colour of a blackbody is red at a temperature of 800–900 K and turns into yellowish-white (warm-white) between 2700 and 3000 K, cool-white around 5000 K and pale bluishwhite between 8000 and 10,000 K. Practical types of thermal radiators have chromaticity positions very close to those of the blackbody radiator. Therefore, their colour appearance can directly be specified by the colour temperature. Non-thermal whitish light sources may have spectra widely different from the continuous spectra of blackbody radiators. Their chromaticity points usually lie not on the blackbody locus of the chromaticity diagram. If the chromaticity point of such light source lies only slightly away from the blackbody locus (still in the whitish area of the chromaticity diagram), its colour can still be compared with that of a blackbody radiator. In that case, the term correlated colour temperature (CCT) is used. For this purpose, CIE has defined lines of constant CCT (Schanda 2007). These lines are calculated in the, for other purposes obsolete, u–v chromaticity diagram (with u ¼ u0 and v ¼ 2/3v0 ). In this diagram, the CCT lines run perpendicular to the blackbody locus (Fig. 2.10).

Fig. 2.10 Small part of the u–v chromaticity diagram around the area of the blackbody locus. Lines of constant correlated colour temperature (CCT) run perpendicular to the blackbody locus. The crosswise distance from the blackbody locus is indicated with Duv. The light source indicated with the black dot has a CCT value of 5500 K and a Duv value of +0.003

0.40

v

spectral lo cus 37

365

0

0.38

375

0.36

4000

T CC 5000

0.34

0.30

385 nm

2500

v

+Du

warm

blackbody locus

D uv

6000

0.32

380

3000 K

green-yellowish

pinkish

8000 10000 cool

0.28 0.26 0.16

0.18

0.20

0.22

0.24

0.26

0.28

0.30

u

2.4 Dominant Wavelength and Excitation Purity

37

The correlated colour temperature of a non-thermal light source is obtained by interpolating between the values of the two nearest CCT lines to the chromaticity point of the lamp (CIE 2004). The example indicated with the black dot in the u–v diagram of Fig. 2.10 thus has a CCT value of 5500 K. The CCT lines can simply be mathematically transformed into other coordinates and displayed in other chromaticity diagrams as has been done in the x–y diagram of Fig. 2.2.

2.3.2

Distance from Blackbody Locus, Duv

When the chromaticity point of a light source is further away from the blackbody locus, it quickly gets an unacceptable non-whitish tint. With points above the blackbody locus the light gets a green-yellowish tint and with points under the blackbody locus a pinkish one (see Fig. 2.10). A relatively new measure for the crosswise distance of a chromaticity point from the blackbody locus is called Duv. It is measured in the u–v chromaticity diagram (Ohno 2014). Points above the blackbody locus get a plus (+) sign and points below a minus (–) sign. The example of the lamp indicated with the black dot in Fig. 2.10 has a Duv value of 0.003. Duv is defined in an American Standard (ANSI 2017) and is being discussed by CIE. For a more accurate assessment of the white quality of a light source, it is undoubtedly a valuable complement to the CCT value. Recent studies on preferred chromaticity of white LED lighting concluded that chromaticity points slightly under the blackbody locus result in the highest preference (Dikel et al. 2014; Ohno and Fein 2014; Ohno and Oh 2016; Perz et al. 2016; Wang and Wei 2018). A Duv value of approximately 0.015 is optimal for the CCT range of 2700–6500 K. A possible cause may be that light sources with slightly negative Duv values render the human skin more pleasantly. Wei and Houser (2016) pointed to the fact that slightly negative Duv values are likely to have better colourrendering properties.

2.4

Dominant Wavelength and Excitation Purity

Where the correlated colour temperature is used to characterise the colour appearance of whitish light with a single value, dominant wavelength together with excitation purity can be used to characterise coloured light. The dominant wavelength is obtained by connecting the x–y chromaticity point of a coloured light source in the chromaticity diagram with a line going through the equal-energy white point E, as is done for the example of point P in Fig. 2.2 with the dotted line. Where the line crosses the nearest boundary of the diagram, the dominant wavelength is obtained at the intersection with the spectrum locus (CIE 2011). In the example of Fig. 2.2 point D shows a dominant wavelength of 513 nm for

38

2 Colour

chromaticity point P. The hue of the coloured light source is similar to the hue of monochromatic radiation of the dominant wavelength. Purples are no spectral colours but mixtures of red and blue or violet. Chromaticity coordinates falling in this area of the diagram can therefore not be assigned a dominant wavelength. The closer the chromaticity point of the coloured light source is to the boundary of the diagram the more saturated or pure the colour is. Excitation purity ( pe) is a measure of this aspect of the colour of a coloured light source. It is defined by the distance between the light source’s chromaticity point and the white point E, relative to the distance between point E and the spectral locus at the position of the dominant wavelength (CIE 2011). In the example of Fig. 2.2, pe ¼ EP/ED. Excitation purity values vary between 1 for the highest purity (points on the spectrum locus) and 0 for the lowest purity (points at white point E).

2.5

Colour Rendering

In addition to the colour appearance of a light source, the colour rendering of surfaces lit by a light source is also of great importance. Of course, this aspect concerns whitish light sources, i.e. light sources that are on or close to the blackbody locus of the chromaticity diagram. The colour-rendering goal to be reached with a whitish light source may be different for different types of lighting application. Often the target is “true” colour rendering. In such cases, colour fidelity is the goal. In some applications, a light source that shifts a particular colour or colours in a specific direction, for example towards more saturation and thus to more colourful colours, may make the visual scene more pleasant. The goal is then not “true” colour rendering but something like “preferred” or “flattery” colour rendering. It has become clear that a one-metric system for colour rendering is simply not enough for a complete evaluation of colour-rendering properties of LEDs with their widely different white-light spectra. A combination of a fidelity measure and a saturation measure is what is needed. The commonly used fidelity metric is the CIE general colour rendering index Ra. The next section describes the details of Ra. The Illuminating Engineering Society of North America has proposed a novel colour fidelity metric, with the name colour fidelity index, Rf. This metric, also recognised, for scientific use, by CIE will be described here as well. The Illuminating Engineering Society of North America has also introduced a colour saturation metric that relates to colour preference. That metric, called gamut index, Rg, meant to be used in conjunction with Rf, will be discussed as well. Colour preference is very much application dependent (Lin et al. 2017; Teunissen et al. 2017) and often needs more information than a one-value preference-related metric can give. For this purpose, graphs that visualise how surface colours change when illuminated with a particular light source are needed. Examples of such graphs will be shown at the end of this section.

2.5 Colour Rendering

2.5.1

39

General Colour Rendering Index Ra

In 1964 CIE defined as colour-rendering metric the CIE General Colour Rendering Index Ra based on a comparison of a set of eight colour samples with defined reference sources (CIE 1965, 1974, 1995). This metric is widely adopted and still today’s internationally recommended colour-rendering metric. Most colourrendering metrics discussed in the literature are based on a comparison of the colour rendering of the source in question with that of a reference light source. Three factors influence the quality of such a metric: the colour-surface samples used, the type of reference light source used and the calculation procedure to calculate the perceived colour difference of a sample lit with the actual light source and the reference source.

2.5.1.1

Colour Samples

The type and number of test colours chosen for the CIE system were based on tests with 164 hypothetical spectra being typical for the fluorescent lamps with the type of phosphors used at the time of investigation (Ouweltjes 1960). As a result, eight test colours of moderate chroma (saturation), approximately the same lightness and covering all hues, were chosen for specifying the general colour rendering index Ra. Figure 2.11, top, shows these samples as they appear approximately under daylight D65 (R1–R8). For more specific evaluations, CIE also recommends six so-called special colour samples, four of high saturation, one representing Caucasian skin and another one representing green leaves. Figure 2.11 also shows these colour samples (R9–R14). Lighting designers often consider particularly the saturated red colour sample (R9) a valuable addition to the general colour rendering index. The spectral reflectance of the samples defines their colours, i.e. the percentage of light they reflect for each wavelength. Figure 2.12 gives an example of the colour samples red, green and blue (R9, R11 and R12). For calculation purposes, the spectral reflectance of all samples is available in tabular form (CIE 1995).

R1

R2

R3

R4

R5

R6

R9

R10

R11

R12

R13

R14

R7

R8

Fig. 2.11 Colours used in the CIE standardised colour-rendering system as approximately seen under daylight D65. Colours R1–8 used for the general colour rendering index Ra and R9–14 for specific individual evaluations

40

2 Colour

Fig. 2.12 Spectral reflectance of CIE colour samples R9 (red), R11 (green) and R12 (blue)

Reflectance 1.0 0.8 0.6 0.4 0.2 0

400

2.5.1.2

450

500

550

600

650

700

λ (nm)

Reference Light Sources

The reference light source against which the actual light source has to be compared is dependent on the correlated colour temperature (CCT) of the light source to be tested. For light sources with CCT values lower than 5000 K, the reference light source is a blackbody radiator operating at the same CCT as the light source to be tested. For light sources with CCT values above 5000 K, the reference is a standardised daylight spectrum, again with the same CCT as the light source to be tested. These latter spectra are reconstituted from actual daylight measurements at different parts of the world. They are referred to as CIE reconstituted daylight D spectra. The colour temperature of reconstituted daylight D is indicated by an addition of two digits, like D50 or D65 for 5000 and 6500 K, respectively. The choice of reference light sources was a bit arbitrary. Nevertheless, there is some logic behind the selection of the reference sources defined for the CIE colour-rendering system. The blackbody reference is similar to incandescent lamps traditionally used in residential homes when there is little or no daylight penetration at all. The daylight reference is used for higher CCT lamps because they are often used in office and industrial type of buildings where daylight is combined with the artificial lighting during much of the time. One fundamental consequence of the system of reference light sources can be illustrated with a simple example. Consider two light sources, one with a CCT of 3000 K and the other with 6000 K. The spectra of these light sources, in this case, both give no colour deviations of the colour samples when compared against their reference source of 3000 and 6000 K, respectively (i.e. both seem to have “ideal” colour rendering). The perceived colour of the samples, however, looks different when comparing the two test sources against each other: the samples seen under the 3000 K source look a bit more yellowish and the samples under the 6000 K somewhat more bluish. Colour rendering defined on the basis of reference sources does not represent an absolute measure. This is a consequence of what has been explained in Sect. 2.1: an object surface does not have one “real” colour. For a lighting designer considering the colour rendering of his or her designs, it is

2.5 Colour Rendering

41

important to take both the effect of colour rendering index and correlated colour temperature into account.

2.5.1.3

Calculation Procedure

For the calculation of the colour deviation of each colour sample, lit with the test source and the reference source, respectively, the CIE (1964) WUV uniform colour space is used with its corresponding colour difference formula. As discussed in Chap. 1, chromatic adaptation to the colour of the light source that illuminates the samples influences perceived colours. Therefore, in a 1974 update of the calculation procedure (CIE 1974), a chromatic adaptation correction was added to the calculation procedure. The so-called Von Kries (1911) correction is used for this purpose. More accurate correction methods have been developed since, but are not incorporated because it might be confusing when well-known existing lamps get a slightly changed colour rendering index. The colour rendering index for each colour sample, Ri, is obtained by appropriate scaling of the calculated colour difference. The general colour index, Ra, is the calculated average of the results for the eight test colours (nos. 1–8 in Fig. 2.11). Scaling is done by a simple mathematical transformation so that the value in the case of no colour deviation (ideal colour rendering for the actual colour temperature) equals 100 and that warm white fluorescent lamps (with the technology of the 60s of last century) result in an Ra value of approximately 50. The CIE recommendation also states that in case of a calculation result lower than zero, that value is set at zero. (This indeed may happen, for example in the case of monochromatic light sources). In this way, the colour rendering index has a scale from 0 to 100 (no colour rendering to “ideal” colour rendering). When light sources have a smaller Ra difference than 5 points, the colour-rendering difference will, mostly, not be noticeable (CIE 1995).

2.5.1.4

Use of General Colour Rendering Index Ra

Most lighting standards worldwide use Ra to give colour-rendering recommendations for many lighting applications. For indoor lighting in situations where people are present for longer periods, such as residential, office, industrial and retail lighting, the minimum requirement is usually an Ra value of 80. The fact that Ra is based on an average of the rendering of eight colours means that it does not give information on the colour rendering of individual colours. Two lamps with the same Ra value may have different colour-rendering qualities for different colours. For high Ra values greater than 90, these differences are small. However, with decreasing value of Ra, the differences in the rendering of individual values are likely to become larger. As has been explained already, a direct comparison of colour-rendering indices is only possible between lamps in the same correlated colour temperature category. As

42

2 Colour

mentioned before, even light sources with an “ideal” Ra value of 100, but different CCT value, show colours differently: high CCT light sources make colours appear more bluish, and low CCT sources make them more reddish. Not in all lighting applications a high Ra value should be aimed at. In some applications, a good rendering of specific colours is more important than an average rendering of all colours. Poor rendering of, in particular, the colour red often results in a bad appreciation of an illuminated scene (Bae et al. 2015). Light with good rendering of the human skin or of meat or vegetables may be the first important aspect in certain applications. Checking the individual colour-rendering values, Ri, for the relevant colours of the set of eight colours used in Ra or the six additional special colours is then important.

2.5.2

General Colour Fidelity Index, Rf

From the description of the general colour rendering index Ra given above, it is evident that some models and procedures on which it is based are outdated. This concerns especially the WUV colour space and the chromatic adaptation correction used. The number and types of colour samples of the system are not uniformly distributed over the colour space and probably not representative of colours occurring in interiors. After the arrival of LEDs on the market, it was soon realised that Ra is not always a good and complete enough predictor of colour-rendering properties for this new category of light sources. Extensive research on improved colour fidelity measures and especially on two-metric colour-rendering evaluation methods has been carried out during the last decade. This has led to many proposals to replace or extend the CIE-Ra method (Hashimoto et al. 2007; Rea and Freyssinier-Nova 2008; Davis and Ohno 2010; Van Der Burgt and Van Kemenade 2010; Smet et al. 2011, 2013, 2016; Smet and Whitehead 2011; Li et al. 2012; Houser et al. 2013; David 2014; David et al. 2015; De Beer et al. 2015; Smet and Hanselaer 2015). The Illuminating Engineering Society of North America has published a Technical Memorandum, IES TM-30-15, with a method using a fidelity- and preferencerelated metric that synthesises many of the different studies’ results (IES 2015). The method also includes visualisation graphics. For details of the development history of this method, reference is made to David et al. (2015). The novel colour fidelity metric, proposed in IES TM-30-15, is referred to as general colour fidelity index, Rf. Since 2017 it is also recommended by CIE (2017) for “accurate scientific use”. The CIE recommendation incorporates some minor, but fundamental, improvements in the calculation method. For the colour-rendering characterisation of commercial lamps CIE, for the time being, still sticks to the CIE-Ra method. IES published in 2018 an updated version of TM-30 taking all improvements of CIE-2017 into account (IES 2018). So, as far as fidelity index is concerned IES TM-30-18 and CIE 2017 are now identical. Both the CIE and IES Publications provide access to a software calculation tool for the calculation of Rf. The most recent comparison

2.5 Colour Rendering

43

studies between the Ra and Rf system indeed show that Rf performs better than Ra (Wei et al. 2019; Royer 2018).

2.5.2.1

Colour Samples

For the calculation of the fidelity index Rf, 99 test colours are used. These were selected from a set of 105,000 widely different types of objects, natural and humanmade. The 99 colours finally selected are evenly distributed in the colour space and do not include extremes in saturation or darkness. The set is also what is called “spectrally uniform”: when the reflectance functions of all samples are combined they cover evenly enough the total wavelength range. It prevents engineering an unrealistic good colour-rendering spectrum by creating light contributions at specific small wavelength areas at which the total of samples do hardly reflect light or, the opposite, reflect exceptionally much light. The reflectance tables of all 99 colour samples are available in both the CIE and IES Publications. Figure 2.13 shows the colours of the 99 samples.

01

02

03

04

05

06

07

08

09

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

64

65

66

67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96

97

98

99

Fig. 2.13 Colours of the 99 samples used in calculating colour fidelity index, Rf, as approximately perceived when illuminated with D50 daylight (IES 2015, 2018; CIE 2017)

44

2.5.2.2

2 Colour

Reference Light Sources

The reference light sources are the same as those used for the calculation of Ra, with a minor change for the CCT range between 4000 and 5000 K. For light sources with CCT values lower than or equal to 4000 K, the reference light source is a blackbody radiator operating at the same CCT as the light source to be tested (same as for Ra). For light sources with CCT values at and above 5000 K, the reference is (again as for Ra) a standardised daylight spectrum, with the same CCT as the light source to be tested. To avoid a minor discontinuity at 5000 K, the reference source between 4000 and 5000 K is defined as a mix of light of a blackbody radiator and standardised daylight. The proportion of standardised daylight increases linearly from 0% at 4000 K to 100% at 5000 K.

2.5.2.3

Calculation Procedure

For the calculation of the colour deviation for each colour sample, lit with the test source and the reference source, respectively, the CIECAM02-UCS uniform colour space is used with its corresponding colour difference formula. It has a built-in chromatic adaptation correction, which is more accurate than the method used for Ra. As has been discussed in Sect. 2.2.3, CIECAM02-UCS is the most recent and most uniform colour space. Instead of using colour-matching functions for the 2 standard observer, as is done for Ra, the functions for the 10 standard observer are used, which are considered to be somewhat more accurate for the determination of colour fidelity. The colour fidelity index for each colour sample, Rf,i, is obtained from the calculated colour deviation. The general colour fidelity index, Rf, is obtained by averaging the results of the 99 individual test colours. Scaling is so defined that the final scale of Rf is the same as that of Ra: viz. from 0 to 100. Moreover, the scaling has also been defined so that the average value of Rf for commonly available lamps for general lighting (with Ra values larger than 60) is equal to the average value of Ra (IES 2018; CIE 2017). This is done to reduce confusion between the Rf and Ra systems (CIE 2017). The scaling also avoids the occurrence of negative values.

2.5.2.4

Use of General Fidelity Index, Rf

From the description above it is clear that Rf represents a fundamentally more accurate colour fidelity index than Ra does. For the light sources, today on the market, there exists a reasonably good correlation between Rf and Ra. Figure 2.14 demonstrates this. It shows a plot of the Rf and Ra values of 184 commercial lamps, consisting of 37 fluorescent lamps, 13 high-intensity discharge lamps and 134 LED lamps. The value of Ra 80 is worldwide in many lighting standards the lower limit for colour rendering for indoor lighting applications. The big light grey square in Fig. 2.14

2.5 Colour Rendering Fig. 2.14 Correlation between the general colour fidelity index, Rf, and general colour rendering index, Ra, computed from the spectra of 184 commercially available lamps with Ra values higher than 60

45 100

LED Gaso

Rf 90

80

Ra > 80 Rf < 80

70

60 60

70

80

90

100

Ra

shows that the number of commercial lamps fulfilling the 80-limit for Ra, but not for Rf, is small. It makes a changeover from the old system into the new one easier. When keeping in the new system the same limit of 80, there will be only a few existing lamps or installations that do fulfil specifications in the old system but don’t in the new one. The good correlation between Rf and Ra is in line with many studies that have shown that different proposals for improved colour fidelity measures do not give substantially different results from the Ra system (Smet et al. 2011; Houser et al. 2013; Gu et al. 2017; Teunissen et al. 2017). However, with the possibility of engineering LEDs that emit light with small spectrum lines at specific wavelengths (for example with the use of quantum dots), the colour fidelity index Rf method is more “future-proof” than the old Ra method. CIE (2017) recommends the use of Rf for accurate scientific use but not as a replacement of Ra, for rating light sources and use in international and regional standards. The CIE Publication states that extensive evaluation of Rf is needed, while there remain some technical issues for further research. The latter concerns especially CIECAM02-UCS, which is widely accepted, but still being researched and improved. IES-TM30 recommends for evaluation of colour-rendering properties of light sources, and the use of Rf in combination with the preference-related metric, gamut index, Rg, which is an integral part of IES-TM-30, and described in the following sections.

46

2 Colour

2.5.3

Gamut Index Rg

2.5.3.1

Gamut Area

All colour spaces and chromaticity diagrams are designed so that more saturated colours are located outwards and unsaturated colours inwards. The saturation aspect of colour rendering of a specific light source is thus directly dependent on the direction of the colour shift of colour samples: a shift outwards increases saturation, and a shift inwards decreases saturation. Figure 2.15 illustrates this. The chromatic coordinates of 16 different colour samples are projected on the a0 –b0 plane of the CIECAM02-UCS colour space. The samples are shown as they appear when illuminated with light sources A and B and their reference light source, respectively. With light source A, the majority of the colour samples shift outwards, so on average this light source increases saturation. With light source B the shift is mostly inwards, and saturation decreases on average. The area obtained by connecting the chromaticity points of the colour samples when illuminated with a specific light source is called the gamut area. From the above, it is clear that the larger the gamut area of a light source relative to a reference source, for a defined set of colour samples, the larger the saturation of colours obtained with that source. The gamut area itself (without comparing it with a reference source), or the gamut volume for a three-dimensional colour space, has been proposed as a colourrendering metric for light sources in various studies (Rea and FreyssinierNova 2008; Davis and Ohno 2010; Hunt and Pointer 2011; Boyce 2014). In the case of a monochromatic light source, all colour points would shift to the same

Fig. 2.15 Gamut areas based on the chromaticity coordinates of 16 colour samples for test light sources A and B and for the reference source, projected on the a0 –b0 plane of the CIECAM02-UCS colour space (A and B have, in this example, an equal CCT value and therefore the same reference source)

b’ A Ref B

a’

2.5 Colour Rendering

47

location: just one point in the diagram. Thus, the larger the gamut area for a given set of colour samples the better the colour rendition of a light source is. In IES TM-30 the gamut area index, based on a comparison of the gamut area of a light source with that of a reference source, is introduced as a colour rendering index for the saturation aspect of light sources.

2.5.3.2

IES Gamut Index Rg

The IES gamut index Rg (IES 2015, 2018) compares the size of the gamut area of the actual light source to be tested, with the gamut area of the reference light source of the same colour temperature. The reference source is the same as defined for the determination of the colour-rendering fidelity index Rf. Also, the same 99 colour samples are used as those employed for the calculation of Rf. To make working with so many colour samples practical they are divided into 16 groups (called bins) according to their position in the a0 –b0 plane. For each bin, the average of the a0 –b0 coordinates of the colour samples in that bin is calculated and used to construct the gamut area. In fact, the coordinates of the 16 colours shown in Fig. 2.15 are the averaged colour points of the 16 bins. The defined TM-30 limits of the 16 bins are shown in Fig. 2.15 with the white radial lines. Gamut index Rg is the percentage of the gamut area of the test source relative to the area of the reference source: Rg ¼ 100 

Gamut areatest source Gamut areareference source

An Rg value larger than 100 indicates an increase, on average, in saturation and a value smaller than 100 a decrease. An increase or decrease in saturation, of course, can only be obtained if the light source considered induces colour shifts relative to the reference source. It implies that Rg values deviating from 100 are always accompanied by an Rf value lower than 100. The IES method is developed explicitly as a coupled system of two metrics that characterise two different aspects of colour rendering: fidelity and saturation preference. The values of the two metrics cannot be maximised simultaneously. Depending on the actual application, the values have to be balanced. For light sources with Rf values larger than 80, Rg has approximate values between 80 and 120 as can be seen from Fig. 2.16 where the Rf and Rg values of the 184 commercial lamps, earlier used, are shown. Where the fidelity metric Rf is suitable for use in regulations and standards, the preference-related metric Rg is not. It is a tool for the lighting designer to help with the choice of that light source that blends best with the colours of the interior design of the space to be illuminated.

48

2 Colour

Fig. 2.16 Rf–Rg plot of 184 commercially available lamps

100 LED Gaso

Rf 90

80

70

60 80

2.5.4

90

100

100

Rg, IES

120

Colour Vector Graphics

A preference-related metric, such as the gamut index Rg, represents the overall effect of a light source on the saturation of colours. A detailed insight of which colours are saturated or desaturated and in which hue direction colours shift can only be obtained through visualisation with colour graphics. Van Kemenade and Van Der Burgt (1995) proposed the use of colour vectors for such visualisation as additional information to colour-rendering indices. Many different methods for the presentation of colour vectors have subsequently been proposed. Both IES (2015, 2018) and CIE (2017) have developed colour vector graphics.

2.5.4.1

IES Colour Vector Graphic

IES TM 30 normalises, for creating the colour vector graphic, the gamut of the reference light source to a circle in the a0 –b0 diagram (Fig. 2.17 top: black circle). The colour shift of the averaged chromaticity point of each of the 16-colour bins is indicated by an arrow (the vector) relative to the circle representing the reference source. The ends of the vectors are interconnected by a line so that both the size and direction of colour shift in each part of the diagram are visualised. Where the line lies outside the circle of the reference source, saturation increases, and where the line is inside the circle, saturation decreases. So, the light source represented on the left of Fig. 2.17 (top) saturates green-yellow colours strongly and red-orange colours a little. A small part of the bluish colours desaturates just a little. On average this light source results in strong saturation which is also clear from its high gamut index (Rg ¼ 120). The light source on the right (top) results in strong desaturation of a large part of the colours (Rg ¼ 85). Vectors more or less parallel to the circle indicate a hue shift and not, or not so much, a saturation shift.

2.5 Colour Rendering

IES TM 30 -15

49

IES TM 30 -15

Reference source Test source

Rf = 70 Rg = 120

CIE 224:2017

Rf = 70 Rg = 85

CIE 224:2017

Fig. 2.17 IES TM 30-15 (top) and CIE 224:2017 (bottom) colour vector graphics visualising colour shifts as obtained with the same two light sources relative to their reference sources

2.5.4.2

CIE Colour Vector Graphic

CIE (2017) developed an alternative graph for visualising colour shifts based on the same 99 colours of the IES method. Colours are also grouped in bins, here in eight bins, uniformly distributed over the hue range. A representative colour sample is assigned to each bin. The colour shift of each of these representative eight samples, when illuminated with the test light source relative to the reference light source, is calculated in terms of deviations of a0 and b0 coordinates. Corresponding vectors are presented on a hue circle (Fig. 2.17, bottom). The zero-shift position is the centre of the small circles which are drawn in the approximate colour of the central hue of

50

2 Colour

each of the eight bins. Individual shifts of the individual colours within each bin are indicated with grey dots. They give insight into the spread of colour shift within each bin. As in the IES diagram, a strong radial direction represents mainly saturation shifts and vectors more parallel to the circle, shifts in hue. While the CIE proposal contains more information about the spread of the results, which can be important for research reasons, the IES TM-30 graphic is more intuitive and therefore more suitable for general use.

2.5.5

Colour Discrimination

Colour discrimination is a special aspect of colour rendering. It describes whether slightly different colours, when seen simultaneously, can be distinguished as being different. This aspect of colour rendering is essential there where colour differences cannot be tolerated. Illustrative examples are the paint industry, the automobile damage repair sector and the dentist practice. For research purposes as well as for testing colour discrimination in practice, the Farnsworth-Munsell 100 Hue (FM-100) test is often used (Farnsworth 1957). This test was developed for testing colour vision deviations in persons. It makes use of a commercially available set of four trays, each containing 25 small cylinders of which the top surface is coloured. Versions with fewer numbers are also available (Fig. 2.18). Each cylinder in a tray has a slightly different hue. The trays contain magenta-orange, yellow-green, bluepurple and purple-magenta hues, respectively. The person carrying out the test has to place the randomly mixed cylinders in the right sequence of (apparent) hue. The score is calculated from the errors made in this sequence. Of course, a colour discrimination measure, calculated from the lamp spectrum, would facilitate the design of proper lighting, there where colour discrimination is critical. In different studies, such measures have been proposed, but comparative

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

64

65

66

67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

84 85

Fig. 2.18 Farnsworth-Munsell 100 Hue (FM-100) coloured test cylinders in a version with 85 pieces

2.5 Colour Rendering

51

studies have given inconclusive or contradictory results (Thornton 1972; Boyce 1976; Boyce and Simons 1977; Mahler et al. 2009; Royer et al. 2012). Esposito and Houser (2019) carried out a study using the Farnsworth-Munsell test (85 test colour version) with 24 systematically varied light source spectra with a wide range of general colour fidelity indexes Rf (from 65 to 95) and gamut indexes Rg (from 80 to 120) all with a CCT of approximately 3500 K. The tests were done with an illuminance of 650 lux on the samples. The study shows that all proposed colour discrimination metrics tried so far have only a small correlation with the score of the results of the FM-100 test (coefficient of determination R2 smaller than approximately 0.60). This includes metrics that combine colour fidelity index and gamut area. Based on the same study, the authors propose a new measure of colour discrimination, Rd. It quantifies the number of cylinder transpositions in the FM-100 test in comparison to the order of the cylinders when illuminated by CIE standard illuminant C. The authors tentatively propose the following interpretation of Rd scores. An Rd score of zero stands for superior colour discrimination (no transpositions), a score between 4 and 12 for “average” colour discrimination and a score of 16 or higher for “poor” colour discrimination. The authors are preparing a research paper that includes access to an Excel sheet for the calculation of Rd from the spectrum of a light source. This will permit other researchers to study and judge the new candidate metric for colour discrimination.

2.5.6

Surface-Colour Metamerism

Metamerism is the term used for the property of light sources having different spectra but the same chromatic coordinates and thus producing the same perceived colour. Surface-colour metamerism stands for the property of different colour samples lit by light sources with different spectra showing the same perceived colour. Such colour samples are called metameres. Fortunately, many coloured objects are metamere for many different light sources. Object colours that are not metamere may under one type of light source appear as having the same colour, while they appear as two different colours under a different kind of light source. Imagine how surprised you would be when buying in a shop a dress in two colours, and outside, under daylight, it suddenly appears to be a dress of one colour. Figure 2.19 shows an example of where this situation occurs. Reflection functions A and B combined with that of the daylight spectrum at the top result in the two objects perceived as the same brownish colour. The same objects combined with the artificial light spectrum at the bottom are seen entirely different: one still as brownish but the other as green. The high reflectivity of object A at long wavelengths (deep red) combines well with the relative high spectrum values of daylight at those wavelengths (top of Fig. 2.19). It makes the object appear as red-brownish. However, in that same wavelength area, lamp L has no radiation (bottom of Fig. 2.19). Where lamp L has a peak in orange-red, object A hardly has reflectivity The peak in the green area of lamp L combines well with the secondary

52

2 Colour

Fig. 2.19 Two non-metamere objects A and B, with reflection functions indicated by the dashed curves, perceived as the same brown colour seen under a daylight spectrum (top) but as two different colours (brown and green) seen under the spectrum of lamp L (bottom) (adapted from Van Der Burgt 2016)

Relative Reflectance (A and B) Relative Energy (Daylight and Lamp L) 1.0

A

A

B

B

Daylight 0.5

0 400

450

1.0

500

550

600

650

Lamp L

700 A

A

B

B

0.5

0 400

450

500

550

600

650

700

λ (nm)

peak in the reflection function of object A, making the object appear as green. A similar evaluation of the combination of the two spectra with the reflection function of object B shows that it appears as brownish under both spectra. Metamere mismatch may be a problem in lighting applications such as textile retail. For such applications, it is important to check if the problem arises with the products actually used. Some shop owners check articles regarding this aspect before purchasing them. The problem may also occur in medical environments where observing body tissues is critical. David et al. (2019) developed a metameric uncertainty index (Rt) that can be used to estimate the probability of noticeable metameric mismatches induced by a given light source. The calculation method of this index uses the framework of IES TM-30 which was discussed in the previous section.

2.6

Summary of Colour Metrics

All colour metrics discussed in this chapter are listed in Table 2.1 together with the aspect they try to specify. They are based on the standard colorimetric observer defined by the colorimetric matching functions of Fig. 2.5. All metrics can be calculated from the spectrum of the actual light source, provided that it is available in small enough wavelength steps (1 or 5 nm). Both the CIE and IES software calculation tools for the calculation of Rf also allow for the calculation of many other of the metrics shown in Table 2.1. Most modern spectrophotometers have embedded software that supplies the values of these metrics from the measured spectrum.

References

53

Table 2.1 Colour metrics and the aspects they specify Metric name Chromaticity coordinates Correlated colour temperature Distance from blackbody locus Dominant wavelength Excitation purity General colour rendering index General fidelity index Gamut index Discrimination index (candidate)

X–Y–Z; x–y; u–v; u0 –v0 ; L–a–b; a0 –b0 CCT Duv λD pe Ra Rf Rg Rd

Specification aspect Colour (most complete) Whiteness of light Whiteness of light Colour of coloured light Colour of coloured light True rendition of colours by light True rendition of colours by light Saturation of colours by light Colour discrimination

References ANSI (2017) ANSI Standard C78.377-2017. American standard for electric lamps—specifications for the chromaticity of solid state lighting products Bae G, Olkkonen M, Allred SR, Flombaum JI (2015) Why some colours appear more memorable than others; a model combining categories and particulars in color working memory. J Exp Psychol Gen 144(40):744–763 Bouma PJ (1971) Physical aspects of colour, 2nd edn. MacMillan and Co. Ltd., London Boyce PR (1976) Illuminance, lamp type and performance on a colour discrimination task. Lighting Res Technol 8:195–199 Boyce PR (2014) Human factors in lighting, 3rd edn. CRC Press, Boca Raton, FL Boyce PR, Simons RH (1977) Hue discrimination and light sources. Lighting Res Technol 9:125–140 CIE (1932) Colorimétrie, Resolutions 1–4. In: Recueil des travaux et compte rendu des séances, Huitième Session Cambridge—Septembre 1931, Bureau Central de la Commission, The National Physical Laboratory Teddington. Cambridge at the University Press, Cambridge, pp 19–29 CIE (1964) Proceedings of the CIE 15th Session, Vienna, vol A (CIE 11A) 1963:35. Bureau Central de la CIE, Paris CIE (1965) CIE Publication E-1.3.2: Method of measuring and specifying colour rendering properties of light sources, 1st edn., Vienna CIE (1974) CIE Publication 13.2: Method of measuring and specifying colour rendering properties of light sources, Vienna CIE (1975) Progress report of CIE TC-1.3 Colorimetry, Vienna. In: Proceedings CIE 18th Session, London, pp 161–172 CIE (1995) CIE Publication 13.3 Method of measuring and specifying colour rendering of light sources, 3rd edn., Vienna CIE (2004) CIE Publication 159:2004 A colour appearance model for colour management systems: CIECAM02, Vienna CIE (2011) CIE Publication CIE S 017:2011 ILV: International lighting vocabulary, Vienna CIE (2015) CIE Publication 170-2:2015 Fundamental chromaticity diagram with physiological axes—part 2: spectral luminous efficiency functions and chromaticity diagrams, Vienna CIE (2017) CIE Publication 224:2017: CIE 2017 Colour fidelity index for accurate scientific use, Vienna CIE (2019) CIE Publication 15:2019 Colorimetry, 4th edn., Vienna

54

2 Colour

David A (2014) Color fidelity of light sources evaluated over large sets of reflectance samples. LEUKOS 10(20):59–75 David A, Fini P, Houser KW, Ohno Y, Royer MP, Smet KAG, Wei M, Whitehead L (2015) Development of the IES method for evaluating color rendition of light sources. Opt Exp 23 (12):15888–15906 David A, Esposito T, Houser K, Royer M, Smet KAG, Whitehead L (2019) A vector field color rendition model for characterizing color shifts and metameric mismatch. LEUKOS 1–16. Prepublished Feb 1, 2019. https://doi.org/10.1080/15502724.2018.1554369 Davis W, Ohno Y (2010) Color quality scale. Opt Eng 49(3):033602-1–033602-16 De Beer E, Van Der Burgt P, van Kemenade J (2015) Another color rendering metric: do we really need it, can we live without it? LEUKOS 12:51–59 Dikel EE, Burns GJ, Veitch JA, Mancini S, Newsham GR (2014) Preferred chromaticity of colortunable LED lighting. LEUKOS 10(2):101–115 Esposito T, Houser K (2019) A new measure of colour discrimination for LEDs and other light sources. Lighting Res Technol 51:5–23 Fairchild MD (2005) Colour appearance models, 2nd edn. Wiley, Chichester Farnsworth D (1957) The Farnsworth-Munsell 100-Hue Test for the examination of colour discrimination: manual. Munsell Color Company, Grand Rapids, MI Gu HT, Luo MR, Liu XY (2017) Testing different colour rendering metrics using colour difference data. Lighting Res Technol 49:539–560 Guild J (1931) The colorimetric properties of the spectrum. Phil Trans R Soc Lond 230A:149–187 Hashimoto K, Yano T, Shimizu M, Nayatani Y (2007) New method for specifying color-rendering properties of light sources based on feeling of contrast. Color Res Appl 32(5):361–371 Houser KW, Wei M, David A, Kramer MR, Shen XS (2013) Review of measures for light-source color rendition and considerations for a two-measure system for characterizing color rendition. Opt Exp 21:10393–10411 Hunt RWG, Pointer MR (2011) Measuring colour, 4th edn. Wiley, Chichester IES (2015) Publication IES TM-30-15: IES method for evaluating light source color rendition. IES, New York IES (2018) Publication IES TM-30-18: IES method for evaluating light source color rendition. IES, New York ISO/CIE (2016) Publication 11664-5:2016 Joint ISO/CIE Standard: colorimetry—part 5: CIE 1976 Luv colour space and u0 , v0 uniform chromaticity scale diagram ISO/CIE (2019a) Publication ISO 11664-1:2019/CIE S 014-1:2019 Joint ISO/CIE Standard: colorimetry—part 1: CIE standard colorimetric observers ISO/CIE (2019b) Publication ISO 11664-3:2019/CIE S 014-3:2019 Joint ISO/CIE Standard: colorimetry—part 3: CIE tristimulus values ISO/CIE (2019c) Publication ISO 11664-4:2019 CIE S 014-4:2019 Joint ISO/CIE Standard: colorimetry—part 4: CIE 1976 Lab colour Space Li C, Luo MR, Li C, Cui G (2012) The CRI-CAM02UCS colour rendering index. Color Res Appl 37(3):160–167 Lin Y, Wei M, Smet KAG, Tsukitani A, Bodrogi P (2017) Colour preference varies with lighting application. Lighting Res Technol 49:316–328 Luo MR, Cui G, Li C (2006) Uniform colour spaces based on CIECAM02 colour appearance model. Color Res Appl 31(4):320–330 MacAdam DL (1942) Visual sensitivities to color differences in daylight. J Opt Soc A 32:247–274 Mahler H, Ezrati J-J, Vienot F (2009) Testing LED lighting for colour discrimination and colour rendering. Color Res Appl 34:8–17 Munsell AH (1929) Munsell book of color. Munsell Color Company, Inc., Baltimore, MD Ohno Y (2014) Practical use and calculation of CCT and Duv. LEUKOS 10(1):47–55 Ohno Y, Fein M (2014) Vision experiment on acceptable and preferred white light chromaticity for lighting. In: Proceedings of CIE 2014: Lighting quality and energy efficiency, Kuala Lumpur, Malaysia

References

55

Ohno Y, Oh S (2016) Vision experiment II on white light chromaticity for lighting. In: CIE Publication x042:2016: Lighting quality & energy efficiency, Melbourne, Australia, pp 175–184 Ouweltjes JL (1960) The specification of colour rendering properties of fluorescent lamps. Die Farbe 9:207–246 Perz M, Baselmans R, Sekulovski D (2016) Perception of illumination whiteness. In: CIE Publication x043:2016: Proceedings of the 4th CIE expert symposium on colour and visual appearance, Prague Rea MS, Freyssinier-Nova JP (2008) Color rendering: a tale of two metrics. Color Res Appl 33 (3):192–202 Royer MP (2018) Comparing measures of average color fidelity. LEUKOS 14(2):69–85 Royer MP, Houser KW, Wilkerson AM (2012) Color discrimination capability under highly structured spectra. Color Res Appl 37:441–449 Schanda J (ed) (2007) Colorimetry, understanding the CIE system. Wiley, Hoboken, NJ Smet K, Hanselaer P (2015) Memory and preferred colours and the colour rendition evaluation of white light sources. Lighting Res Technol 48(4):393–411 Smet K, Schanda J, Whitehead L, Luo R (2013) CRI2012: A proposal for updating the CIE colour rendering index. Lighting Res Technol 45(6):689–709 Smet K, Whitehead L (2011) Meta-standards for color rendering metrics and implications for sample spectral sets. In: Proceedings of the 19th color and imaging conference, San Jose, CA Smet K, Ryckaert WR, Pointer MR, Deconinck G, Hanselaar P (2011) Correlation between color quality metric predictions and visual appreciation of light sources. Opt Exp 19:8151–8166 Smet K, Whitehead L, Schanda J, Luo MR (2016) Toward a replacement of the CIE color rendering index for white light sources. LEUKOS 12:61–69 Stockman A, Sharpe LY (1999) Cone spectral sensitivities and color matching. In: Gegenfurtner K, Sharp LT (eds) Color vision: from genes to perception. Cambridge University Press, Cambridge, pp 53–87 Teunissen C, van der Heijden FHFW, Poort SHM, de Beer E (2017) Characterising user preference for white LED light sources with CIE colour rendering index combined with a relative gamut area index. Lighting Res Technol 49:461–480 Thornton WA (1972) Color-discrimination index. JOSA 62:191–194 Van der Burgt P (2016) Metamerism. In: Luo MR (ed) Encyclopedia of color science and technology, vol 2. Springer Science+Business Media, New York, p 931 Van Der Burgt P, Van Kemenade J (2010) About color rendition of light sources: the balance between simplicity and accuracy. Color Res Appl 35(2):85–93 Van Kemenade JTC, Van Der Burgt PJM (1995) Towards a user oriented description of colour rendition of light sources. In: CIE 23rd Session, New Delhi, vol 1, pp 43–46 Von Kries JA (1911) In: Nagel W (ed) Handbuch der Physiologisches Optik, vol 2. Leopold Voss, Hamburg, pp 366–369 Wang Y, Wei M (2018) Preference among light sources with different Duv but similar colour rendition: a pilot study. Lighting Res Technol 50:1013–1023 Wei M, Houser KW (2016) What is the cause of apparent preference for sources with chromaticity below the blackbody locus? LEUKOS 12:95–99 Wei M, Royer M, Huang H-P (2019) Perceived colour fidelity under LEDs with similar Rf but different Ra. Lighting Res Technol 1–12. Prepublished Dec 23, 2018. https://doi.org/10.1177/ 1477153519825997 Wright WD (1928–29) A re-determination of the mixture curves of the spectrum. Trans Opt Soc Lond 30:141–164

Chapter 3

Visual Performance

Abstract Visual performance for tasks of different size and contrast as a function of background luminance provides information about what lighting levels in interior working situations are required as a minimum. Visual performance has been studied as threshold and suprathreshold performance. For easy to moderately difficult tasks, as occurring in many offices, it is shown that visual performance is not a key issue for determining what lighting level is required. In situations with more difficult visual tasks, visual performance becomes an issue. With the described RVP model, calculation of relative performance for actual tasks gives insight into required lighting levels. Visual performance of older workers deteriorates considerably, and their performance should always become a consideration in setting lighting levels. Disability glare, the form of glare that is responsible for a negative influence on visual performance, has a neglectable effect on visual performance under most interior lighting conditions. The spectrum of light influences the threshold performance measure of visual acuity through its effect on the size of the pupil. A method that compares different spectra regarding the efficiency of providing visual acuity is described. Under many working conditions, however, this is of limited relevance since most of the visual tasks are far above the threshold of visibility.

Without light, our sensory system for vision does not work. How well we visually can perform with light is dependent on the quality of our vision system, the quality of the lighting and the difficulty of the visual task. The difficulty of the visual task relates, on the one hand, to its physical properties, the most important ones being size and contrast, and on the other hand to what type of performance is required: detection or recognition. There are different measures for visual performance such as speed and accuracy of performing the task. Studying the relationship between visual performance and lighting gives insight into what lighting parameters are relevant and how the values of these parameters influence visual performance. This knowledge is indispensable in specifying and designing lighting installations or drafting lighting standards for the many situations that can be encountered in practice. Fortunately, there are many research results to draw on. Most of these studies were done in the second half of the last century. Some of them have become © Springer Nature Switzerland AG 2019 W. van Bommel, Interior Lighting, https://doi.org/10.1007/978-3-030-17195-7_3

57

58

3 Visual Performance

classic examples of visual performance studies. Today their results are as valid as they were at the time the experiments were done. The relationship between our visual performance and lighting does not change over time. In most indoor situations, lighting requirements should not be based solely on visual performance. Visual performance requirements together with visual comfort requirements define the lighting quality needed. Chapter 4 deals with the relation between lighting and visual comfort. Chapters 5–7 discuss which additional lighting requirements are needed to ensure beneficial non-visual biological effects of lighting needed for health, well-being and alertness.

3.1

Visual Task

Different activities involve different visual tasks. Many visual tasks have a two-dimensional character such as reading and writing. Quite some industrial tasks also have a two-dimensional character. Of course, three-dimensional tasks do occur as well. Viewing the human face is one of the more important ones. The two-dimensional task is studied mainly as an achromatic task. The difference between the luminance of the target itself and the luminance of its background makes such a target visible. Luminance contrast C is defined by C¼

Lobj  Lback Lback

with Lobj ¼ luminance of the object Lback ¼ luminance of its background If the object is darker than its background, like black print on paper, the contrast is negative. If on the other hand it is brighter than its background, the contrast is positive. Usually, the task receives the same amount of light as its immediate background. For perfectly diffuse reflecting surfaces, the contrast is then solely dependent on the reflectances of object and background: C¼

E  ρobj  E  ρback ρobj  ρback ¼ E  ρback ρback

with E ¼ illuminance on task and background ρobj ¼ reflectance of object ρback ¼ reflectance of background

3.1 Visual Task

59

10 point

10 point

10 point

10 point

C ~ 0.30

C ~ 0.45

C ~ 0.65

C ~ 0.95

Fig. 3.1 Contrasts with different values (values are approximate because of the influence of the printing process of this book)

Fig. 3.2 Two different types of visual tasks. Left: related to detection of a target of uniform luminance seen against its background also of uniform luminance and right: detection of fine details with alternating luminances

Lobj

Lback

Lmax

Lmin

The contrast of a visual task is thus purely a property of the task itself. To give an impression of how different contrasts look like, Fig. 3.1 shows contrasts with four different values. Since the object is darker than the background, it concerns negative contrasts. Sometimes a task requires the detection of small details with alternating luminances. Figure 3.2, right, is an illustration of such a task where the difference between the light and dark bars has to be detected. In such a situation, the difference in alternating luminances determines the detection possibility. Therefore, a different definition of contrast is used: C¼

Lmax  Lmin Lmax þ Lmin

with Lmax ¼ maximum luminance of the detail Lmin ¼ minimum luminance of the detail The visual task is, apart from its contrast, also characterised by its size as seen by the observer. The angle subtended by the object at the eye of the observer defines it. As is evident from Fig. 3.3, both the physical size of the task and the observation distance play a role. The visual angle may be expressed as a flat angle in minutes of arc (α in Fig. 3.3) or by the three-dimensional solid angle in steradians (ω in Fig. 3.3). Children can see objects sharp from a much shorter distance, even less than 25 cm, than older people. Therefore, the same physical-sized task may result in a larger viewing angle for children. Table 3.1 gives the visual angle (expressed in minutes of arc) and solid angle (expressed in micro-steradians) for different-sized letters (expressed in points) for two different viewing distances. The main text of this

60

3 Visual Performance

α

ω

Fig. 3.3 Visual angle under which an object is seen, from two different observation distances. The visual angle can be expressed by the plane angle α in minutes or by the solid angle in steradians (ω) Table 3.1 Flat visual angle α (minutes) and solid angle ω (steradians) for different-sized letters when viewed from different distances based on lower case height of times character types Letter size Point 4 8.5 10

α at 25 cm Minute 8.3 17.5 20.6

ω at 25 cm Micro-steradian 4.6 20.0 28.2

α at 50 cm Minute 4.1 8.8 10.3

ω at 50 cm Micro-steradian 1.1 5.1 7.1

book is printed in a “Times” letter of 10 point, while the text of the captions is printed in a Times letter of 8.5 point. The small 4-point letter is representative of many difficult industrial tasks. There is some controversy, whether the solid angle or the flat visual angle is a more suitable criterion for object size (CIE 2002a). For letter type of objects, the solid angle takes into account the total inked area of the letter, while the flat angle takes into account the most critical size in only one direction of the letter. Table 3.1 has taken the lower case height as the basis for the flat visual angle. The conversion from the flat visual angle to solid angle is based on circularly shaped targets.

3.2

Threshold Visibility

Well-known measures for visibility of two-dimensional objects are the threshold of contrast and the threshold of visual acuity. Visual acuity relates to the capacity for distinguishing fine details. It is measured by bringing the size of the object to its threshold of vision. Fundamentally, visual acuity is also dependent on the contrast of the detail with its background. Many visual acuity tests are carried out with black objects on a white background. An example of such a test is the black letter test used by opticians when doing eye tests. The measure of threshold contrast relates to the least detectable contrast of an object. It is measured by bringing the contrast to its threshold. Obviously, this can be done for sets of different-sized objects and different background luminances. Of course, for most indoor lighting applications visibility just at the threshold of visibility is not sufficient. For safe, efficient and comfortable working conditions, visibility should be clearly above the threshold of visibility. Nevertheless, studies on

3.2 Threshold Visibility

61

the relationship between the values of different lighting parameters and the threshold of visibility are important. They provide a fundamental understanding of what lighting parameters determine the limitations of vision and what values of these parameters can safeguard vision to stay far enough above its limit.

3.2.1

Visual Acuity

Fundamental visual acuity research often uses gratings as test objects. Gratings are two-dimensional objects consisting of alternating dark and light bars as shown in Fig. 3.4. Van Nes and Bouman (1967) did extensive visual acuity measurements. They used gratings with a transition between the dark and light bars according to a sine-wave function (Fig. 3.4). In the example of Fig. 3.4, the upper set of gratings has a higher background luminance (Lb) than the two lower sets. The bottom set has a contrast value (C) that is only half of that of the top two sets. The grating with the largest number of bars that still can be distinguished as having separate bars defines visual acuity. The measure is the number of pairs of light and dark bars that can be distinguished per degree of visual angle (Fig. 3.5). Since each pair of light and dark bars is called a cycle, the measure is expressed as cycles per degree and is referred to as the spatial frequency. Note that larger spatial frequencies imply that smaller details can be detected. Figure 3.6 shows the results of Van Nes and Bouman regarding threshold contrast and threshold of detail size for four different background luminances. Mostly, the higher the background luminance, the easier it is to see details: same-sized details can be seen at a lower contrast or, at the same contrast, smaller details can be seen. The background luminance determines the state of adaptation of the observers. As Fig. 3.4 Sine-wave gratings with different contrast, C, and background luminance, Lb, values

Fig. 3.5 Spatial frequency expressed in cycles/degree; in these examples 3 and 6 cycles/degree

1 degree

1 cycle

62

3 Visual Performance

1

C th

0.1

visible

Lb (cd/m2) 0.3

0.01 3 30

300

invisible 0.001 100 50

20

10

5 2 cycles/degree

1

0.5

0.2

0.1

D th

Fig. 3.6 Relationship between the combination of contrast threshold, Cth, and detail threshold, Dth, of gratings for different background luminances, Lb. The detail size is quantified in terms of the number of alternating dark-light cycles per degree of visual angle. Visibility based on detail detection in more than 50% of observations. Figure adapted from Van Nes and Bouman (1967)

explained in Chap. 1, a higher adaptation luminance increases the sensitivity of the visual system through neural changes and increases the area where the image is sharp by decreasing the pupil size. The straight dashed line indicates the absolute limit for visual acuity: smaller details are not visible whatever the contrast or background luminance. The limit lies around 60 cycles/min, a value in line with other research results (Westheimer 1964; Hoekstra et al. 1974). Resolving power of the eye is limited by light scattering in the various eye layers and by the spatial distance between individual photocells of the retina. At first sight, it may be surprising that, with large detail sizes, corresponding to spatial frequencies smaller than some 5 cycles/degree, detail vision becomes more difficult with a further increase in detail size. It, however, correctly shows that gradual transitions spread over a relatively wide area are difficult to notice. Consider, for example, a sinus-wave alternating dark-light pattern on a wall seen from a distance of 3 m. The visual angle of 1 corresponds to an area on the wall of 50 cm. A frequency of 5 cycles/degree, in its turn, corresponds to only 1 darklight pattern stretched over this 50 cm area. A slightly higher number of narrower dark-light patterns of the same luminance difference, spread over the same 50 cm, are easier to distinguish. At a larger frequency of for example 10 cycles/degree, two

3.2 Threshold Visibility

63

Fig. 3.7 Grating, which demonstrates the course of the curves of Fig. 3.5. The size increases (logarithmically) from left to right and the contrast increases (again logarithmically) from bottom to top. The thin white line indicates the threshold of visibility of the gratings as viewed by the author of this book from a distance of 60 cm at a lighting level of 500 lux

dark-light, smaller, patterns cover the 50 cm of the wall. These indeed are easier to distinguish. The right side of the drawing of Fig. 3.7 makes this phenomenon visible. The drawing demonstrates the course of the curves of Fig. 3.6. In the drawing, the contrast increases (logarithmically) from bottom to top just like the vertical axis of Fig. 3.6. The detail size of the alternating light-dark patterns increases (logarithmically) from left to right, like the horizontal axis of Fig. 3.6. The thin white line indicates the threshold of visibility for the author of this book when viewing Fig. 3.7 from a distance of 60 cm. It follows a similar course as the curves of Fig. 3.5. The threshold for many readers is probably somewhat better than that of the author (“their” white curve may be at a lower position), because of the effect of age on vision (see also Chap. 8). When increasing the light on the picture, the threshold of visibility moves to lower contrasts: a demonstration of the positive influence of the background luminance.

3.2.2

Threshold Contrast

The most extensive contrast threshold measurements of targets of uniform luminance have been carried out from the mid-1940s to the early 1980s, by Blackwell in the United States. For many object sizes, contrasts and background luminances, observations were made by large groups of observers of all ages. Figure 3.8 shows part of the results of a total of 436,000 observations made by 35 observers in the age group of 20–30 years (Blackwell 1981). The targets were presented to the observers

64 Fig. 3.8 Contrast threshold, Cth, of circular targets, observed under different visual angles for 0.2 s, uniformly lit with light of correlated colour temperature of 2848 K in dependence of background luminance, Lb. Observers in the 20–30-year age group (Blackwell 1981). Best-fit curves are based on measurement points including those outside the range displayed here

3 Visual Performance

2

1 min

0.5

2 min

Cth 1

0.2

4 min

0.1 0.05

20 min 0.02

0.01 1

2

5

10

20

50

100 200

500

1000

Lb (cd/m2)

for 0.2 s. This value was selected to be typical for fixational pauses in carrying out visual tasks such as reading (Ginsburg 1951). As is evident from Fig. 3.8, the visual angle and thus the task detail have a significant influence on the required contrast threshold. A larger background luminance has the effect of lowering the threshold contrast. Within the background luminance range typical for interior lighting applications, from some 10–200 cd/ m2, this effect is relatively small. This is entirely different from road lighting applications where the luminances are much lower. Within the interior lighting range, a higher background luminance cannot make a two times smaller object detectable at its threshold as becomes evident by comparing the 4-min target with the 2-min target or the 2-min target with the 1-min target in Fig. 3.8. The results of threshold contrast observations are expressed often as the reciprocal of the threshold contrast. It is called the contrast sensitivity CS: CS ¼

1 j Cth j

with CS ¼ contrast sensitivity (for the actual task and background) Cth ¼ contrast threshold (for the actual task and background) To make these values relative, CIE (1981a, b) recommends setting the contrast sensitivity value CS at a value of 100 cd/m2 equal to unity. These relative contrast sensitivity values, RCS, can be seen as the potential of our vision system to make contrasts detectable. The higher the relative contrast sensitivity, the easier it is for the vision system to make contrasts detectable. Figure 3.9 gives the relative contrast sensitivity values, RCS, obtained from the data of Fig. 3.8, for three visual angles. As

3.3 Suprathreshold Visibility Fig. 3.9 Relative contrast sensitivity, RCS, in dependence of background luminance, Lb, for different target sizes for the age group of 20–30 years and the 4-min target also for the age group of 60–70 years. Curves based on the data of Fig. 3.8

65 0.2

RCS 1

age 20-30 20 min

0.5

4 min

0.2

4 min, age 60-70

1min

0.1 1

2

5

10

20

50

100 200

500

1000

Lb (cd/m2) the curves for smaller objects are much steeper than for larger objects, the positive effect of background luminance is considerably larger for the smaller objects. Blackwell did his experiments with observers of different age groups. As an illustration, the small stippled curve in Fig. 3.9 gives the results for the 60–70-year age group, for the 4-min disc (Blackwell 1981). The dramatic deteriorating effect of age on contrast vision is evident. Where the 20–30-year group can detect a 1-min disc, the 60–70-year group can only detect a disc about four times larger (4 min). Chapter 8 discusses more details of age effects. Blackwell developed, largely based on his contrast threshold measurements, a visibility model that relates task performance to lighting levels as a basis for interior lighting recommendations. A Technical Report of CIE (1981a, b) describes this model. However, visual tasks can have different visual components that can be so complex that their visibility prediction with simple, uniform-luminance disc type of targets only is impossible (Clear and Berman 1990; Bailey et al. 1993; Rea 2012; Boyce 2014). Moreover, also non-visual aspects such as ergonomic and environmental conditions, duration and strenuousness of work should play a role in setting lighting levels.

3.3

Suprathreshold Visibility

It has been explained in the previous section that investigations into threshold visibility can provide a basic understanding of what lighting parameters determine the limitations of our vision system and what values of these parameters determine whether vision is above its limit. In almost every human activity, visual performance should be well above its visibility threshold to perform efficiency. Efficiency depends on speed and accuracy with which visual tasks can be detected and

66

3 Visual Performance

Fig. 3.10 Principle of the stimulation-response relationship of our sensory systems

Response 1.0 knee

plateau

0.8 0.6 escarpment

0.4 0.2 bottom

0 0.01

0.1

1

10

100

Stimulation

identified. Many researchers have carried out suprathreshold investigations into the relationship between lighting level and speed and accuracy of work. Much of the experimental suprathreshold performance results can be explained by neurophysiological models of the way the photoreceptors and connected neurological network react to light stimuli (Naka and Rushton 1966; Lipetz 1971). These models predict the response (visual performance) as a reaction to light stimuli. They show that, fundamentally, the stimulation-response curve has an S-shaped form when the stimulation is displayed on a logarithmic scale and the response on a linear scale. This S-shaped curve is mathematically more accurately referred to as a hyperbolic tangent (tanh) curve and is shown in Fig. 3.10. This curve shape is typical of all our senses. At very low levels of stimulation, the bottom of the curve, no response is produced. With increasing stimulation, a small and gradually increasing response is obtained. In the middle range of stimulation, a modest increase invokes a large response: the escarpment of the curve. Then gradually the response decreases again: the knee of the curve. Finally, at the end of the curve, a further increase in stimulation hardly increases the response: the plateau of the curve. Where visual performance is important, the stimulation (lighting level) should be such that the response (visual performance) is at least in the knee section of the curve close to the plateau for tasks that are relevant to the actual application. Many experiments, therefore, are aimed at finding the detailed relationship between lighting level and performance for that region. Consequently, results are often only shown for the knee part of the S-shaped curve.

3.3.1

Landolt-Ring Task

Weston in the United Kingdom has set the example of suprathreshold research with his exemplary experiments of speed and accuracy measurements of office and industrial visual tasks with different sizes and contrasts (Weston 1953, 1961). He used Landolt rings as the task. A Landolt ring is a standardised ring with a thickness

3.3 Suprathreshold Visibility

67

D 1/5d

d

1/5d

Fig. 3.11 Standard dimensions of Landolt rings (left) and (right) an example of a test sheet as used by Weston in his suprathreshold visibility tests

of one-fifth of the ring’s outer diameter and with a gap with a width of also one-fifth of the diameter (left of Fig. 3.11). The right side of Fig. 3.11 shows an example of Weston’s test sheets on which Landolt rings are printed with random orientations of the gap in the Landolt ring. The test persons had to identify, starting at the top line and working from left to right, as many rings as possible having a specified orientation of the gap. In many test rounds, the effects of different illuminances, ring sizes and contrasts were studied. Speed was corrected for the time needed for the act of marking so that a measure of visual performance only is obtained. The observers made for this purpose an extra test in which the correct orientation of the gaps was already clearly marked in red ink. Speed, S, is expressed as the reciprocal value of the average time (in minutes) to detect one ring correctly (excluding the time for the act of marking):
S ¼ 1=

 total test time  time for marking number of rings correctly identified

The accuracy, A, is expressed as the fraction of correctly marked rings out of the total number of rings with the specified orientation: A¼

correctly identified rings total number of rings with specified orientation

68

3 Visual Performance

Performance (1/sec) 0.6

Performance (1/sec) 0.6

D 4.5 min

C 0.97

3 min 0.4

0.4

0.3

0.3

0.56 0.39 0.28

0.2

0.2

1.5 min C=0.56

0.1

D=3 min

0.1 0

0 10

100

1000

5000

E (lux)

10

100

1000

5000

E (lux)

Fig. 3.12 Performance as a function of illuminance on the task (E). Left: for different detail sizes based on the size of the gap (D) and contrast value (C) of 0.56. Right: for different contrast values (C) and detail size of 3 min. The average age of observers 37 years (Weston 1953, 1961). Best-fit curves are based on measurement points including those outside the range displayed here

The measure used by Weston for the overall performance, P, is the product of speed and accuracy: P¼SA Figure 3.12 gives part of Weston’s results: on the left for different detail sizes (D) all for a contrast value (C) of 0.56 and on the right for different contrast values, all for a detail size of 3 min. The contrast value of 0.97 corresponds to black rings on white paper, whereas the lower contrast values correspond to black rings on different shades of grey paper. Up to some 500 lux, performance increases for all detail sizes and contrasts. The increase in performance with illuminance level is larger for more demanding tasks, i.e. for smaller sizes and smaller contrasts. This same effect can be experienced when reading the text of Fig. 3.13. While increasing the illuminance: the readability of the low contrast and smaller letters improves more than the others. This only works if started at a relatively low illuminance level. From 200 to 500 lux onwards the steepness of the curves of Fig. 3.12 levels off. The curves are on or near to their highest point, the plateau, meaning that performance approaches saturation. CIE (2002a) compared Weston’s results with those of later studies done by different researchers, all with different types of visual tasks (number-search task: Muck and Bodmann 1961, two-letter task: McNelis 1973, complex Landolt-ring task: Loe and Waters 1973, reading task: Smith and Rea 1978). A next section describes Muck and Bodmann’s search-task experiments. CIE concludes that the results of the later studies fit reasonably well with Weston’s data, especially if extremes in contrast values and background luminances are disregarded. Weston’s

3.3 Suprathreshold Visibility Fig. 3.13 Starting with a relatively low illuminance level, increasing the illuminance level on the page of this book improves readability of visually difficult tasks more than easy tasks

69

At relatively low illuminance levels, readability of supra-threshold, small-contrast letters improves more (faster and less errors) with a same increase in illuminance than large-contrast letters do. At relatively low illuminance levels, readability of supra-threshold, large-contrast letters improves less with a same increase in illuminance than small-contrast letters do. At relatively low illuminance levels, readability of supra-threshold, small-sized letters improves more (faster and less errors) with a same increase in illuminance than large-sized letters with a same contrast do

data are therefore suitable for getting a good understanding of how different lighting conditions affect visual performance for various tasks in suprathreshold conditions. CIE developed a mathematical model with which the results of Weston’s visual performance values can be calculated as a function of background luminance, contrast, visual angle and age of the observer. The performance values are made relative by relating them to nearly the highest performance value achievable, obtained for a 4.5-min visual angle, a contrast of 0.9 and a background luminance value of 1000 cd/m2. Appendix C gives the set of formulas of the calculation model. We used the model to calculate so-called RVP bodies to demonstrate the relationship between relative visual performance (RVP), background luminance (Lb) and contrast (C). Figure 3.14 shows, for a visual angle of 8.8 min (same size as an 8.5-point letter seen from 50 cm) and 4.1 min (same size as a 4-point letter size from 50 cm), the RVP bodies as calculated for observers of 25 and 60 years. The extremes of contrast (C < 0.30) and of background luminance (Lb < 3 cd/m2) are not taken into account. In most indoor lighting applications, these values do not occur. The area of the RVP body where relative visual performance is close to saturation (RVP > 0.9) is coloured green: the plateau. Here, lighting level and contrast improvements hardly result in further improvements in the already high performance. Where visual performance is on the steep downward part of the RVP body (RVP < 0.7), the area is coloured red (the escarpment). The increase of lighting level and contrast has here an important role in improving visual performance. The in-between area is coloured orange. For the larger task (8.8 min) visual performance for young observers is at a high level for the whole range of luminance and contrast values. It means that for this condition lighting level has not a significant influence. This is dramatically different for the older observers: visual performance decreases considerably. For contrast values, lower than some 0.65, visual performance is on the escarpment part of the curve. Increasing lighting levels improves the situation, although visual performance cannot reach the same level as that of young observers. For the task with small detail size, visual performance decreases strongly: only for very large contrast values performance is still on the plateau. For small-sized visual

70

3 Visual Performance

8.8 min, 25 yrs

8.8 min, 60 yrs

0.6

0.5

0.5

0.4

0.4

0.3

0.3

4.1 min, 60 yrs

3

10

30

0.6

0.6

0.5

0.5

0.4

0.4

0.3

0.3

3

10

30

100

0.3

Lb ( cd/m 2

0.45

0.65

0.30

0.45

C

3

0.7

30

0.7

100

0.8

300

0.9

0.8

10

100

0.9

300

1

1

RVP

1

1

RVP

)

C

)

4.1 min, 25 yrs

Lb ( cd/m 2

C

0.65

)

0.3

Lb ( cd/m 2

0.65

C

0.45

0.65

0.45

0.30

Lb ( cd/m 2

3

0.7

0.6

10

0.7

30

0.8

100

0.9

0.8

300

0.9

300

1

1

RVP

1

1

RVP

)

Fig. 3.14 RVP bodies. Relative visual performance (RVP) as a function of object contrast (C) and background luminance (Lb). Top: visual angle 8.8 min (same size as an 8.5-point Times letter viewed from 50 cm); bottom: 4.1 min (same size as a 4-point Times letter viewed from 50 cm). Left 25 years of age and right 60 years of age. Calculated from CIE’s calculation model (2000a) based on Weston’s data (formulas given in Appendix C)

tasks, lighting level does play an important role in bringing visual performance to acceptable levels for young observers. Of course, for such small task, lighting level is also significant for improving performance for the older age group. However, even then, only contrasts with extremely high values can be performed moderately acceptable.

3.3 Suprathreshold Visibility

71

Weston wanted to use his performance curves also as a direct tool to produce lighting-level recommendations for interior lighting. The same arguments, mentioned in the previous section, when explaining that Blackwell’s threshold visibility curves are unsuitable for this purpose, also hold for Weston’s curves.

3.3.2

Search Task

Weston’s task is especially representative for reading tasks. Muck and Bodmann (1961) used number-searching tasks, which are more typical for visual inspection work. For this purpose, they used test sheets with a random distribution of all numbers from 1 to 100, each placed in a small circle. Figure 3.15 shows an example of such a sheet. Different sheets were used with different-sized numbers and contrast values. The subject holds a metal pointer which, when placed on the correct number, releases an electrical contact and so stops a timing device. In this way, the detection time for a specified two-digit number was obtained. As the time needed for the visual search is long compared to that required to act with the pointer, the task is almost exclusively a visual one. Figure 3.16 gives the results in terms of the average detection speed (S)

82

47

1

31

93

78

43

35

87 2

28

62

83

38

25

22

98

17 57

37

60

73

4 13

20

75

32 51

11 66

80

36

88

77

23 68

91

55

58

49 53

44 8

26

42

29

16

85

90

95

24

46 67

81

69

30

9 64

54

48

52

56 30

14

94

12

27 96

76

92

59

74

72

40

19

6

41

45

15

3

65

79

89

84

21

33

7

70

10

34

99

63

97

50

5

71 86

18

100

39 61

Fig. 3.15 An example of a test sheet as used by Muck and Bodmann in their tests with numbersearching tasks

72

3 Visual Performance

Fig. 3.16 Visual performance of a search task expressed in speed (S) as a function of illuminance (E) on the task for different contrast values (C) and a detail size (D) of the task of 4 min (Bodmann 1967). Best-fit curves are based on measurement points including those outside the range displayed here

S (1/sec) 0.05 C=0.93 20-30 yrs

0.04

C=0.63 20-30 yrs

0.03 0.02

C=0.63 50-65 yrs

0.01

D= 4 min

0 10

30

100

300

1000

3000

E (lux)

5000

for a 4-min object size for a contrast value of 0.93 (black numbers on a white background) and 0.63 (black numbers on light grey background). The curves have a similar course as the Weston curves of Fig. 3.12. The absolute values are about a factor of 10 lower: a searching task is more difficult than a reading task. A group of older test persons (50–65 years) scored considerably worse than a younger age group (20–30 years). For the younger group, an increase of illuminance has only a minor positive effect on performance and saturation starts at lighting levels of around 500 lux. The older group, however, benefits substantially from such an increase of illuminance. Here saturation only occurs at levels higher than 3 to 5000 lux. Even then the older test persons cannot perform as well as younger ones at illuminance levels as low as 50 lux. Bodmann’s search test is also suitable as a learning tool for classroom students. It can demonstrate how lighting level influences their performance. Of course, the task can be made more strenuous by printing the sheet of Fig. 3.15 on grey instead of on white paper. By having them do the test at two widely different classroom lighting levels, it also shows how frustrating it can be to have to execute a demanding visual task at a too low lighting level. In this way, they get some insight into the experience many elderly have in many daily situations.

3.3.3

Verification Task

Rea (1981, 1986) designed a number-verification task. The test persons were asked to compare two printed number lists, a reference list on the left and a response list on the right (Fig. 3.17).

3.3 Suprathreshold Visibility Fig. 3.17 Example of Rea’s verification test sheet. Middle digits were observed from 50 cm resulting in a visual angle of 19 min for the height of the digits

73

68739

68739

86395

86395

71194

71124

28740

28740

23718

23718

96271

96271

58709

58709

93628

93628

67455

67455

36544

36544

19213

19213

47564

47562

85223

85223

53428

53428

73886

73886

93878

93878

34762

34762

42585

42585

63987

63987

56683

56683

Reference list

Response list

Different reference lists, printed in different grey tints on white paper, were used to get different contrasts between the numbers and the background. The response sheets, always printed in high contrast, contained one to three discrepancies at random locations. The test persons, between 19 and 24 years old, were asked to compare the reference and response lists, from top to bottom, as quickly as possible and to mark discrepancies with a pen on the response list. Total time required to finish a sheet was measured. False and missed responses were measured as well, but were found to play only a minor role. The total time recorded was corrected for the time needed to carry out the act of marking, and for the time required to read the response list. In this way and also by the experimental design of the experiment, the corrected time was largely a result of visual factors only. Figure 3.18 shows, for some typical contrast values, the performance results regarding speed based on the corrected time. The course of the curves shows the same trend as those of the studies of Weston and Bodmann. With increasing background luminance (or illuminance providing that luminance), performance increases. At higher luminances, the increase becomes gradually less, especially at high contrasts for which saturation starts from slightly more than 100 cd/m2 onwards (some 500 lux). Starting with low

74

3 Visual Performance

Fig. 3.18 Visual performance of the numerical verification task of Fig. 3.17, expressed in speed S (corrected to be representative for the visual aspect only) as a function of background luminance (L ) and approximate illuminance (E) that causes this luminance, for different contrast values (C). The average age of observers is 22 years. The best-fit curves are based on Rea’s visual performance model (Rea 1986)

Speed (1/sec) 0.08 C=0.90

0.07 C=0.16

0.06 C=0.11

0.05

C=0.10

0.04

0.03 30

10 50

100

100 200

L (cd/m2)

500

300

1000

E (lux)

contrast values, increasing those values just a little improves performance considerably. However, from medium contrast values onwards, say from some 0.20, the improvement of performance with a further increase of contrast is only small.

3.3.4

Rea’s Visual Performance Model

The purpose of Rea’s verification experiments was to lay the foundation for a model that enables the calculation of suprathreshold visual performance for a broad range of task sizes, contrasts and adaptation luminances. A requirement for such a model is that it relates exclusively to visual performance and not to non-visual task performance aspects such as ergonomical, motivational and emotional ones and that it does not include processing time. The verification experiment fulfils that requirement quite well. Subsequent studies by Rea and Ouellette (1988), using reaction time to the sudden presentation, random in time, of a square target, fulfils that requirement even better. The reaction time was recorded simply by releasing a button, making the time of processing so short that no correction was needed. Non-visual factors hardly can influence the result. To quote Boyce (2014): “detecting the presence of something as against nothing is about the simplest visual task possible”. With the results of these new experiments, arithmetic expressions were developed that enable the calculation of relative visual performance (RVP) based on reaction time (Rea and Ouellette 1991). Visual performance values are made relative by taking the

3.3 Suprathreshold Visibility

75

performance for a target of high contrast and large size as a reference; that is for such a target, the relative visual performance RVP is set to 1.0. The average age of the test persons participating in the reaction time tests was 21 years. The model was completed by making it age dependent. It is based on estimates of age-dependent reduction in transmission of the eye lens and decrease of pupil size (Weale 1961, 1963). Also, the reduction of apparent contrasts because of age-dependent scattering in the eye lens is taken into account. The age effect is even worse because of adverse effects in the neural network and permanent damage to part of the photoreceptors with an increase of age. The CIE calculation model based on Weston’s research, already discussed in a previous section, does not take into account these latter effects. Rea and Ouellette (1991) claim good similarity between their RVP model and their earlier verification task experiments. As far as the influence of luminance level is concerned, Rea’s RVP model indeed shows similar dependency as in this study and also as in Weston’s and the other earlier studies on suprathreshold performance. However, the influence of contrast of Rea’s RVP model is weaker than in all the other studies. Figure 3.19, where for a visual angle of 8.8- and 4.1-min Rea’s model is compared with the earlier discussed CIE model based on Weston’s experiments, illustrates this. Rea’s model uses a solid angle instead of a flat visual angle as a criterion for the task size. Although this may give a small error in the comparison, it has hardly any influence on the conclusion on the luminance and contrast sensitivities of both systems as expressed above. The purpose of the earlier referred CIE’s 2002 publication was “to provide an answer to the question which model is suitable to predict visual performance for tasks required in office work and most other indoor activities” (CIE 2002a). The publication concludes that the model based on Weston’s data provides such a good prediction. It also states that Rea’s reaction-time model is not suitable for such prediction as far as the contrast dependence is concerned. A possible reason could be that the reaction-time task is fundamentally differently processed than the other visual tasks dealt with here. In later experiments by Eklund et al. (2001) the amount of work done on a data entry task over an 8-h workday was used as a measure of visual performance. The test persons had to enter into a computer alphanumerical strings of ten characters with different sizes and contrasts, printed on paper, a task regularly carried out in offices. From the amount of work done in 8 h, work speed was calculated. The relative work speed dependency from background luminance, contrast and size showed good agreement with RVP values calculated from Rea’s RVP model (Eklund et al. 2001). There remains some uncertainty, whether Rea’s RVP model takes the contrast dependence strongly enough into account for the non-reaction type of visual tasks. When evaluating different lighting conditions based on the RVP model, this should be kept in mind. Appendix D gives the set of formulas for the calculation of RVP from Rea’s RVP model.

76

3 Visual Performance

Rea ± 8.8 min (5.1 μstr ), 25 yrs

0.7

0.6

0.6

0.5

0.5

0.4

0.4

0.3

0.3

3

10

30

Rea ± 4.1 min (1.1 μstr) , 25 yrs

0.8

0.7

0.7

0.6

0.6

0.5

0.5

0.4

0.4

0.3

0.3

3

10

30

100

0.65

0.3

Lb ( cd/m 2

0.45

0.65

0.45

0.3

)

C

3

0.8

10

0.9

30

0.9

300

1

1

RVP

1

1

RVP

100

100

Weston 4.1 min, 25 yrs

300

C

)

)

Lb ( cd/m 2

0.3

Lb ( cd/m 2

0.65

C

0.45

0.65

0.45

0.30

Lb ( cd/m 2

3

0.7

10

0.8

30

0.9

0.8

300

0.9

300

1

1

RVP

1

1

RVP

100

Weston 8.8 min, 25 yrs

C

)

Fig. 3.19 Relative visual performance (RVP) based on Weston’s and Rea’s data, respectively. Calculated from CIE (2002a) and Rea and Ouellette (1991). Formulas are given in Appendices C and D

3.4

Disability Glare

Glare can take either of the two forms: disability glare and discomfort glare. Disability glare is the form that is responsible for a negative influence of glare on visual performance. Glare that causes a feeling of discomfort is referred to as discomfort glare. It will be shown here that disability glare has a neglectable effect on visual performance under most interior lighting conditions. On the other hand, discomfort glare can cause problems in interior lighting situations. Chapter 4 will discuss discomfort glare.

3.4 Disability Glare

77

Fig. 3.20 Light scatter in the eye due to glare

θ

ne scene scene scene scene e c s ne sce

scene

fovea

Eeye

The mechanism by which the loss of visual performance due to the presence of glare takes place can be understood by considering the light scattering occurring in the eye of the observer as illustrated in Fig. 3.20. An image of the lit scene in front of the eye is focused on the retina of the eye. At the same time, however, light coming from a luminaire that is too close to the direct line of sight is partly scattered in the cornea, eye lens and eyeball. Part of that scattered light is redirected towards the fovea, where it acts as a bright veil drawn across the field of vision. This veil can be considered as having a luminance—the equivalent veiling luminance (Lveil)—proportional to the quantity of light scattered into the direction of the retina. The result of this veil is twofold. Firstly, it increases the adaptation luminance of the eye which, in turn, improves the contrast sensitivity of the eye: a positive effect. Secondly, however, it decreases the effective contrast of the targets because of the “masking” effect of the veil. The negative effect of a reduction of effective contrast is larger than the positive effect of a better contrast sensitivity due to the larger adaptation luminance. The overall negative effect of disability glare on visual performance can be calculated since Holladay (1927) developed, based on visibility tests, a formula that predicts the equivalent veiling luminance. Its value is dependent upon the illuminance on the eye (Eeye) and the angle (θ) between the viewing direction and the direction of light incidence from the glare source (see Fig. 3.20). The calculation formula has been refined by CIE (2002b) as Lveil ¼ Eeye 

 
 a  4  10  1 þ 70 θ2

with a ¼ age of the observer, for 0.1 < θ < 30 . Crawford (1936) showed that for multiple glare sources the total equivalent veiling luminance of a lighting installation can be described by adding the equivalent veiling luminances of all the individual sources; thus Lveil ¼

n X 1

Lveil, i

78

3 Visual Performance

Now the effective adaptation luminance and the effective contrast can be calculated while taking into account the effect of the actual veiling luminance of a lighting installation. The effective adaptation luminance becomes Ladapt ¼ Lb þ Lveil The effective contrast of the object decreases from C to Ceff: C¼

L o  Lb Lo

and

C eff ¼

ðLo þ Lveil Þ  ðLb þ Lveil Þ Lb ¼C ðLb þ Lveil Þ Lb þ Lveil

By using these effective contrast and adaptation luminances as input for the RVP model, the impact of disability glare on suprathreshold performance can be calculated. To demonstrate the effect of disability glare, Fig. 3.21, right, gives the results of RVP calculations (based on Weston’s data) for the 4.1-min target (4-point letter size) and observer age of 25 years for a glare situation characterised by a veiling luminance value that is 15% of the actual background luminance value. This roughly corresponds to a glare situation that is common under road lighting conditions that fulfil international standards (it corresponds to a TI value—the glare measure for road lighting—of 10% at a road lighting level of 1 cd/m2). In interior lighting situations, veiling luminance values are usually much lower. Figure 3.21, left, gives RVP values for the same conditions without glare.

4.1 min, 25 yrs

4.1 min, 25 yrs, Lv/Lb = 0.15

0.6

0.6

0.5

0.5

0.4

0.4

0.3

0.3

3

10

30

100

0.3

Lb ( cd/m 2

0.65

C

0.45

0.65

0.45

)

0.3

Lb ( cd/m 2

3

0.7

10

0.8

0.7

30

0.8

100

0.9

300

0.9

300

1

1

RVP

1

1

RVP

C

)

Fig. 3.21 Relative visual performance (RVP) as a function of object contrast (C) and background luminance (Lb) for a visual angle of 4.1 min and observer age of 25 years. Left: without glare; right: with glare. Calculated with CIE calculation model based on Weston’s data

3.5 Indirect Glare

79

From the graphs, it is clear that glare has only a small effect on visual performance for medium to high background luminance and contrast values (the orange and green area). For low to very low background luminance and contrast values, the glare effect becomes gradually more noticeable (the lower end of the escarpment of the body). As has been stated at the beginning of this section: disability glare indeed has a neglectable effect on visual performance under most interior lighting conditions.

3.5 3.5.1

Indirect Glare Non-self-luminous Tasks

Light from bright light sources reflected in glossy and semi-matt materials may disturb visual performance. Reflected light by specular areas not at the task itself but near to it may distract the eyes and draw them away from the task towards the brighter area. It is called the phototropism mechanism (Enoch and Birch 1981). When looking with both eyes towards a particular spot on glossy material, the bright spot of a light source is seen at slightly different places with the left and right eye. This causes what is called the binocular rivalry problem: an involuntary tendency to try to fuse the images (De Boer 1977). Finally, light reflected on the task itself scatters as shown in Fig. 3.22 with the example of a pencil task. Just as the scattering of light within the eye in the case of disability glare (see the previous section) is characterised by a veiling luminance, the light scattering into the direction of the observer, originating in the visual task itself, is also characterised by a veiling luminance. The effect of the veiling reflection is, again just as with disability glare, a reduction of task contrast. Added to this adverse effect for visual performance is a decrease in sharpness of the outer contours of the task. Both effects are visible in Fig. 3.22 (bottom). The effect of contrast reduction and the consequent loss of visual performance are dependent upon: Fig. 3.22 Pencil text on mat paper showing contrast reduction and loss of sharpness of contours (De Boer and Fischer 1981). Top: completely diffuse sphere illumination, bottom: strongly directional lighting

80

3 Visual Performance

Fig. 3.23 Offending zone in which luminaires cause the largest indirect glare problems in glossy tasks

Offending zone

• The reflective properties of the task and its background • The position of the light sources relative to the position of the task and the position and viewing direction of the observer • The size and luminance of the light sources Of course, the more glossy the task, the larger the adverse effects on visual performance. The most offending luminaire positions can readily be determined by drawing vision lines from the observer’s eye towards the edges of the visual task and onwards to the ceiling (Fig. 3.23). Larger luminaires occupy more easily, in whole or in part, the offending zones. High-luminance luminaires produce higher bright spots and larger veiling luminances. The actual contrast reduction for a specific task can be quantified by the contrastrendering factor (CRF). It is defined as the ratio of the contrast under the actual lighting condition relative to the contrast of the same task obtained in a reference condition of completely diffuse lighting (CIE 1981a, b): CRF ¼

C actual task ðactual lightingÞ C actual task ðdiffuse lightingÞ

The reference lighting is also referred to as sphere illuminance: luminous intensities equal from all directions around a sphere (see the sketch, top right, Fig. 3.22). Completely diffuse lighting results in only small veiling reflections so that the best CRF value is close to 1. In the case of light directed towards the task only from behind the observer, CRF may be slightly higher. If the precise reflection properties of a task and its background are known, the contrast values of the actual and the reference lighting situation can be calculated by adding the veiling luminances for the observer position and viewing direction to the Ltask and Lbackground components of the contrast formula. Very few complete reflection data sets for this purpose are available (Boyce and Slater 1981). The most completely documented task, for many observation and light incident angles, is that measured and published by Blackwell and DiLaura (1973). It is a relatively glossy “classroom task”, a pencil task on matt background. Although this task may not be the actual task often occurring in today’s offices, it can be useful to compare different lighting installations by calculated CRF values in those situations where glossy tasks often occur. Lynes (1982) published a data set for a pencil task for 20 viewing. For CRF calculations, the luminaires

3.5 Indirect Glare

81

should be split into smaller elements (all with the same light distribution) with dimensions small compared with the distance from the task. It should be realised that a CRF comparison has to be made for each working position. Striving for a situation in which all positions in a space have a large CRF value may result in a diffuse lighting situation where hardly any shadows occur. This, in turn, gives a dull and monotonous situation in which three-dimensional objects, including the human face, are not so easy to recognise and may appear unpleasant. Chapter 4 will explain that a certain directionality of lighting is necessary. Unfortunately, this does not always go together with maximum CRF values. In this context, it is good to note that everywhere, where physically possible, most people will automatically and unconsciously adapt their position and the position of the task to avoid disturbing reflections from the light sources. Just a slight turning or tilting of the paper being viewed at may avoid problems. Printing glossy material is often done more professionally than in the past by selecting the paper quality carefully and by having not the text but only the photographs printed in high gloss.

3.5.2

Self-Luminous Devices

Self-luminous devices vary from screens of desktop and laptop computers to tablets and smartphones. Direct glare from light reflected in screens may give the same problems as described in the previous chapter about non-self-luminous devices. Phototropism, binocular rivalry, contrast reduction because of veiling reflections together with a less sharp letter and symbol contours, all may occur. In the case of screens set to display negative contrasts (dark background and bright symbols) the first two problems are much larger than for screens set in positive contrast (bright background and dark symbols). This is because on a screen in negative contrast the conspicuity of the bright reflections in the dark screen is much higher than on a bright display. On a dark display, the luminance of the bright reflection spots may even be so high that they give direct rise to disability glare (they behave as secondary light sources). Fortunately, today most of the screens are used in positive contrast, probably with the sole exception of CAD workstations where a negative contrast setting facilitates the discrimination of different-coloured thin lines. The problem of contrast reduction caused by veiling reflections and the accompanying blur of contours cannot be changed by the contrast setting. Contrast reduction, together with the other problems mentioned, is very much dependent on the quality of the screen. There is a wealth of literature dating from the late 70s and the 80s of the last century about the problem of lighting and self-luminous display units (CIBS 1981; CIE 1984; IES 1989). With the negative contrast and dark-green display screens of those days that literature was important and relevant. Fortunately, many of the problems have been solved today at the source of the problem: the screen. High-quality display screens are available in which the screen is treated to reduce reflections (for example, coating, professional roughening, quarter-wave and mesh layers).

82

3 Visual Performance

Fig. 3.24 Offending zone in which luminaires may cause reflected glare to the user of a desktop computer. A wall behind the user solves reflection problems from the display

Offending zone

For desktop computers with a fixed tilted position of the screen, the offended zone can be determined where luminaires should not be placed in order not to radiate light in directions that can be reflected into the eyes of the user (Fig. 3.24). From Fig. 3.24 it is directly apparent that the luminaires just behind the user are not a problem. This means that having the user working in a relatively small office, or with his back at a short distance from a not too bright wall, solves many if not all reflection problems. If there are luminaires at a larger distance behind the user, these luminaires should have low luminances at higher elevation angles if the users work with low-quality display screens. Smartphones and tablets usually do present a smaller problem because, as with glossy paper material, it is easy to tilt and rotate the screen so that reflection problems are minimised. Many tablets have flexible holders which facilitate fixing the tablet in a suitable, reflection-free position.

3.6 3.6.1

Influence of Spectrum Pupil Size

The light source spectrum has, under certain conditions, an impact on visual acuity. This is because of the effect the spectrum has on the size of the pupil. Where initially the effect of the spectrum on pupil size was solely explained by the influence of rods, it is now clear that pRGC cells in the retina discovered in 2002 also play a role (see Sect. 1.5 of Chap. 1). These photosensitive ganglion cells even seem to have a dominant influence (Markwell et al. 2010; McDougal and Gamlin 2010; CIE 2015). The relative contributions of pRGCs, rods and to a lesser extent cones are not constant but change with light level and light duration. As the existence of pRGCs was still unknown when most of the research work on pupillary reflex was done (Alpern and Campbell 1962; Bouma 1962; Berman 1992), new research is needed to understand the mechanism of it fully. Chapter 5 will show that the maximum spectral sensitivity of pRGCs is obtained at a short wavelength in the blue range of the spectrum. The maximum sensitivity of rods also lies in the blue range (Fig. 1.5) at an only slightly longer wavelength. This is the reason that short-wavelength light (light with a relative large blue amount and

3.6 Influence of Spectrum Fig. 3.25 Average pupil diameter when viewing a wall (2  2 m from 1 m) lit with 2000 K high-pressure sodium (HPS) and 2800 K incandescent (Incand) lamps as a function of wall luminance. Observers between 17 and 20 years (Berman et al. 1987)

83

Pupil (mm2) 40 Incand

30

HPS

20

10 0

25

50

75

100 125 L (cd/m2)

thus high correlated colour temperature CCT) constricts the pupil more than longer wavelength light does (light with lower CCT). As an illustration of this phenomenon, Fig. 3.25 shows a comparison of the pupil size obtained under yellowish highpressure sodium light (CCT around 2000 K) and incandescent light (CCT around 2800 K). The pupil area with incandescent lamp light is between 5 and 8.5 mm2 smaller than with light from high-pressure sodium lamps of the same lighting level. A smaller pupil influences the quality of the retinal image. As rods and pRGCs are located only in the peripheral area of the retina, outside the fovea, the effect of the light spectrum on pupil size exists only with a full visual field. Research in which the visual field is restricted to the fovea only does, therefore, not show the effects described here.

3.6.2

Visual Acuity

A smaller pupil size has three effects on vision: • The negative effect of less light penetrating the eye, resulting in less light on the retina (lower retinal illuminance) • The positive effect of a better depth of field • The positive effect of less blurring and distortion of the retinal image by spherical aberrations in the eye lens and by imperfections of the eye material In 1987, Berman started a series of investigations in which he proved that for adaptation luminance levels larger than some 50 cd/m2 the overall effect of a smaller pupil size is positive as far as visual acuity is concerned (Berman et al. 1987, 1993, 1996; Berman 1992). Others have confirmed these effects (Navvab 2001, 2002). Figure 3.26 shows an example of results obtained by Berman et al. The trends shown in this figure are clear: for demanding tasks (small size, small contrasts at low background luminances), visual acuity improves with shorter

84

3 Visual Performance

Fig. 3.26 Visual acuity performance expressed as the proportion of correctly identified orientations of a Landolt ring (2-min gap), presented 200 ms, as a function of background luminance (Lb) for two contrast values (C). Large surround field lit by polychromatic bluegreenish (blue curves) and red-pinkish (pink dashed curves) light. Adaptation luminance 53 cd/m2. Age range between 18 and 45 years (trend curves calculated from Berman et al. 1993)

Visual acuity performance 1 C=0.40 0.9 C=0.40

0.8

C=0.24

0.7 blue-greenish light red-pinkish light

C=0.24 0.6 10

20

50

80

Lb (cd/m2)

wavelength light (higher correlated colour temperature). The other investigators mentioned above confirm these trends.

3.6.3

Equivalent Visual Efficiency Method

Berman (1992) argued that the pupil reflex was driven solely by rods. He showed that his experimental results of pupil size for a certain spectrum could be predicted from the proportions of radiation of that spectrum in the V0 (λ) and V(λ) areas, respectively. The larger the relative proportion in the scotopic V0 (λ) area (rod vision), the smaller the pupil size. The ratio of these proportions is called the scotopicphotopic ratio S/P: P E ðλÞ  V 0 ðλÞ S=P ¼ P E ðλÞ  V ðλÞ with the scotopic V0 (λ) and photopic V(λ) values expressed in absolute values and E(λ) the spectrum of the light source. Table 3.2 gives typical S/P values for various light sources. Note that different spectra may have the same correlated colour temperature but a different S/P ratio. So the colour temperature is just a rough indication of the S/P ratio value. IES of North America (2013) used Berman’s prediction of the pupil size to develop a method that permits the comparison of two light source spectra with regard to the efficiency with which they provide visual acuity. This so-called equivalent visual efficiency (EVE) method calculates the (photopic) lighting levels

3.6 Influence of Spectrum Table 3.2 Approximate S/P ratios and corresponding EVE factors for various light sources

85 Light source High-pressure sodium 2000 K Incandescent 2800 K Fluorescent/LED 3500 K Fluorescent/LED 4000 K Fluorescent/LED 6500 K Daylight 6500 K

S/P ratio 0.65 1.4 1.4 1.6 2.1 2.5

EVE factor 1.85 1 1 0,90 0.72 0.63

for two different light sources required to provide equal visual acuity (for demanding tasks seen under full-field vision conditions):
L2 ¼ L1

ðS=PÞ1 ðS=PÞ2

0:8

with L1 and L2 ¼ light level required with light sources 1 and 2, respectively (S/P)1 and (S/P)2 ¼ S/P ratio of light sources 1 and 2, respectively In Table 3.2 EVE factor values for equal visual efficiency are compared with a 3500 K fluorescent lamp or LED light source (with S/P ratio of 1.4) as the reference. It shows that, for visual acuity, it can be advantageous to use light sources with a relatively high S/P ratio. This can be particularly interesting for situations where demanding tasks, near to the threshold of visibility, have to be carried out for extended periods. Whether it is wise to lower light levels according to the EVE method under such conditions remains questionable. Would it not be wiser to give the extra performance advantage by using a light source with a high S/P value to workers working under such visually demanding circumstances? This especially so, because non-visual aspects such as emotion and mental effort also influence the pupil size. An increase in pupil size has been recorded in different situations with an increased workload (Just et al. 2003). As discussed previously in this chapter, under working conditions typical for many offices and factories, most of the visual tasks are far above the threshold of visibility (on or very near to the plateau of visual performance). The visual acuity advantages of the use of light sources with high S/P ratios as described in this section are here not relevant. EVE factors should, therefore, not be applied. Houser (2014), one of the authors of the IES publication about the EVE model, gives a more detailed argumentation for this. Berman revisited his earlier research data after the discovery of the pRGC cells. He concluded that they also correlate well with the, at that moment predicted, spectral sensitivity of pRGC cells (Berman and Clear 2008). In the meantime, more accurate information has become available on pRGC sensitivity (see Chap. 5). As these pRGC cells have a more dominant role in the pupillary reflex than the rods, it would be worthwhile to investigate if an even better correlation is obtained between pupil changes and parameters relating to the sensitivity of pRGCs instead of rods.

86

3 Visual Performance

References Alpern M, Campbell FW (1962) The spectral sensitivity of the consensual light reflex. J Physiol 164:478 Bailey I, Clear R, Berman S (1993) Size as a determinant of reading speed. J Illum Eng Soc 22:102–117 Berman SM (1992) Energy efficiency consequences of scotopic sensitivity. J Illum Eng Soc 21(1):3–14 Berman SM, Clear RD (2008) Past vision studies can support a novel human photoreceptor. Light Eng 16(2):88–94 Berman SM, Jewett DL, Bingham LR, Nahass RM, Perry F, Fein G (1987) Pupillary size differences under incandescent and high pressure sodium lamps. J Illum Eng Soc 16(1):3–20 Berman SM, Fein G, Jewett DL, Ashford F (1993) Luminance-controlled pupil size affects Landolt C task performance. J Illum Eng Soc 22(2):150–165 Berman SM, Fein G, Jewett D, Benson B, Law T, Myers A (1996) Luminance-controlled pupil size affects word-reading accuracy. J Illum Eng Soc 25(1):51–59 Blackwell HR (1981) Description of a comprehensive family of relative contrast sensitivity (RCS) functions of luminance to reflect differences in size of task detail, task eccentricity and observer age. J IES 11(2):52–63 Blackwell HR, DiLaura DL (1973) IERI. Application procedures for evaluation of veiling reflections in terms of ESI: II. gonio data for the standard pencil task. J Illum Eng Soc 2(3):254–283 Bodmann HW (1967) Quality of interior lighting based on luminance. Trans Illum Eng Soc (Lond) 32(1):23–37 Bouma H (1962) Size of the static pupil as a function of wavelength and luminosity of the light incident on the human eye. Nature 193:690–691 Boyce PR (2014) Human factors in lighting, 3rd edn. CRC Press, Boca Raton, FL Boyce PR, Slater AI (1981) The application of CRF to office lighting design. Lighting Res Technol 13(2):65–79 CIBS (1981) Technical memorandum 6, Lighting for visual display units. CIBS, London CIE (1981a) International Commission on Illumination Publication 019.21-1981, An analytic model for describing the influence of lighting parameters upon visual performance, vol 1, 2nd edn. CIE, Vienna CIE (1981b) International Commission on Illumination Publication 019.22-1981, An analytic model for describing the influence of lighting parameters upon visual performance, vol 2, 2nd edn. CIE, Vienna CIE (1984) International Commission on Illumination CIE Publication 60:1984, Vision and the visual display unit work station. CIE, Vienna CIE (2002a) International Commission on Illumination Publication 145:2002 The correlation of models for vision and visual performance. CIE, Vienna CIE (2002b) International Commission on Illumination Publication 146:2002 CIE collection on glare. CIE, Vienna CIE (2015) International Commission on Illumination CIE Technical Note 003:2015, Report on the first international workshop on circadian and neurophysiological photometry, 2013. CIE, Vienna Clear RD, Berman S (1990) Speed, accuracy and VL. J Illum Eng Soc 19:124–131 Crawford BH (1936) The integration of the effects from a number of glare sources. Proc Phys Soc (Lond) 48:35 De Boer JB (1977) Performance and comfort in the presence of veiling reflections. Lighting Res Technol 9(4):169–176 De Boer JB, Fischer D (1981) Interior lighting, 2nd edn. Kluwer Technische Boeken, Deventer Eklund NH, Boyce PR, Simpson SN (2001) Lighting and sustained performance: modelling dataentry task performance. J Illum Eng Soc 30:126–141 Enoch JM, Birch DG (1981) Inferred positive phototropic activity in human photoreceptors. Philos Trans R Soc Lond Ser B Biol Sci 291(1051):323–351

References

87

Ginsburg LM (1951) Ocular movements and fixations in reading. Am J Optom 28:605–615 Hoekstra J, Van der Groot DPA, Van den Brink G, Bilsen F (1974) The influence of the number of cycles upon the visual contrast threshold for spatial sine wave patterns. Vis Res 14:365–368 Holladay LL (1927) Action of a light source in the field of view in lowering visibility. J Opt Soc Am 14:1 Houser KW (2014) To use or not to use TM-24? Leukos 10:57–58 IES (1989) IES Rp-24 IES recommended practice for lighting offices containing computer visual display terminals IES (2013) IES TM-24-13 An optional method for adjusting the recommended illuminance for visually demanding tasks within IES illuminance categories P through Y based on light source spectrum Just MA, Carpenter PA, Miyake A (2003) Neuroindices of cognitive workload: neuroimaging, pupillometric and event-related potential studies of brain work. Theor Issues Ergon Sci 4:56–88 Lipetz LE (1971) The relation of physiological and psychological aspects of sensory intensity. In: Loewestein WR (ed) Principles of receptor physiology. Springer, Heidelberg Loe DL, Waters IM (1973) Visual performance in illumination of differing spectral quality. Environmental Research Group, University College, London Lynes JA (1982) Designing for contrast rendition. Lighting Res Technol 14(1):1–14 Markwell EL, Feigl B, Zele AJ (2010) Intrinsically photosensitive melanopsin retinal ganglion cell contributions to the pupillary reflex and circadian rhythm. Clin Exp Optom 93(3):137–149 McDougal DH, Gamlin PD (2010) The influence of intrinsically-photosensitive retinal ganglion cells on the spectral sensitivity and response dynamics of the human pupillary light reflex. Vis Res 50(1):72–87 McNelis J (1973) Human performance—a pilot study. J Illum Eng Soc 2:190–196 Muck E, Bodmann HW (1961) Die Bedeutung des Beleuchtungsniveaus bei praktische Sehtätigkeit. Lichttechnik 13:502–507 Naka KI, Rushton W (1966) S-potentials from colour units in the retina of fish (Cyprinidae). J Physiol 185:536–555 Navvab M (2001) A comparison of visual performance under high and low color temperature fluorescent lamps. J Illum Eng Soc 30(2):170–175 Navvab M (2002) Visual acuity depends on the color temperature of the surround lighting. J Illum Eng Soc 31(1):70–84 Rea MS (1981) Visual performance with realistic methods of changing contrast. J Illum Eng Soc 10 (3):164–177 Rea MS (1986) Toward a model of visual performance: foundations and data. J Illum Eng Soc 15:41–57 Rea MS (2012) The Trotter Paterson lecture 2012: whatever happened to visual performance? Lighting Res Technol 44:95–108 Rea MS, Ouellette MJ (1988) Visual performance using reaction times. Lighting Res Technol 20 (4):139–153 Rea MS, Ouellette MJ (1991) Relative visual performance: a basis for application. Lighting Res Technol 23(3):135–144 Smith S, Rea MS (1978) Proofreading under different levels of illumination. J Illum Eng Soc 7:47–52 Van Nes FL, Bouman MA (1967) Spatial modulation transfer in the human eye. J Opt Soc Am 57:401–406 Weale RA (1961) Retinal illumination and age. Trans Illum Eng Soc (Lond) 26(2):95–100 Weale RA (1963) The aging eye. HK Lewis, London Westheimer G (1964) Pupil size and visual resolution. Vis Res 4:39–45 Weston HC (1953) The relation between illumination and visual performance, Reprint IHRB Rep. No. 87 (1945) and Joint Rep. (1935). Medical Research Council, HMSO, London Weston HC (1961) Rationally recommended illuminance levels. Trans Illum Eng Soc (Lond) 26(1):1–16

Chapter 4

Visual Satisfaction

Abstract Visual performance as described in the previous chapter relates to the lighting of the task. The lighting of the whole space determines whether the overall appearance of the space is experienced as visually satisfying. The brightness of a space, the distribution of the luminance in that space, the directionality of the light, the degree of discomfort glare and the colour tint of light determine visual satisfaction. For the characterisation of the visual appearance of a room two different metrics are proposed: the average luminance in a horizontal band with a width of 40
, and the mean room surface exitance. The directionality of lighting determines the appearance of three-dimensional objects and faces in a space. The concept of flow of lighting allows for the calculation of the main direction and strength at a point in space as a result of all light rays at that point. The vector-to-scalar ratio can quantify, and light tubes visualise, the flow of lighting. It enables a detailed analysis of the spatial and formgiving potential of lighting designs. The unified glare rating UGR concept is used as a measure of the degree of discomfort glare. However, for glare sources with a non-uniform luminance, such as many LED matrix luminaires, the UGR concept needs modifications. The spectrum of the glare source influences also discomfort glare: short-wavelength light sources result in more discomfort glare than long wavelengths. It is questionable if correlated colour temperature-based rules work good enough to predict visual satisfaction with light sources of different tints of whiteness.

Visual performance as described in the previous chapter relates to the lighting of the task. Of course, lighting should not only be restricted to those areas of a space where visual tasks are being carried out. The lighting of the whole space determines whether the overall appearance of the space is experienced as visually satisfying. Visual satisfaction relates to such aspects as spaciousness, attractiveness, relaxedness, intimacy and worker’s stimulation. Veitch et al. (2011) reported in the publication “linking study appraisals to work behaviour” that “people who appraise their lighting as good, also appraise the room as more attractive, are in a more pleasant

© Springer Nature Switzerland AG 2019 W. van Bommel, Interior Lighting, https://doi.org/10.1007/978-3-030-17195-7_4

89

90

4 Visual Satisfaction

mood and more satisfied with the work environment, and more engaged in their work”. The overall brightness of the space, the distribution of the brightness in the space, the directionality of light, the degree of discomfort glare and the colour tint of light determine visual satisfaction. Many investigations into visual satisfaction make use of subjective assessment type of studies. Their results, together with the already described results of visual performance studies, are of great importance for lighting recommendations and lighting design. Too many lighting recommendations concentrate mainly on ensuring visual performance, while the aspect of visual satisfaction is underrated. Brandston, a renowned international lighting designer with over 50 years of experience, stated: “I always light the spaces first and then supplement for the tasks. I do not light for the tasks and then supplement for the spaces” (Brandston 2010).

4.1 4.1.1

Spatial Brightness Brightness-Luminance Relation

Fechner (1860) introduced psychophysics as a professional discipline. In his 1860 book “Elemente der Psychophysik” he introduced his famous Weber-Fechner law, which says that the subjective sensation of sight, sound and weight (all psychological elements) is not proportional to the intensity of the stimulus (a physical element). In lighting terms, this means that a variation of luminance of a surface (the stimulus) does not result in a proportionally equal variation of experienced brightness of that surface (the sensation). For example, increasing the luminance of a surface by a factor of two results in a brightness increase that is clearly less than the factor of two. The Weber-Fechner law says that the sensation is proportional to the logarithm of the intensity of the stimulus. A century after this law was defined, it became clear that an exponential exponent improves the relationship (Stevens 1961): Sensation ¼ c  Stimulusa where c and a are constants (according to Wade and Swanston 2013, the constant a is 0.6 for loudness and 0.5 for brightness). The luminance of the surface itself does not solely determine how bright a surface is experienced. It is also determined by the luminance, size and position of the surrounding surfaces (Marsden 1969; Stevens 1975; Bodmann and La Toison 1994). An important reason for this is that with an increase in the size of the surrounding surface, a larger surface is projected on the retina so that a larger part of the retina participates in the vision process. In the central part of the fovea there are mainly M- and S-cones, while in the peripheral area of the fovea also blue-sensitive L-cones are located (see Fig. 1.6 of Chap. 1). In the non-foveal part of the retina also photosensitive retinal ganglion cells and even rods may play a role (Berman et al.

4.1 Spatial Brightness

91

1990; Berman 2008; Viénot et al. 2009; Vidovszky-Nemeth and Schanda 2012). This phenomenon has important consequences for how lighting is experienced. An extensive study done by Marsden (1970) under interior lighting conditions gives a good insight. He compared the subjective brightness of room surfaces, as experienced by observers, with the measured luminance of the same surfaces, in a room where the luminance of the various room surfaces could be changed. He defined the brightness scale relative to a reference situation of a surface wall in the room with a luminance value of 30 cd/m2. The brightness of this reference situation was, arbitrary, set at a value of 10. Subsequently, the observers were asked to appraise the brightness of other surfaces relative to the reference condition. For example, a surface appearing twice as bright as the reference situation obtains a brightness value B of 20. One appearing half as bright gets a value of 5. Marsden derived from his experiments an exponential relationship between the brightness (B) of a surface in the room, the luminance value Lsurf of that same surface and the luminance of another surface in the same room having the highest luminance value, Lmax: B¼c

L0:58 surf L0:21 max

Figure 4.1 shows this relationship in graphical form. Before discussing the graph, it should be noted that, although the trend between different test persons is similar, there exist large individual differences between persons as far as perceived brightness is concerned. Fig. 4.1 Relative brightness of a surface in a room with luminance Lsurf, as a function of the luminance of another surface having the highest luminance in that room, Lmax. The thin line represents situations where the actual surface being appraised has the maximum luminance (Lsurface ¼ Lmax). For the area Lsurface < 3 cd/m2 no tests are carried out and therefore the lines are shown as dotted. Lines are based on the relationship derived by Marsden (1970) and Coaton and Marsden (1997)

Relative Brightness 30

Lsurf = Lmax

c

10 6.2

b 300 Lmax

30

3

3 2.6

a

1 1

3

10

30

100

300

Lsurface (cd/m2)

92

4 Visual Satisfaction

Fig. 4.2 The sensation of the brightness of the small squares, having the same luminance, is different because of their different surround luminance

Brightness does increase with a smaller factor than the factor with which the luminance of a surface is increased (compare the thick drawn lines of the investigation results with the thin broken line that gives the situation when brightness would have exactly followed the changes in luminance). The figure also shows that the brightness experienced by a surface decreases when in the surroundings of that surface there are other surfaces with higher luminances. See for example the difference in brightness between point c (Lsurf ¼ 30 and Lmax ¼ 30) where the brightness B has a value of 10 and point b (Lsurf ¼ 30 and Lmax ¼ 300) where the brightness value has decreased to only 6.2. The higher the maximum luminance relative to the actual surface luminance being appraised, the lower the brightness is assessed. Figure 4.2 also demonstrates this. Here the small dark square on the right, seen against a field of a high-luminance surround, has a lower brightness than the small square on the left, although the luminance of the two small squares is the same. For the lighting designer this is important to realise: the larger the luminance differences of the surfaces in a room, the lower the experienced brightness of the darker surfaces. To increase that brightness it is more effective to increase only the luminance of the surfaces with the lower luminance values than increasing the luminance of all surfaces in the room. Compare in Fig. 4.1 point a (Lsurf ¼ 3 and Lmax ¼ 30) with point b where all luminances are a factor of 10 higher (Lsurf ¼ 30 and Lmax ¼ 300) and point c where only the surface luminance is a factor of 10 higher (Lsurf ¼ 30 and Lmax ¼30). Point b, with all luminances a factor of 10 higher, results in a brightness of 6.2, while point c, with only the surface luminance increased by a factor of 10, results in a higher brightness of 10. Coloured surfaces and different lamp spectra also lead to differences in perceived brightness. For highly saturated coloured surfaces the brightness is appraised higher than for less saturated colours (Boyce and Akashi 2002; Boyce 2014). The light sources used by Marsden, although consisting of two different types, were of nearly equal correlated colour temperature. For luminance levels larger than 8 cd/m2, halogen lamps were used with a CCT of around 3000 K, and for luminance levels lower than 8 cd/m2, tubular fluorescent tubes, also with CCT of 3000 K.

4.1 Spatial Brightness

4.1.2

93

Influence of Spectrum

Many studies have demonstrated that the spectrum of a light source that illuminates a room surface affects the perceived brightness of that surface in a way that is not linearly related to how the spectrum affects the luminance of the same surface. One important reason is that given in the previous section for the effect of the size of the surrounding surface on brightness. Not only the fovea determines the brightness of a large surface, but also the peripheral parts of the retina, so that all cone types, photosensitive retinal ganglion cells and perhaps also the rods play a role in defining the brightness. However, the luminance of the same surface is determined by V(λ) which only takes into account the effect of cones (especially the L- and M-cones). Another reason is of chromatic character, the so-called Helmholtz-Kohlrausch effect, illustrated in Fig. 4.3. The coloured squares at the top of Fig. 4.3, all with the same luminance, appear brighter than their achromatic background, with the same luminance. Yellowishgreen colours are an exception. The more saturated the colours are, the stronger the effect is. The bottom shows the colours converted to grey, still with the same luminance. Helmholtz used the term “glowing” to describe the appearance of saturated colours (Kuehni 2016). Chromatic adaptation reduces the effect the spectrum has on brightness. Chromatic adaptation is a colour-balancing process of the visual system that ensures that the appearance of colours under different lighting conditions is, to a certain extent, preserved. It is an aspect of perceptual constancy, discussed already in Sect. 1.8.1. Chromatic adaptation requires approximately 1 min to reach a steady state (Fotios 2006). Since, most probably, the long-term impression of the lighting of a space is more important than the first, short-time, impression, studies that permit the test persons to view the test surfaces for longer than 1 min seem to be relevant for the lighting design practice. CIE published in 2011 a Technical Report that permits the evaluation of brightness in dependence of luminance obtained from different lamp spectra. The method described is valid for a viewing field smaller than 10
. It is therefore of relevance for evaluating the brightness of light sources when looking towards the light source. The

Fig. 4.3 HelmholtzKohlrausch effect. Saturated colours (top) are seen more bright than when the colours are converted to grey while keeping the same luminance (bottom). The one exception is for the yellowish colour

94 Fig. 4.4 Correlated colour temperature ratios, CCT1/ CCT2, for lamp pairs resulting in equal brightness, as a function of their luminance ratio L1/L2 (Fotios et al. 2012)

4 Visual Satisfaction

CCT1/CCT2 2.5 2.0 1.5 1.0 0.5 0 0.6

0.7

0.8

0.9

1.0

1.1

L1/L2 method is, however, not applicable for large bright surfaces and thus not for spatial brightness. In a joint effort of the universities of Sheffield (UK), Pennsylvania (USA) and Aalborg (DK), Fotios et al. (2012, 2015a) reviewed 70 studies regarding the influence of lamp spectrum on spatial brightness. He concluded that 19 of these studies provide credible evidence of spectrum effects on appeared brightness. Studies from which complete enough data of the spectra were available were subsequently used for further analysis. Based on those studies Fig. 4.4 shows luminance ratios for different lamp pairs giving equal brightness, as a function of the corresponding correlated colour temperature ratios of those lamp pairs. Figure 4.4 shows that there are lamp spectra with which the same brightness can be obtained for an approximately 25–30% lower luminance. Figure 4.4 also shows that the correlation between the parameters is quite weak. It implies that the correlated colour temperature of a lamp is not a good predictor for the actual effect a lamp spectrum has on brightness. Berman et al. (1990) tested the so-called scotopic-photopic ratio S/P as a metric to predict the brightness effect of a lamp spectrum. It is the ratio between the scotopicweighted, V0 (λ), and the photopic-weighted, V(λ), lamp spectrum. It is a well-known metric used in road lighting for predicting mesopic vision effects. Berman indeed proved that the S/P ratio is a better predictor for brightness under photopic conditions than the correlated colour temperature. This may be attributed to the S/P ratio better taking into account the contribution of the photosensitive ganglion cells. However, in a recent study, designed as an extension of the Berman study (Fotios et al. 2015b), it appears that the S/P ratio alone is still not a good enough predictor. This new study concludes that the chromatic contribution to brightness (see for example Fig. 4.3) should also be part of a predictor. The study suggests that this can be done by the simultaneous use of S/P ratio and “gamut area”, the latter being a metric related to colour rendering that has been described in Chap. 2. This result hopefully paves the

4.2 Room Appearance

95

way for the derivation of a relationship between brightness and such a dual-lamp spectrum metric. It has already been shown that the effect of the spectrum on brightness is age dependent (Royer and Houser 2012). A relationship should, therefore, also take the age effect into account.

4.2

Room Appearance

The visual appearance of a room determines for an important part whether the lighting has a positive impact on persons in that room (Flynn et al. 1973; Loe et al. 1994; Tiller and Veitch 1995; Veitch and Newsham 2000; De Vries et al. 2015; Oi and Mansfield 2015). Most of the psychophysical studies that have demonstrated this relate to office type of spaces. These studies try to relate, through assessments by test persons, subjective aspects with the luminance and luminance distribution of the surfaces in the space.

4.2.1

Room Surface Illuminance and Luminance

4.2.1.1

Task Surface

At the outset of describing subjective assessment studies, it is important to point to the large assessment differences between individuals. De Boer and Fischer (1981) demonstrated this by evaluating nine different investigations of preferred horizontal illuminance levels on the task surface. A total of 1930 test persons were asked, after having completed a simple working task in the test room, to indicate whether they appraised the horizontal lighting level on their task surface as being “too dark”, “satisfactory” or “too bright”. Figure 4.5 shows the calculated means of the percentages for these appraisals. Fig. 4.5 Calculated means of nine assessment studies of horizontal task surface illuminance in working interiors (full drawn curves). Thin broken line curves: individual assessment “satisfactory” for the individual studies (De Boer and Fischer 1981)

Assessment 100

%

satisfactory too dark

80

too bright

60 40 20 0 100

200

500

1000 2000

5000

10000 20000

Eworking plane (lux)

96

4 Visual Satisfaction

From the figure, it is clear that there is no illuminance level at which all people are satisfied. Even at the point of optimum satisfaction, some 10% of the persons assess the situation as too dark and another 10% as too bright. Where at an exceptionally high lighting level of 10,000 lux the large majority assess the situation as “too bright”, there are nevertheless still some 5% of persons who judge that situation as satisfactory. In the heating and air-conditioning profession it is more commonly accepted that it is impossible to satisfy everybody with the same condition than in the lighting profession. The illuminance level at which the optimum satisfaction is obtained (approx. 2000 lux) is higher than that needed for good visual performance and also higher than the recommended lighting levels in standards and recommendations. This is confirmed by later studies (Van Ooyen et al. 1987; Manov 2007). Van Ooyen et al. (1987) showed that preferred lighting levels on the task area are dependent on the task being carried out. They demonstrated, for example, that working with visual display units (VDUs) requires lower levels than reading paper. The illuminance level on the task surface and the luminance of the walls influence each other’s preference (van Ooyen et al. 1987; Houser et al. 2002; Inoue 2010; Chraibi et al. 2017). Chraibi’s study even shows that the uniformity of the wall luminance influences the preference for the illuminance on the task surface. A less uniform wall luminance permits, for the same preference, a lower task surface illuminance.

4.2.1.2

Room Boundary Surfaces

The lighting of the room boundaries determines for a considerable part the room appearance as a whole. Semantic differential scales are often used to assess appearance. Table 4.1 gives examples of such scales for different subjective aspects, Table 4.1 Examples of semantic differential scales used in room appearance assessments (Loe et al. 1994; Veitch et al. 2011)

Visual lightness Bright–dim Radiant–gloomy Spacious–cramped Glaring–not glaring

Visual interest/attractiveness Interesting–non-interesting Interesting–monotonous Non-uniform–uniform Pleasant–unpleasant Stimulating–subdued Dramatic–diffuse Comfortable–uncomfortable Tense–relaxing Like–dislike Beautiful–ugly Attractive–unattractive Colourful–colourless Sombre–cheerful Large–small

Aspects with an  are investigated in Loe et al. (1994)

4.2 Room Appearance

97

Fig. 4.6 Mock-up conference room with different dominant areas that were investigated by Loe et al. (1994)

C40

C20 B20 B40

40˚

B40

categorised into two groups (Loe et al. 1994; Veitch et al. 2011). One category relates to the judgement concerning “visual lightness” and the other to “visual interest” or “visual attractiveness”. The classic assessment study of room boundary surfaces is that of Loe et al. (1994). The full-scale mock-up office room used for the experiments (Fig. 4.6) was furnished as a small conference room. Eighteen widely different lighting situations were assessed by 12 observers (all with a professional background in lighting) from a fixed position in the middle of the long side of the room. The subjective aspects assessed are indicated with an  in Table 4.1. Lighting varied in both lighting level and lighting distribution. The experimental room was so arranged as to exclude as far as possible a direct view of the luminaires (Mansfield 2010). Therefore, effects of luminaire luminances were not included in this experiment. Detailed luminance measurements of all room surfaces were made for all lighting situations. From these, the average luminance, Lav, and luminance uniformity expressed as Lmax/Lmin were calculated for the whole area, for the areas represented by circles subtending an angle of 20
and 40
, respectively, and for horizontal bands with a width of 20
and 40
, as illustrated in Fig. 4.6. Note that the 20
circle and band normally only include wall surfaces, while the 40
circle and band, for larger rooms, also include part of the floor and ceiling surface. This becomes clear by comparing the small room, sketched with a broken line at the bottom of Fig. 4.6 with the larger room drawn in a full line. All assessment data obtained were investigated using the statistical process of factor analysis. This was done to identify interrelated subjective aspects and bring them under one common factor. In this way, two such factors were obtained. One that can be described as “visual lightness” covers the interrelated subjective aspects of

98

4 Visual Satisfaction

“bright–dim”, “radiant–gloomy” and “spacious–cramped”. The other factor, which can be labelled as “visual interest”, includes the aspects “non-uniform–uniform”, “interesting–uninteresting”, “pleasant–unpleasant”, “stimulating–subdued” and “dramatic–diffuse”. In the next step, the correlation between these two factors and the measured luminances of the different room areas was tested. The factor “visual lightness” correlates best with the logarithmic value of the average luminance of the band with a width of 40
(B40 band). Probably, the reason for the good correlation with this band is that it encompasses the main area of view when a person looks around a space. The factor “visual interest” correlates best with the logarithm of the uniformity ratio Lmax/Lmin, again of the B40 band. Although all of the interrelated subjective assessments showed a good correlation, there are still individual differences. Of the “visual lightness” factor, it is the “bright–dim” assessment that has the highest correlation with the average luminance in the B40 band (coefficient of determination 0.92). Of the “visual interest” factor, it is the assessment of “interesting–uninteresting” itself that has a good correlation with the Lmax/Lmin ratio of the B40 band (coefficient of determination 0.71). Figure 4.6 shows the relationship between the “bright–dim” assessment and the logarithm of the average luminance in the B40 band. The curve is drawn, by Loe et al., on the assumption that it follows the familiar S-shape, showing at high luminance value saturation. The midpoint of the assessment scale, where assessment changes from dim to bright, corresponds to a value of the average luminance in the B40 band of 30 cd/m2. This value can be considered as the minimum required in an office environment to have a sufficient degree of “visual lightness”. For a reflectance value of 0.5, this luminance value corresponds to an illuminance value on the room boundary surfaces of 190 lux. Twenty years later, Kirsch (2014) made a visual lightness assessment study in a cell type of office mock-up. The walls and ceiling consisted of back-lit (3000 K) diffuse acrylic panels that could be dimmed separately to generate many different room surface luminances. Sixty-eight test persons (naïve in lighting, average age 26.3 years) were assessed from behind a desk (with a constant illuminance of 500 lux) for visual lightness and visual attractiveness of the office. Correlation of the average assessments of the test group with the average luminances of different room areas was tested. The best correlation was found to exist with the average luminance of the B40 band (coefficient of determination 0.86). The small squares in Fig. 4.7 give the results. They are in good agreement with the earlier results of Loe et al. The test persons were also asked to assess perceived attractiveness. This subjective aspect also correlated best with the average luminance in the B40 band (coefficient of determination 0.87). Interesting enough, Kirsch checked the chronotype of his test persons. He could show that there is a significant difference between morning and evening chronotypes. Morning types require in the evening higher luminances than in the morning for a same assessment of brightness. With evening types it is just the other way around: they need in the morning higher luminances than in the evening. At the midpoint, the difference is around 20 cd/m2.

4.2 Room Appearance Fig. 4.7 Bright–dim assessment, being representative for the factor “visual lightness”, as a function of the average luminance Lav,B40 of the B40 band (Loe et al. 1994; Kirsch 2014)

99

Assessment bright Loe et al. Kirsch

mid

dim 3

1

10

30

100

300

Lav, B40 (cd/m2) Fig. 4.8 Visual interest as a function of the ratio Lmax/ Lmin,B40 of the B40 band (Loe et al. 1994). The results for the appraisal attractiveness (Kirsch 2014) are given on the same assessment scale

Visual Interest interesting attractive

Loe et al. Kirsch

mid

unattractive

uninteresting

3

15

10

30

85

100

300

Lmax/Lmin, B40

Figure 4.8 gives the uniformity results of the Loe et al. (1994) study for the “interesting–uninteresting” aspect as a function of the logarithmic value of the luminance ratio Lmax/Lmin of the B40 band. Loe et al. did draw the full line, as visually representing the relationship. It shows that the larger the ratio of Lmax/Lmin, and thus the more non-uniform the luminance distribution in the B40 band, the more interesting the room is appraised. Kirsch, in his 2014 study, tested the relationship between the same Lmax/Lmin ratio in the B40 band with the aspect “attractiveness” for different lighting situations all with strong non-uniform room luminance distributions. Attractiveness is most probably closely related to the “interesting” aspect. In this study, with 34 test persons (average age 26.8 years), the luminance in the B40 band was held constant at 40 cd/m2. The small squares in Fig. 4.8 give the results. The thin broken-line curve in Fig. 4.8 is the best trend curve when the results of Loe et al. and Kirsch are combined. The relationship

100

4 Visual Satisfaction

that this curve suggests makes sense. Uniform room luminance distributions lead to an appraisal “uninteresting” and “unattractive”. The more non-uniform the situation becomes (larger Lmax/Lmin ratio), the more interesting and attractive lighting is appraised until the maximum appraisal is obtained. From that situation onwards the appraisal decreases with a further increase in non-uniformity until it becomes unacceptable. Taking the midpoint of the assessments again as a basis, the requirement for a sufficient degree of visual interest is for the ratio Lmax/Lmin in the B40 band to be between approximately 15 and 85. Kirsch also tested the relationship between the aspect of visual lightness and non-uniformity. Although in these tests the average luminance in the B40 band was held constant at 40 cd/m2, the appraisal of visual lightness varied from below acceptable to above acceptable with a change in uniformity. It is another indication that both the aspect of the quantity of “room light” and “distribution of room light” are needed to describe subjective satisfaction obtained from lighting. In a recent study, Oi and Mansfield (2015) analysed again the original data obtained for the factor “visual interest” by Loe et al. (1994). They, first of all, propose, instead of the ratio of Lmax/Lmin, the more robust luminance ratio Lmax/ Lav. They do this because the average value of a set of individual values is more accurate than the single minimum value from that set. Also, average luminance measurements are more accurate than single-point measurements. They also propose the use of luminance histograms and Fourier analysis for the characterisation of the luminance pattern in the room. More work is needed to see if these advanced methods are suitable for the daily lighting design practice. Early studies agreed well with the result of at least 30 cd/m2 wall luminance of the 1994 study of Loe et al. (Tregenza et al. 1974; Flynn et al. 1975; Van Ooyen et al. 1987). Later assessment studies indicate for an adequate quality wall luminance values varying from approximately 20 to 70 cd/m2 (Berrutto et al. 1997; Veitch and Newsham 2000; Loe et al. 2000; Newsham et al. 2004; Kirsch 2014; De Vries et al. 2015; Duff 2015; Duff et al. 2017a, b). The lower end of this range is for situations that involve work with visual display units. One study, that takes the average of wall and ceiling luminance as a basis, reports higher required luminance values of around 100 cd/m2 (Miller et al. 1995). Some of the studies mentioned above give separate values for wall and ceiling luminance. The preferred values for the ceiling are up to a factor of 2 higher than for the walls. The importance of a bright enough ceiling for getting the assessment “good” for the overall space has probably to do with an unconscious comparison with a natural outdoor environment where the sky is the brightest part of the visual scene. Figure 4.6 (bottom) shows that for small offices the ceiling is not part of the B40 band. For this reason, it is perhaps useful to have luminance requirements for the luminance of the B40 band and a separate requirement for the luminance of the ceiling. An extensive assessment study in which an image of an office cubicle was presented to a large number of observers resulted in required wall luminance values between 40 and 60 cd/m2. The required ceiling luminance was approximately 80 cd/ m2 (Newsham et al. 2004). In this study, the luminance of six areas of the image of the office cubicle could be varied independently with the aid of a projector. The

4.2 Room Appearance

101

projection of each subsequent image was selected with a so-called genetic algorithm, so that the most attractive assessed image influenced the next image: the “parent” image influences the “child” image. A study with this set-up and procedure is efficient, faster and cheaper than conventional mock-up assessment studies. Since it showed results well in line with the studies of the mock-up type, this approach is worth following.

4.2.1.3

Metric for Subjective Lighting Quality: B40 Band Based

By calculating the average luminance and luminance ratio for the B40 band, it is possible to evaluate lighting designs with the aid of Figs. 4.7 and 4.8, for each specific observer position and each viewing direction. However, what lighting specifiers and designers would help even more is one single or a restricted number of overall measures that are characteristic for the quality of the lighting installation as a whole. Loe et al. (2000) did a pilot study to investigate such an overall measure. They based their candidate measure on an averaging procedure of the values obtained in the B40 band for many different observer positions and viewing directions. They defined for this purpose a regular grid of approximately 1 m  1 m. On each grid point, they measured for each of the four perpendicular viewing directions the average luminance in the B40 band. Figure 4.9 shows a sketch of the principle of this procedure. They characterised this set of luminances subsequently by its arithmetic mean and standard deviation. The measurements were made in a test room for eight different lighting situations. A small group of 12 test persons made assessments for these lighting situations in the same test room. Half of them made their assessments while viewing the room from a fixed position in the middle of one of the walls, while the other half made their assessments while moving around the room as they wished. The results showed a high correlation between the “bright–dim” assessment and the mean value of the set of measured average luminances (coefficient of determination 0.93). A mean value of at least 40 cd/m2 is required, being close to the 30 cd/m2 of the 1994 study. The “interesting–uninteresting” assessment showed a moderately high coefficient of determination with the standard deviation of the average luminances of all B40 bands (coefficient of determination 0.64). The standard deviation needed is

Fig. 4.9 Procedure for averaging the average luminances in the B40 band for four viewing directions at a large number of observer positions

102

4 Visual Satisfaction

Fig. 4.10 Example of different zones in a “visual interest–visual lightness” diagram, indicating different requirements for different types of lighting applications (Loe 2016)

high

Visual interest (St Dev for B40)

Zone A

Zone B Zone C

low low

high

Visual lightness (Lav, B40) approximately 10 cd/m2. Although this investigation is only a pilot, it shows a possible way towards two single overall measures for specifying lighting installations to ensure adequate visual lightness and visual interest, respectively. The luminance in the B40 band can easily be measured with a conventional illuminance meter with the cosine-corrected photocell mounted with its face vertical and fitted with screens that limit the field of view vertically 40
and horizontally 90
. The instrument has to be calibrated for measuring luminance. It may be expected that lighting applications other than office lighting require different values. Applications as office lighting, industrial lighting and restaurant lighting would occupy different zones in a “visual interest–visual lightness” diagram. Loe (2016) sketched a schematic diagram to illustrate this (Fig. 4.10). Perhaps such a diagram could be used as a basis for specifying lighting recommendations for different applications.

4.2.2

Mean Room Surface Exitance, MRSE

4.2.2.1

Indirect Illuminance at the Eye

Cuttle (2008, 2010, 2018) proposed an entirely new way to characterise perceived brightness and adequacy of the lighting of a space. He uses as a measure the amount of light, reflected from all the surfaces and objects in that space, arriving at the eye of an observer, without including the direct light on the eye from the light sources. This is the illuminance at the eye due to the indirect light coming from all surfaces. The direct light arriving at the eye is excluded because it does not contribute to a positive room appearance but to the negative factor glare. The photometric measure for the amount of light leaving a surface is “exitance”; it is expressed in lumen per square

4.2 Room Appearance Fig. 4.11 Cross section of two offices with the same mean room surface exitance (MRSE) value of 100 lm/ m2. Top: all room surface reflectances equal, resulting in same illuminance values for all viewing directions (average at each point in the space is 100 lux). Bottom: different room surface reflectances resulting in different illuminances for the different viewing directions (average at each point in the space is 100 lux)

103

100 100 100 100

100 100

100

100

100 100

100 100

180

200 100 190 60

50 50

170

10 40

100 50

meter of the radiating surface, lm/m2. The difference between exitance and illuminance is that exitance relates to the amount of light leaving a surface and illuminance to the amount of light arriving at a surface. The unit of both measures is lm/m2. Only in the case of illuminance the unit lm/m2 has its own specific name: lux. Cuttle showed that, for rooms with all room surfaces having the same reflectance, the mean of the exitances from all surfaces is equal to the indirect illuminance at any given point in the room, for all orientations of the plane of reference of that illuminance. The eye of an observer in a room is also such a point. Therefore, for rooms with room surfaces with a same reflectance, the single value, mean room surface exitance (MRSE) of the lit room, is equal to the illuminance on the eye of an observer in that room. This holds for each viewing direction and each location in the room. Figure 4.11, top, illustrates this for a room with an MSRE value of 100 lm/m2. For rooms with different room surface reflectances, the situation is more complex. The illuminance on the eye at a point in space then varies with viewing direction (Fig. 4.11, bottom). MRSE now equals the average illuminance at a point, where the averaging refers to all planes perpendicular to all viewing directions. The single value MRSE gives information about the mean illuminance on the eye but gives no information about the illuminances for the different viewing directions. Cuttle expected MRSE to correlate well with the perceived brightness and subjective adequacy of the lighting of a room when assessed with complete free viewing directions. In a series of experiments, it has indeed been shown that there exists a good correlation between MRSE and perceived brightness and subjective adequacy of a room (Duff 2015; Duff et al. 2017a, b). Duff and Duff et al. did two studies in a small office type of room. They made for these studies 27 different lighting scenes with various combinations of indirect, direct and mixed lighting

104 Fig. 4.12 Average of subjective appraisal for room brightness, Ba, as a function of mean room surface exitance, MRSE, as obtained from appraisals in a uniform-to-moderate uniform-lit office and an extreme non-uniform-lit office, respectively (Duff 2015; Duff et al. 2017b)

4 Visual Satisfaction

Appraisal (brightness) 7 uniform to moderate uniform non-uniform

4

1 20

40

60

80

MRSE

100

(lm/m2)

(4000 K and Ra of 80) and with various combinations of room reflectances. In one study the lighting varied between uniform and moderately uniform, while in the other study the different lighting situations were, for an office lighting application, extremely non-uniform. Twenty-six young observers (average age 20.8 years) assessed the perceived brightness of each lighting scene. The test persons were asked to relate brightness to the entire space, and not solely to their immediate field of view. Figure 4.12 gives the average results of all observers. The two studies show a strong linear relationship between perceived brightness and MRSE, with coefficient of determination of 0.79 and 0.95, respectively. The studies also tested a possible relationship between perceived brightness and horizontal illuminance. As to be expected, no noticeable relationship was obtained. That the curves show no levelling off at high MRSE values is most probably because of the limited range of MRSE values studied. Taking the midpoint of the brightness assessment scale as the minimum requirement, it follows from Fig. 4.12 that an MSRE value of slightly more than 100 lm/m2 is needed. The test persons were also asked to give their appraisal for the perceived adequacy of the illumination (named by Cuttle as “PAI”). Figure 4.13 shows the relationship obtained between perceived brightness and perceived adequacy of illumination PAI on the basis of the average appraisals of all test persons. For the more uniform-lit office, there exists a good correlation between perceived brightness and adequacy of lighting. Increase in perceived brightness improves the general satisfaction with the lighting. For the very non-uniform-lit office, the relationship is different: with increasing perceived brightness, the overall satisfaction with the lighting hardly improves. It suggests that for more extreme luminance distributions, as may occur especially in the non-office type of lighting applications, subjective adequacy is not solely related to perceived brightness. Here, apart from room brightness, another aspect plays a role as well. This is in line with the results of Loe et al. and Kirsch, discussed before, which indicate that a measure for both the quantity and the uniformity of light is needed to describe visual satisfaction.

4.2 Room Appearance

105

Fig. 4.13 Average subjective appraisal of perceived brightness as a function of the percentage of test persons who assessed the lighting as being adequate (PAI), for a uniform-to-moderate uniform- and an extremely non-uniform-lit office, respectively (Duff 2015; Duff et al. 2017b)

Appraisal (brightness) 7 uniform to moderate uniform non-uniform

4

1 0

10

20

30

40

50

60

70

Perceived adequacy of illumination (PAI) (%)

4.2.2.2

Formulas for the Calculation of MRSE

Cuttle (2008, 2010) derived the formula for the calculation of MRSE through a “thought experiment” for an imaginary space having surface reflectances being the same for ceiling, walls and floor. For the conditions of the thought experiment MRSE equals the total of the first reflected luminous flux from each surface divided by the room absorptance, being the sum of the surface areas multiplied with their absorptance values. For the equal-reflectance condition, the MRSE formula can be written as MRSE ¼

FRF Aα

with FRF ¼ first reflected flux (sum of direct flux reflected from each surface) Aα ¼ room absorptance (sum of the surface areas multiplied with their absorptance values) The general formula to calculate MRSE, valid for rooms with different room surface reflectances, for which not only the first reflection but also all further interreflections have to be taken into account, is derived by Duff et al. (2016a) as P MRSE ¼

ðM s  A s Þ P ¼ As

P

Ls  π  As P As

with MRSE ¼ mean room surface exitance (lm/m2) Ms ¼ mean exitance of each surface within the room As ¼ area of each surface within the room Ls ¼ average surface luminance of each surface within the room

106

4 Visual Satisfaction

Duff et al. (2016b) have produced a software script for the calculation of MSRE that can be used by software developers to implement in their lighting software. Duff et al. (2016a) also described a procedure for measurement of MSRE in practical situations. It uses captured high dynamic range (HDR) images made with a reflex digital camera (image luminance measuring device, see also Chap. 18).

4.2.2.3

MRSE-Based Metric for Subjective Lighting Quality

It has already been mentioned that for lighting specifiers and designers it is helpful to have one single or a restricted number of overall measures that characterise the subjective quality of a lighting installation as a whole. In Sect. 4.2.1.3, a two-metric system, based on the luminance in the so-called B40 band, has been discussed. Cuttle (2013) proposes another two-metric system for the same purpose. The first metric is MRSE for perceived adequacy of illumination (PAI). As discussed above, it relates to a sufficient reflected flux in a room or space. The second metric relates to the distribution of light in the space, which Cuttle calls “hierarchy of illumination (IH)”. The intention here is to provide the lighting designer with a tool to give emphasis to selected areas or objects in a space with direct light from the luminaires. The metric used for this second aspect is called TAIR, which stands for “target/ambient illuminance ratio”. It is the ratio of local illuminance incident on a target surface or object (indirect and direct) to the ambient illumination level as expressed by MRSE. In a formula TAIR ¼

Etg E tg , indir þ Etg , dir ¼ MRSE MRSE

with Etg ¼ illuminance on target. The proposed lighting design procedure with Cuttle’s two-metric system is based on the prerequisite that MRSE values are specified in lighting standards for different types of applications. The designer makes his or her design so that it fulfils the relevant MRSE requirement by distributing the indirect flux. Then the designer decides which surfaces and objects need how much extra emphasis and expresses that in TAIR values. It may be done by taking architectural and aesthetical criteria into account. Next, the designer directs the direct luminaire flux so that his or her TAIR requirements are fulfilled as well. The addition of TAIR as a second metric to be used in combination with MSRE is important. This because it permits emphasising different areas with light, for both visual satisfaction and visual performance reasons. It has been pointed out already that for spaces with different room reflectances, MSRE describes the average quantity of reflected light on the eye, but includes no information about the differences for various viewing directions. It has also been demonstrated that other studies indicate that there exist so-called dominant areas in a room. They have a larger influence on perceived brightness than other less dominant areas. Horizontal surfaces seem to have a larger effect on perceived brightness than

4.2 Room Appearance

107

vertical surfaces (Kato and Sekiguchi 2005). Work of Kirsch (2014) is interesting since it varied lighting while keeping the MSRE value constant. In six non-uniform luminance tests, the different room boundary surfaces were varied while keeping the average of each surface luminance the same. This implies that the MRSE value for all six situations was the same. Nevertheless, the average appraisals of the test persons for visual brightness were different for these situations. Rather extreme non-uniform luminances, including stepwise luminance transitions over a wall, were included in this study. These occur in practice also, for example in the case of windows in a wall which have near-zero reflectance. The fact that MSRE has no information about the luminance distribution in the space reduces the suitability of MSRE as a metric for visual satisfaction. Raynham et al. (2019) suggest that MSRE and PAI may break down in more complex real environments. Complex here relates to the shape of the room and “dramatically” variation of the lighting conditions across the room. In an L-shaped room, there are positions where an observer cannot see all room surfaces. The surface exitance of invisible parts of the room can thus have no influence on that observer’s adequacy of illumination. The authors, therefore, propose instead of MRSE the use of mean indirect cubic illuminance (MICI). It is the average of the six indirect illuminances on the faces of a cube. In the case of all room surfaces having the same exitance, MICI equals MRSE at all points in the space. The authors also show that in a variety of non-complex rooms, the average value of MICI is nearly equal to the MRSE value. MRSE is independent of location within a room and does not differentiate for different viewing directions. MICI is dependent on the position in the room and does not differentiate for different viewing directions because of the averaging over the six faces of the cube. In an earlier study, Raynham (2016) examined the consequences of adopting a proposed MRSE value of 100 lm/m2 in two different-sized offices with a variety of surface reflectances. He concludes that, relative to today’s office-lighting practice, a significant extra luminous flux is required with a significant upward component. The latter is because aiming light at the surfaces with the higher reflectances (like the ceiling) is most effective for increasing MRSE. More research is needed to see if the combination of mean surface exitance MRSE (or the alternative of mean indirect cubic illuminance MICI) and target/ ambient illuminance ratio TAIR can be used as a general approach for interior lighting designs.

4.2.3

Items for Further Study

Some of the studies dealt with above made use of test persons with a lighting background, and others with persons naïve in lighting. Future studies are needed to find out whether or not there is a difference between these types of test persons. The experience of test persons with lighting may also play a role. That could mean that people from different parts of the world with significant different lighting

108

4 Visual Satisfaction

conditions in their homes and working places would appraise lighting situations differently. The majority of the tests have been carried out by test persons not carrying out a real office task. One study where room appearance was assessed while carrying out different office tasks showed an influence of the actual office activity on the appraisal of room luminance uniformity (Kobayashi et al. 2001). With desk type of work and conversation more uniform situations were appraised better. Non-uniform situations were preferred when thinking and when taking a break. The influence that the actual activity has on preferred lighting earns more attention. Different investigations about the effect of age report conflicting results (Boyce 1973; Knez and Kers 2000; Juslén et al. 2005). Specific studies on this subject are needed. This also holds for the effect of gender where two studies indicate a clear gender effect on the relation between mood and lighting (Knez 1995; Knez and Kers 2000). The last aspect regarding room appearance that needs further study relates to perceptual constancy, described in Chap. 1 (Sect. 1.8.1). Because of perceptual constancy, vision may partly neglect the illumination (or brightness) and appraise the lightness (i.e. reflectance) of surfaces instead. It leads to the question under which conditions the reflectance of a room surface is more influential than the illuminance on that surface (Cuttle 2004; Boyce 2014). CIE offers for future studies advice with a publication titled “guidance towards best practice in psychophysical procedures used when measuring relative spatial brightness” (CIE 2014).

4.3

Directionality and Modelling

The directionality of lighting relates to the difference in intensity with which the light is incident on objects from different directions. It determines the appearance of threedimensional objects and faces of persons in space. A too low degree of directionality (and thus a high diffuseness) gives soft or no shadows, which results in a dull environment in which identification of objects and faces is difficult. Too strong directionality produces harsh shadows in which details of the object are hidden, making the object appear unattractive and difficult to identify. The modelling effect of lighting is good when it reveals the details and texture of objects and results in an aesthetically pleasing environment. The balance between the diffuse and directional components of a lighting installation determines the quality of directionality. The ratio of cylindrical to horizontal illuminance (Ecyl/Ehor) at a point has been investigated as a possible measure for modelling. The cylindrical illuminance is the mean vertical plane illuminance at a point averaged over all vertical orientations of the plane. Therefore, this ratio characterises the directionality in the vertical plane relative to the horizontal plane. The ratio of Ecyl/Ehor gives a reasonably good prediction of the modelling quality in the case of general lighting suspended from or mounted on the ceiling (Hewitt et al. 1965). Based on the perceived appearance of

4.3 Directionality and Modelling

109

the face of a statue, values of around 0.3–0.6 proved to provide good modelling. Faces of statues need slightly larger directionalities than natural faces for the same preference (Fischer 1970). Therefore the values to be applied in lighting practice should be somewhat lower. The ratio provides only reliable enough modelling information for lighting from the ceiling. Therefore, the concept of cylindrical to horizontal illuminance ratio is not a general fit concept.

4.3.1

Flow of Lighting

To mathematically study the effect of light on objects in space, Gershun (1936) introduced the expression “light field” and visualised it with “flow lines”. Lynes et al. (1966) and Cuttle et al. (1967) introduced a promising and practical concept, which they called “flow of lighting”. This concept allows for the calculation of the main direction and strength of the light at a point in space as a result of all light rays at that point. Although some national interior lighting recommendations mention the concept, it is not often used in the lighting design practice. The required calculations for complete lighting installations are difficult to do without a computer. Today, of course, it is easy to incorporate the required calculation procedures into lighting calculation software. So, today, there is no real barrier for lighting designers and specifiers to use the concept of flow of lighting. This is of importance because with modern LED lighting it is easier to provide specific directionalities to light than it was with the larger light-emitting surfaces of gas-discharge lamps. Cuttle showed that flow of lighting could be quantified by the so-called vector-toscalar ratio, Evector/Escalar (Cuttle et al. 1967; Cuttle 1971, 2008). It contains two measures, the illuminance vector and scalar illuminance. In mathematics the expression “scalar” stands for a number that has no directional aspect, this in contrast to “vector” that stands for a number that has both a value and direction. The illuminance vector characterises the direction and magnitude of the light flow at a point in the lit space. The final directional effect of it depends on the ambient or overall lighting at that same point. The higher the overall lighting at a point is, the lower the directional effect. The scalar illuminance (Escalar) at a point characterises the overall illumination at that point. It is the mean illuminance on the surface of an (infinitesimally small) sphere at that point (Fig. 4.14, right). It is also called mean spherical illuminance. The maximum difference in illuminance between opposite sides of an (infinitesimally small) flat disc at a point defines the magnitude of the illuminance vector (Evector) at that point (Fig. 4.14, left). The direction of the vector is defined by the orientation of the disc for which the maximum difference is obtained. It is the direction of the normal to the disc surface at that orientation, away from the side having the lowest illuminance (Fig. 4.14, left). With no directionality of lighting, the vector/scalar ratio is zero, and the maximum value is 4, which is the case in a situation with a single parallel light beam in a complete black interior. Appendix E gives the set of formulas for the calculation of Evector/Escalar.

110

4 Visual Satisfaction

Fig. 4.14 Right: The illumination vector at a point is defined by that (infinitesimally small) disc having the largest difference in illuminances on the opposing sides of the disc. The vector is directed away from the side having the lowest illuminance. Right: The scalar illuminance is the mean illuminance on the surface of an (infinitesimally small) sphere

Fig. 4.15 Simple matt white sphere on a stick to evaluate in practice the illuminance vector in a lit space

Cuttle (2008) proposed the use of a simple matt white sphere on a small stick to reveal the potential of lighting to produce shading patterns at different locations in a lit space. Such a sphere directly visualises the illuminance vector (Fig. 4.15). Cuttle (Cuttle et al. 1967; Cuttle 1971) did extensive experiments with a cubicle made up of tubular fluorescent lamps. It enabled the creation of a wide range of different vector/scalar ratios, both regarding direction and strength. Test persons, looking at the face of a person sitting in the middle of the cubicle, could adjust

4.3 Directionality and Modelling

111

Table 4.2 Vector/scalar ratio with appearance effect and application suggestions (Cuttle 2008) Evector/Escalar 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0

Appearance

Application

Dramatic Very strong Strong Moderately strong Moderately weak Weak Very weak

Strong contrasts, details in shadows not discernible Suitable for displays; too harsh for human features Pleasant appearance for distant faces (formal) Pleasant appearance for near faces (informal) Soft lighting for subdued effects Flat shadow-free lighting

different directional lighting situations to provide “just too soft”, “preferred” or “just too harsh” perceived modelling. In this way, he found that values for the vector/scalar ratio in the range of 1.2–1.8 are preferred for office types of environments. For other types of lighting application, different modelling may be preferred. In display lighting, for example, much larger values may be wanted. Table 4.2 gives vector/scalar ratio values with corresponding appearances and suggested applications as proposed by Cuttle. Fischer (Fischer 1970; de Boer and Fischer 1981) extended the study of modelling of the human face to inanimate objects with an experimental set-up similar to that used by Cuttle. The objects were flowers, a tennis ball and an unlighted incandescent lamp. His results for preferred values of the vector/scalar ratio for faces were in agreement with the 1.2–1.8 range of Cuttle. For objects, he found that a somewhat stronger directional effect, with vector/scalar ratios in the range of approximately 1.5–2.5, is needed to arrive at a preferred modelling situation. Morgenstern et al. (2014) found that the directionality of lighting that typically occurs in natural scenes is also preferred for human features. Natural scene directionalities coincide with the vector/ scalar ratio range of 1.2–1.8 (Morgenstern et al. 2014; Xia et al. 2017a). Protzman and Houser (2005) did an innovative exercise to relate vector/scalar ratios to the quality of modelling. They used published data from experiments carried out in the 1970s by Flynn et al. (1973, 1975) and Flynn (1977). Flynn used a small conference room for experiments which resulted in subjective impressions for six different lighting settings in that room. Protzman and Houser used a lighting software package to create a computer model of Flynn’s conference room with the six lighting conditions tested by Flynn. The room size and shape, surface characteristics, luminaire selections and layouts, and luminance distribution were based on photographs of room and lighting situations at the time. With the computer model, it was now possible to make vector/scalar ratio calculations at the grid points of horizontal calculation planes. These calculated values were subsequently related to the original, 1970s, assessments. This led to interesting findings. Spaces with a variation in the vector/scalar ratio are preferred to spaces with little variation under the condition that the vector/scalar ratios at individual grid points have values between 1.2 and 1.8, the range suggested by Cuttle. A combination of an average vector/scalar ratio lower than 1.2 coupled with little individual variation results in perceptions of spaciousness. This exercise was exploratory, but it shows a possible

112

4 Visual Satisfaction

way of how the vector/scalar ratio can be applied in lighting practice. Ngai (2016) gave an example of the practical use of vector/scalar ratio by using the vector/scalar ratio in a study to evaluate task-ambient lighting systems that save energy and enhance visual appearance at the same time. For the measurement of the illumination vector and scalar illuminance so-called cubic illumination meters have been developed. They are based on the measuring of illuminances on six faces of a cube (Cuttle 1997, 2008, 2014; Xia 2017b). Xia (2017b) provides guidance for the easy construction of a cubic illumination meter with commercially available components.

4.3.2

Light Tubes

The flow of lighting can be visualised by so-called light tubes (Muryy 2009; Mury et al. 2009). Light tubes are a further development of the earlier mentioned flow lines of Gershun (1936). The direction of a light tube is in each position parallel to the light vector while the opening width of the tube is inversely related to the magnitude of that vector. This magnitude is equal to the average light flux “flowing” through the tube at that position. It means that at locations where the tube is small, and the magnitude of the vector is large, the net light transport is large (high flux density). Near to light sources, from where the tubes originate, they are small (large flux transport). The tubes end up on the room surfaces, where the net light transport has become small, and the tubes consequently are wide. Light tubes are not the same as light rays and can, therefore, contrary to light rays, be curved. Light tubes are based on the combined effect of all light incident at a point from all directions (from both direct and indirect contributions). Figure 4.16 illustrates the difference between light rays and light tubes. Fig. 4.16 Difference between light rays and light flow. Small arrows represent light rays and the broad arrow the illumination vector. The diameter of the light tubes is inversely proportional to the magnitude of the vector. Top: Large illumination vector and corresponding small light tube; bottom: smaller illumination vector and corresponding wider light tube. Note: Discs on the left are in reality infinitesimally small

4.3 Directionality and Modelling

113

With today’s available computer graphic software, methods for light tube visualisations of complete lighting installations have been developed (Mury et al. 2009; Pont 2013; Huang and Sanderson 2014; Kartashova et al. 2016; Xia et al. 2017a, b). Xia et al. (2017b) introduced light tube visualisations that not only show information about direction and light flux density but also about diffuseness. Diffuseness here is defined entirely based on a mathematical description of the physical light distribution. It represents the inverse of the vector/scalar ratio. Their light tube visualisations use colour saturation to indicate diffuseness and colour brightness to visualise light density. Figure 4.17 gives an example of such visualisation. This type of visualisations allows a detailed analysis of the spatial and formgiving potential of lighting designs. It combines two early concepts, “lighting field”

Legend

Normalised light density

Fig. 4.17 Visualisation of light tubes for a room with three diffuse light sources mounted in a ceiling along one of the long walls. A tube’s direction is parallel to the light vector, and its width is inversely proportional to the magnitude of that vector. Low colour saturation of the tube’s drawing indicates a high level of diffuseness. Low brightness of the drawing means low light density. Picture with permission from Xia et al. (2017b)

1.0 0.8 0.6 0.4 0.2 0 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Normalised diffuseness

114

4 Visual Satisfaction

Fig. 4.18 The appearance of a white matte sphere for the conditions given in the legend of Fig. 4.17. Oblique light incidence from above and from the front (altitude angle 70
and azimuth angle 20
). Picture with permission from Xia et al. (2017b)

Normalised light density 1.0 0.8 0.6 0.4 0.2 0 0.0

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Normalised diffuseness 4.0

Evector/Escalar

0

(Gershun 1936) and “flow of lighting” (Lynes et al. 1966; Cuttle et al. 1967), into a modern tool for the professional lighting designer. That designer must get a good feeling of what different combinations of diffuseness (or the reversed metric directionality) and density visually mean. The computer rendering shown in Fig. 4.18 of how diffuse white spheres appear under different combinations of diffuseness and flux density helps develop this feeling.

4.4

Discomfort Glare

As has been discussed already in Chap. 3, glare can take either of the two forms: disability glare and discomfort glare. Disability glare is the form that is responsible for the negative influence of glare on visual performance. In Chap. 3 it has been shown that disability glare has a neglectable effect on visual performance under most interior lighting conditions. Glare that causes a feeling of discomfort is referred to as discomfort glare. It can cause problems in interior lighting situations. The discomfort effect of glare may lead to feelings of irritation and fatigue. The eyes may contract. In extreme circumstances, discomfort glare may result in painful feelings and after prolonged exposure in headaches. The adverse effects of discomfort glare vary to a great extent in type and seriousness between persons.

4.4.1

Fundamental Approach

The physiological and psychological mechanism underlying discomfort glare sensations is not yet clearly understood. Different theories have been proposed. An

4.4 Discomfort Glare

115

influencing factor is certainly the pupillary contraction (Fry and King 1975; Howarth et al. 1993). The activity of facial muscles surrounding the eyes as a reaction to discomfort glare has also been studied (Berman et al. 1994; Murray et al. 2002). Stone (2009) suggested a pain control mechanism in the brain, protecting the retina from bright light. Wilkins (Wilkins et al. 1984; Wilkens 2016) proposes as a possible cause for discomfort the retinal image differing too much from natural scenes, particularly in respect of spatial frequencies of the brightness pattern of the scene (for spatial brightness frequencies see also Sect. 3.2.1). Bargary et al. (2015) studied functional magnetic resonance imaging (fMRI) scans of the brain produced in the presence of discomfort glare. Their results suggest that the physiological properties of receptive fields and their ON-OFF centre-surround ganglion cells may play a fundamental role in evoking discomfort glare sensations. Receptive fields have been dealt with in Sect. 1.5 of Chap. 1. Donners et al. (2015), Scheir et al. (2018, 2019) and Safdar et al. (2018) use as a basis for their discomfort glare considerations the neural response to bright light and the mechanism of receptive fields. These studies demonstrate that models built upon centre-surround receptive fields and their ganglion cells that transform luminance signals into edge detection signals are promising candidates for quantifying discomfort glare from a physiological point of view. This can be understood by considering the large circular object shown in Fig. 4.19, left, as a large-sized luminaire of uniform luminance. This picture has been explained already in detail in Chap. 1 (Sect. 1.5). As Fig. 4.19, left, shows, only the ganglion cells that overlap with the edges of the luminaires result in output activity of the ganglion cells. A matrix LED luminaire with a multitude of bright LEDs has more edges and consequently excites more ganglion cells (Fig. 4.19, right). Therefore, more information is sent through the optic nerve towards the visual cortex in the brain. This engages the brain more, which, in turn, may be the reason for a higher degree of discomfort glare of LED-matrix luminaires. This latter aspect will also be discussed in Sect. 4.4.2.3. Still, some work has to be done before this fundamental approach may lead to discomfort glare models suitable for use in lighting design. Until a generally accepted fundamental approach, based on the mechanism underlying discomfort glare sensations, results in a workable method to quantify discomfort glare, empirical studies form the background for the metrics used to quantify discomfort glare.

4.4.2

Unified Glare Rating, UGR

Discomfort glare has long been studied empirically by measuring the subjective assessment of different glare situations in which the physical parameters influencing the glare sensation were varied. These physical parameters are (see Fig. 4.20) as follows: • The luminance of the glare source in the direction of the observer (Ls in cd/m2): the larger the source luminance the larger discomfort glare sensations.

116

4 Visual Satisfaction + +

+

+ +

+

+

+

+

+

+

+

+

+

+

+

+

+

+ +

+

+ +

+

+

+

+

+ +

+

+

+ +

+

+

increased firing rate

+

+

+

+

+

+

+

+

+

+

+

+

activity suppressed

+

+

+

+

+ +

+

+

+

+

+ +

+ +

+

+

+ +

+

+

+ +

+

+

+

+ +

+ +

+ +

+ +

+

+

+

+

+

+

+ +

+ +

+ +

+

+

+

+ +

+ +

+ +

+

+ +

+

+

+ +

+

+

+

+ +

+ +

+

+ +

+

+

+

+

+ +

+ +

+ +

+ +

+ +

+

+

+ +

+

+

+

+

+

+

+

+ +

+

+

+ +

+ +

+

+

+ +

+ +

+

+

+

+

+

+ +

+ +

+

+ +

+

+ +

+

+ +

+

+ +

+ +

+

+ +

+

+

+ +

+ + +

+

+

+

+ + +

+ +

+

+

+

+

+ +

+

+ +

+

+ +

+ +

+ +

+

+ +

+

+

+ +

+

+ + +

+

+ +

+

+ +

+

+

+

+

+ +

+ +

+

+

+

+

+ +

+

+

+ +

+

+ +

+

+ +

+

+

+

+ +

+

+ +

+

+ +

+

+

+

+ +

+

+

+

+

+

+ +

+

+ +

+

+

+

+

+ +

+ +

+

+ +

+

+

+ +

+

+ +

+

+

+ +

+

+

+

+

+

+

+ +

+

+

+ +

+

+

+ +

+

+

+ +

+

+

+

+ +

+

+

+

+

+

+

+

+

+

+

+

+

+

+ +

+ +

+

+

+ +

+ +

+

+ +

+ +

+ +

+

+ +

+

+ +

+

+

+

+

+

+

+

+

+ +

+

+ +

+

+ +

+

+

+ +

+

+

+

+

+

+

+

+

+ +

+ +

+

+

+

+

+ +

+

+ +

+

+

+

+ +

+

+

+

+ + +

+

+

+

+

+

+

+

+

+ +

+ +

+ +

+

+ +

+

+ +

+ +

+

+ +

+

+

+

+

+

+ +

+ +

+

+ +

+

+

+

+

+

+ +

+

+ +

+

+

+

+ +

+

+

+

+

+

+

+

+ +

+ +

+ +

+

+

+

+ +

+

+ +

+

+ +

+ +

+

+

+

+

+

+

spontaneous activity only

Fig. 4.19 Edge detection of a light source. The light source with a uniform luminance (left) excites fewer ganglion cells than the source with non-uniform luminance (right) Fig. 4.20 Physical parameters influencing discomfort glare: glare source luminance (Ls), solid angle (ω), position ( p), background luminance (Lb)

p h’ Ls x

Lb

ω y

viewing direction

• The solid angle subtended by the glare source at the observer’s eye (ω in steradian): the larger the solid angle, the larger the part of the visual field that is occupied by the glare source and the larger discomfort glare sensations. • The background luminance that controls the adaptation level of the observer (Lb in cd/m2): the larger the background luminance, the larger the adaptation state and consequently discomfort glare sensations are less; since the room surface reflectances partly determine the background luminance, these reflectances play a role as well: the darker the walls and ceiling are, the larger discomfort glare sensations are. • The displacement of the glare source from the line of sight: the further away the glare source is positioned from the line of sight, the smaller discomfort glare sensations are. The basis of discomfort glare studies has been laid by Luckiesh and Guth (1949) in the United States and Hopkinson (1957) in the UK, followed by Söllner (1965) in

4.4 Discomfort Glare

117

Germany. Their studies, as well as later assessment studies, have led to different formulae that try to predict the discomfort glare sensation with the aid of a glare rating. All these formulae are of the following type: G¼

Lbc

Lsa  ωb  f ðdisplÞ

with Ls: the light source luminance, Lb: the background luminance and a, b and c weighting exponents (different in the different studies) f(displ) a function of the displacement of the glare source from the line of sight

4.4.2.1

UGR for a Specific Location and Viewing Direction

In 1995, CIE combined the best parts of the different glare evaluation methods in use at that time into one formula. Because of this “consensus method”, the result is called “CIE Unified Glare Rating (UGR)”:   0:25 X L2s  ω UGR ¼ 8  log  Lb p2 with p: position index, dependent on the displacement of the glare source away from the line of sight, obtained from Luckiesh and Guth (1949) log: the logarithm to the base 10 Ls: the light source luminance Lb: the background luminance, the value that can be calculated from the vertical indirect illuminance at the observer’s eye with Lb ¼ Eind/π The UGR rating scale itself is the scale that was used with good experience in the British system for nearly 30 years. A high value indicates significant discomfort glare, while a low value indicates little discomfort glare. Lighting resulting in UGR values lower than 10 corresponds to imperceptible glare and UGR values of 30 to very uncomfortable glare. The difference of one number of the scale represents a just perceptible difference and a difference of three a noticeable difference. As discomfort glare for very small light sources is determined by intensity rather than luminance, the UGR system should not be used for sources with a solid angle smaller than 0.0003 sr (CIE 1995). An intensity-based formula is given in CIE (2002) for light sources smaller than 0.0003 sr. For large light sources with a solid angle larger than 0.1 sr (for example 1 m2 seen from 3 m) the system has not been evaluated and should therefore not be used. An essential property of the CIE Unified Glare Rating formula is that it fulfils the requirement of additivity of multiple glare sources; that is, it allows for simple addition of the ratings of different glare sources to arrive at the overall rating of a

118

4 Visual Satisfaction

complete lighting installation. It also allows for subdivision of luminaires; that is, two subdivided halves of a luminaire, each with the same luminance as the complete luminaire, indeed gives the same UGR rating as the complete luminaire alone. To give insight into the effect of the different components of the formula on discomfort glare, Fig. 4.21 shows, for a single glare source, positioned vertically 10
above the line of sight, the dependency of UGR on solid angle of the glare source, glare source luminance and background luminance. Figure 4.22 gives insight into the effect of displacing the glare source away from the line of sight in both the horizontal and vertical plane, through the influence of position index p on decreasing the value of UGR. The eyebrows and forehead of observers shield some glare source positions at large vertical angles. The corresponding position index values, p, are, therefore, according to CIE (1995), void as indicated in Fig. 4.22. The table with the position index values, p, for different displacements, as needed for the calculation of UGR, is given in Appendix F. The UGR method makes it possible to predict the discomfort glare rating for each observer position and each line of sight for all types of lighting installations. These installations may even consist of different types of luminaires mounted at different heights. In this way, discomfort glare can be evaluated for all relevant situations. It should be stressed that UGR is a lighting installation metric. As the background luminance plays a role in it, it is impossible to characterise the discomfort glare aspect of a specific luminaire without specifying the details of the space in which it is used (room dimensions and reflectances).

Fig. 4.21 The relationship between UGR, solid angle under which the glare source appears (ω), glare source luminance (Ls) and background luminance (Lb). The solid angle scale is also given in terms of the distance from which a glare source with dimensions 0.5  0.5 m is seen

UGR 30 27 24

Lb (cd/m2) Ls

10

.0

00

.0

00

10

21

10

18

.00 10

15 12

00 10

6

10

0 0.0003

0,5 x 0.5 m seen from:

10

0

0

10

9

3

30

10

00

30

0

00

0.001 16m

30

0.003

0.01 5m

0.03

ω

0.1

(steradian) 1.6m

4.4 Discomfort Glare Fig. 4.22 Illustration of the position index p dependency of the displacement of a glare source away from the line of sight. Position index p value shown in bold. Vertical displacement: h0 /y; horizontal displacement x/y (see the coordinating system in Fig. 4.20). As the reciprocal of the square of p is in the UGR formula, the larger the value of p the lower the UGR and thus the less discomfort glare

119

α -3.0

0.0

-1.5

α

1.5

3.0

x/y

0.0

0°

1.0

4.2

0.4

22°

2.3

6.1

1.7

60°

16

14

1.8 1.9

h’/y

4.4.2.2

α=71°

Unified Glare Rating for a Lighting Installation, UGRL

To judge the overall quality of the lighting of a room with respect to discomfort glare, and for specifying glare in lighting standards, there is a need for one single UGR value that characterises the overall glare situation in a lighted room. Such single UGR value is obtained by setting reference conditions for the observer location and viewing direction and defining a standard calculation procedure. The single value metric is referred to as “Unified Glare Rating for a Lighting Installation, UGRL”. The subscript “L” stands for “lighting installation”. The reference conditions and the standard calculation procedure as defined by CIE for the production of UGR tables are used (CIE 1995, 2010). Chapter 13 deals with UGR tables. The reference conditions and calculation procedure used for these UGR tables as well as for the determination of UGRL are discussed below. They are as follows: • Reference conditions: UGR is calculated for observers located at the midpoint of each wall with a horizontal line of view towards the centre of the opposite wall (eye height 1.2 m). For symmetrical installations, only two observers need to be taken into account (Fig. 4.23). The worst UGR value, viz. the highest value, of these observers is the typical one for the entire lighting installation of the room: UGRL. • Standard calculation procedure: The luminaires are not arranged according to the actual situation but distributed uniformly over a rectangular space, with the first rows of luminaires half a spacing from the room borders. • Standard calculation procedure: The spacing-to-height ratio of the luminaires is small, either 1:1 or, better, 1:0.25. Here the luminaire height is taken as the distance above eye level.

120

4 Visual Satisfaction

Fig. 4.23 Reference observer positions for the calculation of one singe UGR value representative for a whole room. Left: an example of an actual luminaire arrangement; right: the standardised arrangement for the calculation of the single UGR value

Actual luminaire arrangement

Spacing to height ratio =0.25

That for the calculation of UGRL not the actual (planned) luminaire positions are used but the uniformly distributed positions (with small spacings) may come as a surprise. Figure 4.23 clarifies the reason. The left part of the drawing shows an example of a planned luminaire arrangement, while the right shows the uniformly distributed arrangement of the luminaires with a spacing-to-height ratio of 0.25. In a lighted room, the worst UGR value usually occurs for an observer located near to a luminaire row. However, the reference observers on the left are relatively far away from these positions and therefore do not give a representative UGR value for the room as a whole. On the right part of the drawing, the reference observers are much closer to where the worst glare condition occurs, thanks to the more densely packed luminaires. In this latter way, the more representative UGR value is obtained, which is also representative of the room on the left with the actual luminaire positions. That in the standardised situation more luminaires are present than in the actual situation has hardly any effect on the calculated UGR value. The additional luminaires indeed increase the number of summations of the summation part of the glare formula, but they also increase the background luminance. These two effects cancel each other. The smaller the spacing-to-height ratio the more representative the calculated UGRL value is for the overall glare situation in a room. For this reason, it is unfortunate that where in the original CIE publication (1995) a spacing-to-height ratio of 0.25 was recommended it was later changed to 1:1. Hopefully, future recommendations will change it back to 1:0.25 (see also Sect. 13.1.4 of Chap. 13). The fact that the standard calculation procedure does not use the luminaires as located in the actual situation but distributed uniformly over a rectangular space limits, strictly speaking, the application of UGRL to rectangular rooms using one type of luminaire mounted at the same height. However, the following procedures give for most situations accurate enough results. For non-rectangular spaces, the actual space is, for the calculation of UGRL, changed into the best-fitting rectangular space. For more than one luminaire type or more than one mounting height, UGRL is calculated for each luminaire type or each mounting height individually. The worst value of UGRL is then taken as the decisive value.

4.4 Discomfort Glare

4.4.2.3

121

Non-uniform Glare Sources

Today the UGR method is the most widely specified method for the restriction of discomfort glare in national and international standards for interior lighting. The practical experience in applying the method has been mainly positive with all types of gas-discharge lamps. This even though Clear (2013) concluded that most of these studies had some built-in weaknesses. With the introduction of LEDs in interior lighting, the situation has been considerably changed. The non-uniform luminance distribution of many LED luminaires makes the UGR system unreliable in predicting discomfort glare in quite some situations. The glare source luminance (Ls) of the glare source is obtained from the luminous intensity of the glare source (Is) in the direction of the observer, and the projected area into the direction of the observer (Ap) of the light-emitting part of the glare source: Ls ¼

Is Ap

Ls is thus equal to the average luminance of the glare source in the direction of the observer. With a mainly uniform luminance distribution across the light-emitting area, as is the case with most gas-discharge lamp luminaires, this works well. Many LED luminaires consisting of a number of individual LEDs, however, have a strong non-uniform luminance (Fig. 4.24). The examples shown all have the same average luminance but a widely different luminance distribution, with different peak luminances, different internal luminance contrasts and different gradients. Many studies done with LED luminaires have demonstrated that the average glare source luminance results in UGR values that no longer relate well enough with glare assessments (Waters et al. 1995; Kim and Koga 2004; Lee et al. 2007; Takahashi et al. 2007; Eble-Hankin 2008; Xia et al. 2011; Higashi et al. 2013; Cai and Chung 2013; Geerdinck et al. 2014; Tashiro et al. 2015; Scheir et al. 2015b; Yang et al. 2017a, b, 2018a, b). Figure 4.25 illustrates this from the results of a discomfort glare assessment study of Yang et al. (2017b). In an officelike room, 20 observers assessed discomfort glare on a 7-point scale. For each assessment, one of 15 different LED panel types of luminaire was used as a glare source. These panels did fall into four groups, according to their structure: • DLED panels with a diffuser placed in front of the whole matrix panel (nearly uniform luminance distribution) • HLED panels with individual diffusers in front of the raw LEDs (slightly less uniform luminance distribution) • RLED panels with raw LEDs (pronounced non-uniform luminance distribution) • PLED panels also with raw LEDs but with two different light outputs of the LEDs on the panel (most non-uniform luminance distribution) Figure 4.25 shows that for a same calculated UGR, the discomfort glare assessment varies widely between the non-uniform and uniform panels. For example, for a

122 Fig. 4.24 Uniform and non-uniform luminance distributions of different LED luminaires, all with same average luminance value. Left: diffuser in front of the whole matrix panel; middle and right: raw LED matrix panel

4 Visual Satisfaction

Ls

av

av

Fig. 4.25 Relationship between assessment of discomfort glare and calculated UGR value for four groups of 15 LED matrix panels, in order of uniformity of luminance: DLED, HELD, RLED and PLED. Panels at, respectively, 10
and 20
above the line of sight. Background luminance 1.6 and 14.3 cd/m2. Solid angle around 0.004 sr. Four different average luminances were assessed for each LED panel (redrawn from Yang et al. 2017b)

av

Assessment (discomfort) 7 6 5 4 3 DLED HLED RLED PLED

2 1 -20

-10

0

10

20

30

UGR

40

same calculated value of UGR of 10, the non-uniform R and P LED panels are assessed on the 7-point scale nearly 2 points worse than the more uniform D and H panels. In other words: UGR, as it is defined today, is not good enough for LED luminaires. Section 4.5.1 mentioned that an entirely new model based on a fundamental physiological approach (receptive field mechanism and pupillary reflex) to discomfort glare will probably not be available soon. Fortunately, quite some of the studies mentioned above, that indicated that non-uniformity is a problem, also proposed different ways to correct the UGR formula to make it also valid for non-uniform luminance glare sources. Some of them propose a different way to determine a more relevant value of the glare source luminance based on the actual luminance distribution: the effective glare source luminance (Hara and Hasegawa 2012; Tashiro et al. 2015; Yang et al. 2017b, 2018b). As an illustration of this

4.4 Discomfort Glare

123

approach, we discuss the Yang et al. study in some detail. They propose a solution based on the internal contrast of the luminance distribution of the glare source, expressed in terms of the standard deviation of the luminance variation over the light-emitting area of the glare source: Contrast ¼

Std Lav

A new glare source luminance is introduced, LUNI, that is calculated from the average glare source luminance (Lav) and the contrast value according to the expression LUNI ¼ ðContrast þ 1Þa=16  Lav with a ¼ scaling factor, equal to the value 11, based on the data set of the authors. The modified glare rating formula taking non-uniformity of glare sources into account then becomes  UGRUNI ¼ 8  log

0:25 X L2UNI  ω  Lb p2



For complete uniform glare sources UGRUNI equals UGR. Figure 4.26 shows the assessments of Fig. 4.25 now related to values of UGRUNI. The correlation between the assessments and UGRUNI is good. The same data set of assessments used with a different modified UGR model, developed by various researchers using a different method to take non-uniformity into account (Tashiro et al. 2015), gives a similar good correlation. Quite some researchers propose, instead of adapting the method of determining the glare source luminance, an adaption of the method of determining the lightemitting area of the glare source: the effective luminous area (Funke and Schierz 2015, 2016; Loe et al. 2015; Scheir et al. 2015a, b, 2017). A high-resolution luminance image of the luminaire forms the basis of the determination of the effective luminous area of the luminaire to be used in the standard UGR formula. Such an image is obtained with an imaging luminance-measuring device (ILMD), which are described in Sect. 18.2.4.1 and 18.2.4.2. The image is converted into a luminance map consisting of many pixels of different luminance values. Scheir et al. use a luminance threshold to distinguish pixels with low luminance values from pixels with high luminance values. They tested as threshold values 500 and 750 cd/ m2. Only pixels with a luminance value above the threshold are taken into account for the determination of the actual luminous area. Figure 4.27 shows for the matrix LED luminaire shown at the top, on the lower-left the total light-emitting surface (as used today in UGR calculations), in the lower-middle the actual luminous area for the 500 cd/m2 threshold and on the lower-right the area for the 750 cd/m2 threshold.

124 Fig. 4.26 Relationship between assessment of discomfort glare and calculated value of UGRUNI for the same LED matrix panels as in Fig. 4.25. Coefficient of determination: 0.95 (redrawn from Yang et al. 2018b)

4 Visual Satisfaction

Assessment (discomfort) 7 6 5 4 3 DLED HLED RLED PLED

2 1 -20

-10

0

10

20

30

40

UGRUNI Fig. 4.27 Matrix LED luminaire (top) with bottomleft: an image of the standard (total) luminous area; middle: pixels with a luminance value larger than 500 cd/m2; right: pixels with a luminance larger than 750 cd/m2. Source: Scheir et al. (2017)

standard area

Lthres > 500 cd/m2

Lthres > 750 cd/m2

For the actual calculation of the light-emitting area, the areas of the remaining pixels are summed to give the effective area for use in the standard glare formula. The resulting UGR values correlate well, especially for the threshold of 750 cd/m2, with the subjective appraisals of 16 naïve test persons (coefficient of determination between 0.78 and 0.88 depending on the task carried out). By considering groups of connected pixels as separate light sources in the luminaire (16 in the example of Fig. 4.27), the correlation improves only slightly (coefficient of determination between 0.80 and 0.91). It, therefore, does not outweigh the extra effort. The elegance of this method is that nothing changes for the lighting designer and the lighting specifier. Lighting standards and calculation software need not be adapted. The lighting designer can with the unchanged standard UGR formula predict discomfort glare for all types of luminaires including matrix LED luminaires by using instead of the total light-emitting area the effective light-emitting area. Luminaire manufacturers must supply as part of their photometric data, instead of the total

4.4 Discomfort Glare

125

light-emitting area, the effective light-emitting area (determined in the manufacturer’s laboratory by a luminance map of the luminaire). The CIE Technical Committee “Discomfort caused by glare from luminaires with a non-uniform source luminance” evaluated several UGR correction methods. Just before the publication of this book, the Committee concluded that the preferred correction method for UGR for non-uniform glare sources is the use of “effective light emitting area” in the standard (unchanged) UGR formula (CIE 2019). The evaluation of the committee showed that this preferred method gives the best agreement with the experimental data. The effective light emitting area is determined from a high-resolution luminance image of the luminaire as described earlier in this Section. If luminaire manufacturers determine and provide the effective light emitting area for their luminaires, UGR can be calculated using the familiar and unchanged lighting application software packages. The adaptation of the UGR method according to this preferred method must be considered as a temporary solution until a fundamental approach based on physiological and psychological mechanisms, for example, as discussed in Sect. 4.4.1, provides practical results.

4.4.2.4

Influence of Spectrum

The discomfort glare sensation is influenced by the spectrum of the light source used. Recent research shows that the photopic eye sensitivity does not solely determine the spectral effect in discomfort glare. Especially also the short-wavelength type of S-cone cells (“blue sensitive” cells) play a role (Bullough 2009; Fekete et al. 2010; Bodrogi et al. 2012; Akashi et al. 2013; Niedling et al. 2013). It explains why light sources with a relatively large amount of short wavelengths (bluish light) result in higher discomfort glare effects than do longer wavelength spectra (yellowish and reddish light). The few studies explicitly done under interior lighting conditions indicate that by varying the CCT between 3000 and 7000 K, the discomfort glare assessment varies with roughly 0.5 points on a 7-point assessment scale (Zhang et al. 2013; Huang et al. 2018). For different-coloured, non-white, LEDs the variation is much larger, up to 2.5 points (Yang et al. 2018a). These studies thus indicate that usually cooler tinted light sources result in a higher degree of discomfort glare than warmer tinted light sources. For a more accurate prediction of the effect, the complete spectrum of the light source and not only its correlated colour temperature should be taken into account. For the UGRUNI model that predicts UGR for both uniform and non-uniform glare sources (discussed in the previous section), an extension has been developed that also predicts discomfort glare from LEDs having different spectra (Yang et al. 2018a, b; Huang et al. 2018, 2019). This modified model is designed to be suitable for white-light LED spectra and coloured non-white LED spectra. This model, taking the complete spectrum into account, is based on only two glare source positions in the vertical observation plane (0
and 20
) and thus needs further validation for more glare source positions in both the vertical and horizontal plane and for additivity of more light sources.

126

4.4.3

4 Visual Satisfaction

Overhead Glare

In Fig. 4.22 it has been shown that CIE (1995) does not specify position index values for angles larger than 60
above the horizontal line of sight in the vertical observation plane and above 71
to the left and right. This is done because glare sources are shielded for these conditions by the eyebrows and forehead of observers and are thus outside the field-of-view area. However, high-luminance luminaires in this area (above or nearly above the observer) nevertheless can evoke an uncomfortable sensation (Fig. 4.28). This is particularly the case when the head moves back and forth under such high-luminance luminaires. Light may be reflected from the nose and cheeks and scattered from the eyebrows and glasses into the eyes resulting in discomfort glare sensations. This aspect of glare is referred to as “overhead glare”. Assessment tests with a single luminaire, with different uniform luminances, positioned at angles between 55
and 75
in the vertical plane of observation were carried out to study overhead glare (Ngai and Boyce 2000; Boyce et al. 2003). When using for these angles, position index values from the original source (Luckiesh and Guth 1949), there exists a good correlation between UGR and overhead glare assessments. This was confirmed for a larger range of angles (55–90
) for a non-uniform LED matrix luminaire in a similar type of study (Xia et al. 2011). In this study, position index values for the larger angles were obtained by extrapolation of the original position index values. A better correlation (coefficient of determination of 0.67) was obtained when using a different set of position index values (Kim et al. 2009). The earlier studies with uniform-luminance luminaires indicated that overhead glare occurs with luminaires at luminances above 16,500 cd/m2. The non-uniform LED matrix luminaires result in uncomfortable overhead glare at average luminances of the LED panel above 8300 cd/m2. These studies seem to indicate that overhead glare can be predicted with the aid of the UGR concept by extended position index values. Further studies are needed to confirm this.

Fig. 4.28 Luminaire positions that may give rise to overhead glare if luminances in the directions towards the observer are too high

overhead glare

± 60°

glare taken into account by UGR

4.5 Light Colour Preference

4.4.4

127

Indirect Glare

Light from bright-light sources reflected in glossy and semi-matt materials may also lead to discomfort sensations. The disturbance of visual performance because of such indirect glare can be severe. Therefore limitation of indirect glare is usually done to limit the adverse visual performance effects. Section 3.5 of Chap. 3 dealt with this subject.

4.5

Light Colour Preference

In Sect. 4.2.2 the influence of spectrum of light on brightness has been discussed in some detail. The colour, or better the tint of whiteness, of the light of a lighting installation itself may also have a direct effect on visual satisfaction. The predominantly prevailing climate influences the preference of light colour in interiors. In a warm climate, the preference is more towards “cooler” light sources of higher correlated colour temperatures (CCT), while in cooler climates preference is more for “warmer” light of lower CCTs. Sales analyses of the various colour types of fluorescent tubes per region do confirm these preferences. Cultural or ethnic differences also influence light colour preference (Park and Farr 2007a; Yakahashi et al. 2013; Yaodong et al. 2014). It is probably an effect of long-term habituation. In Japan, right from the introduction of fluorescent lamps in the 50s of the last century, fluorescent lamps with cooler white light than incandescent lamps were often used in living rooms of domestic interiors, a custom which hardly ever existed in Europe or the Americas. A previous section discussed a study that indicated that the chronotype of test persons influenced the appraisal of visual brightness (Kirsch 2014). It could well be that also the appraisal for the colour appearance of light is affected by chronotype. Persons of different chronotypes have different circadian rhythms. Therefore the circadian rhythm itself may play a role. The next chapter of this book will discuss circadian rhythms. Given this multitude of possible influences on light colour appraisal, it is not surprising that investigations into this subject give conflicting results. Kruithof (1941), one of the early prominent fluorescent lamp developers, was the first who did a study on the relationship between correlated colour temperature and appraisal of the pleasantness of lighting. At the moment of his study, artificial indoor lighting was incandescent lamp lighting. So, the long-term habituation was only with light of a CCT of around 2700 K (with a colour rendering index Ra of 100). Quite probably this has influenced the results of Kruithof’s study. Up to the present day, his study is often used as a reference. He published his findings in a graph, which became known as the “Kruithof curve”, sometimes referred to as “Kruithof law”. It is redrawn in many publications and lighting handbooks, unfortunately sometimes with additions or alterations that do not originate from the original Kruithof publication. Figure 4.29 is an exact, redrawn, copy of the graph.

128 Fig. 4.29 Kruithof curve, giving the relationship between correlated colour temperature (CCT) and illuminance level (E) for different pleasantness appraisals. Up to 2850 K, dimmed incandescent lamps, and, from 2850 K onwards, a varying combination of daylight and fluorescent lamps was used (Kruithof 1941)

4 Visual Satisfaction dimmed incandescent

E (lux)

daylight and fluorescent

5 2

10000 5

unpleasant unnatural

2

1000

pleasing

5 2

cold

100 5

dim

2

10 5

1750

2000 2250 2500

3000

4000 5000

10000

∞

CCT (K)

The “pleasantness appraisal labels” shown were not on Kruithof’s original graph, but are taken, exactly, from the written text of his publication. The appraisal “dim” corresponds to low CCT values in the lower shaded area, and “cold” to higher CCT values of the same area. Kruithof used widely different light sources for the CCT range below (dimmed incandescent lamps), and above 2850 K (daylight and fluorescent tubes). The conclusion of Kruithof’s graph is straightforward: from a pleasantness point of view, at low illuminances, low CCT light sources are preferred and at high illuminances high CCT light sources. Kruithof’s publication does not give more details of the test room that he used other than the wording “laboratory room”. It also does not describe the experimental procedure nor does it give the number of test persons. From later publications of colleagues who worked at the same time in the same (Philips) company, some more details of his study have become available. According to De Boer and Fischer (1981), only two test persons participated in what they call a “pilot test”. Another colleague wrote in 1967: “It is rather strange that Kruithof’s curve has become something of a general law. Dr Kruithof himself is by no means convinced that his values can be used for interior design” (Bodmann 1967). All this means that Kruithof’s experiment allows no conclusions to be drawn. This even more so, because the curve is given without individual measuring points and the publication mentions that the upper curve has been extrapolated with an assumed asymptote at 5000 K. Almost all later studies on a possible relationship between light colour and preference give at least part of their results by comparing them with Kruithof’s results. It is for this reason that we included the Kruithof curve in this book. One aspect of Kruithof’s curve that is interesting, but seldom explicitly referred to in other publications, is the scaling of the horizontal axis. The steps of the axis are not in simple “degree Kelvin” but in so-called reciprocal mega-Kelvin (1/MK). These steps are meaningful because they are based on the sensibility for correlated colour temperature changes, as first recognised by Priest (1933). Figure 4.30

4.5 Light Colour Preference Fig. 4.30 Comparison of Kelvin (K) and reciprocal mega-Kelvin (1/MK) scale to express correlated colour temperature CCT

129

CCT (1/MK) 500

400

300 30 2750

2000

2500

3000

200

100

0

10.000

∞

33 6000

4000 5000

CCT (K) compares Kelvin and reciprocal mega-Kelvin scales. Now it is, for example, clear that a difference between 2750 and 3000 K (incandescent and halogen lamp light) is nearly as noticeable as the difference between 5000 and 6000 K. Reciprocal microKelvins are commonly used in specifying colour filters and are referred to as reciprocal micro-degree or “mired”. Based on a literature search, Fotios (2017) concludes that 29 studies were published in which Kruithof’ relationship was the subject of investigation. He evaluated the experimental design and quality of reporting of these studies against the CIE guidance towards best practice in psychophysical procedures (CIE 2014). On this basis, only nine studies passed the test of credibility, viz. Boyce and Cuttle (1990), Davis and Ginthner (1990), Park and Farr (2007b), Viénot et al. (2009), Ao-Thongthip et al. (2013), Han and Boyce (2003), Dikel et al. (2014), Wei et al. (2014) and Islam et al. (2015). They all concern judgements regarding brightness and pleasantness of an illuminated scene not related to specific objects in that scene. Fotios converted the different appraisal scales used in these studies into the same 7-point rating scale. He subsequently plotted the results of all nine studies in one graph. Figure 4.31 shows a slightly adapted version of that graph. The average rating for the appraisal pleasantness, Apl, is given as a function of correlated colour temperature, CCT. Each line in the graph gives the result as obtained for the same illuminance value. To keep the graph surveyable, we have grouped the illuminance values in two groups, one with values equal or larger than 500 lux (solid lines) and one with values smaller than 500 lux (dotted lines). The two dashed lines are from a study where installed lumen instead of illuminance was specified. The graph shows, for the CCT range of 2500–6500 K, a strong horizontal trend. It means that the correlated colour temperature has hardly any effect on the appraisal of pleasantness. Those lines that are slightly non-horizontal show no consistent direction. This holds for the range of illuminance levels investigated in these studies: 50 lux to slightly more than 1000 lux. With all the knowledge gathered since Kruithof carried out his test in 1941, the conclusion can only be that the Kruithof curves are not valid for general use. Cuttle (2017) gives a possible reason for the Kruithof curves not having general validity. He proposes that it follows from the fact that Kruithof described the positive effect of a certain illuminance–correlated colour temperature combination as appearing “pleasing”. Cuttle points to the fact that a situation may be appearing as pleasing at one moment but not at another moment because of temporal variations in mood, attitude and expectations of the users. He illustrates this with the

130 Fig. 4.31 Relationship between pleasantness appraisal and correlated colour temperature (CCT) for two groups of illuminance values. Each line gives the results obtained for the same illuminance value; two dashed lines: lumen instead of lux values were specified. Graph slightly adapted from Fotios (2017)

4 Visual Satisfaction

Assessment (pleasantness) 7 6

≥ 500 lux < 500 lux

5 4 3 2 1 2500

3500

4500

5500

6500

CCT (K)

example of a hotel reception area. In the morning guests appreciate a bright, lively and stimulating appearance with high lighting level and high colour temperature, while in the evening a subdued and relaxing appearance is preferred with a lower lighting level and lower colour temperature. It would be interesting to study if an approach that takes into account the variations in mood, attitude and expectations results in guidance as to what lighting level–colour temperature combination is preferred in different situations and at different moments. Chronotype of the test persons and circadian timing of the assessments should then also be taken into account as possible influencing parameters. An alternative metric or metrics for correlated colour temperature to characterise the spectrum should then be considered as well.

References Akashi Y, Asano S, Kakuta Y (2013) Visual mechanisms of discomfort glare sensation caused by LEDs. In: CIE Centennial Congress, CIE x038:2013:327–330 Ao-Thongthip S, Suriyothin P, Inkarojrit V (2013) The combined effect of gender and age on preferred illuminance and colour temperature in daily living activities. In: Proceedings Lux Pacifica, Bangkok, pp 441–445 Bargary G, Furlen M, Raynham PJ, Barbur JL, Smith AT (2015) Cortical hyperexcitability and sensitivity to discomfort glare. Neuropsychologia 69:194–200 Berman SM (2008) A new retinal photoreceptor should affect lighting practice. Lighting Res Technol 40:373–376 Berman SM, Jewett DL, Fein G, Saika G, Ashford F (1990) Photopic luminance does not always predict perceived room brightness. Lighting Res Technol 22:37–41 Berman SM, Bullimore MA, Jacobs RJ, Bailey IL, Gandhi N (1994) An objective measure of discomfort glare. J Illum Eng Soc 23(2):40–49

References

131

Berrutto V, Fontoynont M, Avouac-Bastie P (1997) Importance of wall luminance on user satisfaction: pilot study on 73 office workers. In: Proceedings of Lux Europe, 8th European lighting conference, Amsterdam, pp 82–101 Bodmann HW (1967) Quality of interior lighting based on luminance. Trans Illum Eng Soc (Lond) 3(1):22–40 Bodmann HW, La Toison M (1994) Predicted brightness—luminance phenomena. Lighting Res Technol 26:135–143 Bodrogi P, Wolf N, Khanh TQ (2012) Spectral sensitivity and additivity of discomfort glare under street and automotive lighting conditions. Light Eng 20(2):22–26 Boyce PR (1973) Age, illuminance, visual performance and preference. Lighting Res Technol 5:125–140 Boyce PR (2014) Human factors in lighting, 3rd edn. CRC Press, Boca Raton, FL Boyce PR, Akashi Y (2002) The impact of light spectrum on perception, performance and preference in photopic conditions. In: Proceedings of IES conference, paper 33, pp 467–478 Boyce PR, Cuttle C (1990) Effect of correlated colour temperature on the perception of interiors and colour discrimination. Lighting Res Technol 22(1):19–36 Boyce PR, Hunter CM, Inclan C (2003) Overhead glare and discomfort. J Illum Eng Soc 32 (1):73–88 Brandston HM (2010) Comment 3 in Cuttle C. Towards the third stage of the lighting profession. Lighting Res Technol 42:90–91 Bullough JD (2009) Spectral sensitivity for extrafoveal discomfort glare. J Mod Opt 13:1518–1522 Cai H, Chung T (2013) Evaluating discomfort glare from non-uniform electric light sources. Light Res Technol 45:267–294 Chraibi S, Crommentuijn L, Van Loenen E, Rosemann A (2017) Influence of wall luminance and uniformity on preferred task illuminance. Build Environ 117:24–35 CIE (1995) International Commission on Illumination CIE Publication 117:1995, Technical report, Discomfort glare in interior lighting, Vienna CIE (2002) International Commission on Illumination CIE Publication 147:2002 CIE collection on glare, Glare from small, large and complex sources, Vienna CIE (2010) International Commission on Illumination CIE Publication 190:2010, Technical report, Calculation and presentation of unified glare rating tables for indoor lighting luminaires, Vienna CIE (2011) International Commission on Illumination CIE Publication 200:2011, Technical report, CIE supplementary system for photometry, Vienna CIE (2014) International Commission on Illumination CIE Publication 212:2011, Technical report, Guidance towards best practice in psychophysical procedures used when measuring relative spatial brightness, Vienna CIE (2019) International Commission on Illumination CIE Publication 232:2019, Technical report, Discomfort caused by glare from luminaires with non-uniform source luminance, Vienna Clear RD (2013) Discomfort glare: what do we actually know? Lighting Res Technol 45:141–158 Coaton JR, Marsden AM (1997) Lamps and lighting. Arnold, London Cuttle C (1971) Lighting patterns and the flow of light. Lighting Res Technol 3:171–189 Cuttle C (1997) Cubic illumination. Lighting Res Technol 29:1–14 Cuttle C (2004) Brightness, lightness, and providing ‘a preconceived appearance to the interior’. Lighting Res Technol 36(3):201–216 Cuttle C (2008) Lighting by design, 2nd edn. Architectural Press, Oxford Cuttle C (2010) Towards the third stage of the lighting profession. Lighting Res Technol 42:73–93 Cuttle C (2013) A new direction for general lighting practice. Lighting Res Technol 45:22–39 Cuttle C (2014) Research note: a practical approach to cubic illuminance measurement. Lighting Res Technol 46:31–34 Cuttle C (2017) Review of a published article. LEUKOS 13:19–20 Cuttle C (2018) A fresh approach to interior lighting design: the design objective-direct flux procedure. Lighting Res Technol 50:1142–1163 Cuttle C. Valentine W, Lynes J, Burt W (1967) Beyond the working place. In: Proceedings CIE 16th Session, Washington, DC, pp 19–28

132

4 Visual Satisfaction

Davis RG, Ginthner DN (1990) Correlated color temperature, illuminance level and the Kruithof curve. J Illum Eng Soc 19(1):27–38 De Boer JB, Fischer D (1981) Interior lighting. Kluwer Technische Boeken, Deventer De Vries HJA, Heynderickx IEJ, De Kort YAW, De Ruyter B (2015) Wall illumination-beyond room appraisal. In: Proceedings CIE 28th Session, Manchester, pp 284–290 Dikel EE, Burns GJ, Veitch JA, Mancini S, Newsham GR (2014) Preferred chromaticity of colortunable LED lighting. LEUKOS 10(2):101–115 Donners MAH, Vissenberg MCJM, Geerdinck LM, Van Den Broek-Cools JHF, Buddemeijer-Lock A (2015) A psychophysical model of discomfort glare in both outdoor and indoor applications. In: Proceedings 27th CIE Session, Manchester, pp 1602–1611 Duff J (2015) On a new method for interior lighting design. Doctoral thesis, Dublin Institute of Technology, Dublin Duff J, Antonutto G, Torres S (2016a) On the calculation and measurement of mean room surface exitance. Lighting Res Technol 48:384–388 Duff J, Antonutto G, Torres S (2016b) https://www.dropbox.com/1/1C62txkbVpcW1AQ8HPfydu Duff J, Kelly K, Cuttle C (2017a) Spatial brightness, horizontal illuminance and mean room surface exitance in a lighting booth. Lighting Res Technol 49:5–15 Duff J, Kelly K, Cuttle C (2017b) Perceived adequacy of illumination, spatial brightness, horizontal illuminance and mean room surface exitance in a small office. Lighting Res Technol 49:133–146 Eble-Hankin M (2008) Subjective impression of discomfort glare from sources of non-uniform luminance. Doctoral dissertation, University of Nebraska, Lincoln Fechner GT (1860) Elemente der psychophysik. Breitkopf und Härtel, Leipzig Fekete J, Sik-Lanyi C, Schanda J (2010) Spectral discomfort glare sensitivity investigations. Ophthal Physiol Opt 30:182–187 Fischer D (1970) Optimale Beleuchtungsniveaus in Arbeitsraumen. Lichttechnik 22:61–103 Flynn JE (1977) A study of the subjective responses to low energy and nonuniform lighting systems. Lighting Des Appl 7(2):6–15 Flynn JE, Spencer TJ, Martyniuk O, Hendrick C (1973) Interim study of procedures for investigating the effect of light on impression and behaviour. J Illum Eng Soc 3(1):87–94 Flynn JE, Spencer TJ, Martyniuk O, Hendrick C (1975) The influence of spatial light on human judgement. In: Proceedings CIE 18th Session, London, pp 39–46 Fotios SA (2006) Chromatic adaptation and the relationship between lamp spectrum and brightness. Lighting Res Technol 38(1):3–17 Fotios S, Atli D, Cheal C, Houser K, Logadóttir Á (2012) Proceedings of the 3rd international conference on appearance, Edinburgh, UK, pp 169–171 Fotios S, Atli D, Cheal C, Houser K, Logadóttir Á (2015a) Lamp spectrum and spatial brightness at photopic levels: a basis for developing a metric. Lighting Res Technol 47:80–102 Fotios S, Atli D, Cheal C, Hara N (2015b) Lamp spectrum and spatial brightness at photopic levels: investigating prediction using S/P ratio and gamut area. Lighting Res Technol 47:595–612 Fotios SA (2017) A revised Kruithof graph based on empirical data. Leukos 13(1):3–17 Fry GA, King VM (1975) The pupillary response and discomfort glare. J Illum Eng Soc 4:307–324 Funke C, Schierz CH (2015) Extension of the unified glare rating formula for non-uniform LED luminaires. In: CIE Publication CIE 216:2015 Proceedings of the 28th CIE Session, Manchester, pp 1471–1480 Funke C, Schierz CH (2016) What is the effective luminance or effective area of non-uniform LED luminaires for discomfort glare rating with UGR? In: Tagungsband Licht 2016, Karlsruhe, pp 563–570 Geerdinck LM, Van Gheluwe JR, Vissenberg MCJM (2014) Discomfort glare perception of nonuniform light sources in an office setting. J Environ Psychol 39:5–13 Gershun A (1936) Световое поле (light field), Moscow. Translated (1939) by Moon PH and Timoshenko G. J Math Phys 18:51–151 Han S, Boyce PR (2003) Illuminance, CCT, décor and the Kruithof curve. In: Proceedings of the 25th CIE Session, San Diego, CA, pp 282–285

References

133

Hara N, Hasegawa S (2012) Study on discomfort glare rating of the luminaire with LED array. J Illum Eng Inst Jpn 96(2):81–88 Hewitt H, Bridgers D, Simons R (1965) Lighting and the environment: some studies in appraisal and design. Lighting Res Technol 30:91–116 Higashi H, Koga S, Kotani T (2013) The development of evaluation for discomfort glare in LED lighting of indoor work place: the effect of the luminance distribution of luminous parts on subjective evaluation. In: Proceedings of CIE Centenary Conference: toward a new century of light, pp 648–656 Hopkinson RG (1957) Evaluation of glare. Illum Eng 52:305–316 Houser KW, Tiller DK, Bernecker CA, Mistrick RG (2002) The subjective response to linear Fluorescent direct/indirect lighting systems. Lighting Res Technol 34:243–264 Howarth PA, Herons G, Greenhouse DS, Bailey IL, Berman SM (1993) Discomfort from glare: the role of pupillary hippus. Lighting Res Technol 25:37–42 Huang A, Sanderson A (2014) Light field modelling and interpolation using Kriging techniques. Lighting Res Technol 46:219–237 Huang W, Yang Y, Luo MR (2018) Discomfort glare caused by white LEDs having different spectral power distributions. Light Res Technol 50(6):921–936 Huang WJ, Yang Y, Luo MR (2019) Verification of the CAM 15u colour appearance model and the QUGR glare model. Lighting Res Technol 51:24–36 Inoue, Y (2010). Study on illuminance balance between working area and ambient—consideration of initial lighting condition, visual task performance and impression of lighting. In: Proceedings of the CIE 2010 lighting quality and energy efficiency, Vienna, pp 776–782 Islam MS, Dangol R, Hyvärinen M, Bhusal P, Puolakka M, Halonen L (2015) User-acceptance studies for LED office lighting: lamp spectrum, spatial brightness and illuminance. Lighting Res Technol 47:54–79 Juslén HT, Wouters MCHM, Tenner AD (2005) Preferred task-lighting levels in an industrial work area without daylight. Lighting Res Technol 37:219–233 Kartashova T, Sekulovski D, De Ridder H, Te Pas SF, Pont SC (2016) The global structure of the visual light field and its relation to the physical light field. J Vis 16(10):9 Kato M, Sekiguchi K (2005) Impression of brightness of a space judged by information from the entire space. J Light Vis Environ 29(3):123–134 Kim W, Koga Y (2004) Position index for a glare source in the whole visual field. J Illum Eng Inst Jpn 88(11):847–852 Kim W, Han H, Kim JT (2009) The position index of a glare source at the borderline between comfort and discomfort (BCD) in the whole visual field. Build Environ 44(5):1017–1023 Kirsch RM (2014) Lighting quality and energy efficiency in office spaces. Doctoral thesis, Department of Lighting Technology, Technical University, Berlin Knez I (1995) Effects of indoor lighting on mood and cognition. J Environ Psychol 15:39–51 Knez I, Kers C (2000) Effects of indoor lighting, gender and age on mood and cognitive performance. Environ Behav 32(6):817–831 Kobayashi S, Inui M, Nakamura Y (2001) Preferred illuminance uniformity of interior ambient lighting. J Light Vis Environ 25(2):64–75 Kruithof AA (1941) Tubular luminescence lamps for general illumination. Philips Techn Rev 6 (3):65–73 Kuehni RG (2016) Kohlrausch, Arnt. In: Luo R (ed) Encyclopedia of color science and technology. Springer, New York Lee CM, Kim H, Choi DS (2007) A study of the estimation of discomfort glare for LED luminaires. In: Proceedings of the 26th Session of CIE, Beijing, vol D3, pp 33–36 Loe DL, Mansfield KP, Rowlands E (1994) Appearance of lit environment and its relevance in lighting design: experimental study. Lighting Res Technol 26(3):119–133 Loe DL, Mansfield KP, Rowlands E (2000) A step in quantifying the appearance of a lit scene. Lighting Res Technol 32(4):213–222

134

4 Visual Satisfaction

Loe DL (2016) Light, vision and illumination: the interaction revisited. Light Res Technol 48:176–189 Loe D, Deng S, Bian J, Van der Burgt PJM (2015) Evaluation of glare from non-uniform indoor luminaires. In: CIE Publication CIE 216:2015 Proceedings of the 28th CIE Session, Manchester, pp 1860–1869 Luckiesh M, Guth SK (1949) Brightnesses in visual field at borderline between comfort and discomfort (BCD). Illum Eng:650–670 Lynes J, Burt W, Jackson G, Cuttle C (1966) The flow of light into buildings. Trans Illum Eng Soc (Lond) 31:65–91 Manov B (2007) An experimental study on the appraisal of the visual environment at offices in relation to colour temperature and illuminance. Build Environ 42:979–983 Mansfield KP (2010) Comment 2 in Cuttle C. Towards the third stage of the lighting profession. Lighting Res Technol 42:89–90 Marsden AM (1969) Brightness—a review of current knowledge. Lighting Res Technol 1:171–181 Marsden AM (1970) Brightness-luminance relationships in an interior. Lighting Res Technol 2:10–16 Miller NJ, McKay H, Boyce PR (1995) An approach to the measurement of lighting quality. In: Proceedings of the IESNA annual conference, New York Morgenstern Y, Geisler WS, Murray RF (2014) Human vision is attuned to the diffuseness of natural light. J Vis 14:1–18 Murray IJ, Plainis S, Carden D (2002) The ocular stress monitor: a new device for measuring discomfort glare. Lighting Res Technol 34(3):231–242 Mury AA, Pont SC, Koenderink JJ (2009) Representing the light field in finite three-dimensional spaces from sparse discrete samples. Appl Opt 48:450–457 Muryy AA (2009) The light field in natural scenes. Doctoral thesis, Faculty of Industrial Design Engineering, University of Technology, Delft Newsham GR, Marchand RG, Veitch JA (2004) Preferred surface luminances in offices, by evolution. J Illum Eng Soc 33(1):14–29 Ngai PY (2016) Evaluations of low ambient task-surround lighting system in a simulated environment. J Solid State Lighting 3(1):1–17 Ngai PY, Boyce PR (2000) The effect of overhead glare on visual discomfort. J Illum Eng Soc 29 (2):29–38 Niedling M, Kierdorf D, Völker S (2013) Influence of a glare source spectrum on discomfort and disability glare under mesopic conditions. In: CIE Centennial Congress, CIE x038:2013:340–347 Oi N, Mansfield KP (2015) Lighting quality: possibility of luminance distribution as its determinant. In: Proceedings CIE 28th Session, Manchester, pp 1111–1120 Park NK, Farr CA (2007a) The effects of lighting on consumers’ emotions and behavioral intentions in a retail environment: a cross-cultural comparison. J Interior Des 33(1):17–32 Park NK, Farr CA (2007b) Retail store lighting for elderly consumers: an experimental approach. Fam Consum Sci Res J 35:316–337 Pont SC (2013) Spatial and form-giving qualities of light. In: Albertazzi L (ed) Handbook of experimental phenomenology: visual perception of shape, space and appearance. Wiley, Chichester Priest IG (1933) A proposed scale for use in specifying the chromaticity of incandescent illuminants and various phases of daylight. JOSA 23(2):41–45 Protzman JB, Houser KW (2005) On the relationship between object modelling and the subjective response. LEUKOS 2(1):13–28 Raynham P (2016) Room lighting in the absence of a defined visual task and the impact of mean room surface exitance. Lighting Res Technol 48:190–204 Raynham P, Unwin J, Guan L (2019) A new metric to predict perceived adequacy of illumination. Lighting Res Technol 1–12. Prepublished Feb 4, 2019. https://doi.org/10.1177/ 1477153519828416

References

135

Royer MP, Houser KW (2012) Spatial brightness perception of trichromatic stimuli. LEUKOS 9 (2):89–108 Safdar M, Luo MR, Mughal MF, Kuai S, Yang Y, Fu L, Zhu X (2018) A neural response-based model to predict discomfort glare from luminance image. Lighting Res Technol 50:416–428 Scheir GH, Hanselaer P, Bracke P, Deconinck G, Ryckaert WR (2015a) Calculation of the unified glare rating based on luminance maps for uniform and non-uniform light sources. Build Environ 84:60–67 Scheir GH, Hanselaer P, Van de Perre L, Ryckaert WR (2015b) Effect of luminance contrast on the perception of discomfort. In: CIE Publication CIE 216:2015 Proceedings of the 28th CIE Session, Manchester, pp 1870–1876 Scheir GH, Hanselaer P, Ryckaert WR (2017) Defining the actual luminous surface in the unified glare rating. LEUKOS 13:201–210 Scheir GH, Donners M, Geerdinck LM, Vissenberg MCJM, Hanselaer P, Ryckaert WR (2018) A psychophysical model for visual discomfort based on receptive fields. Lighting Res Technol 50:205–217 Scheir GH, Hanselaer P, Ryckaert WR (2019) Pupillary light reflex, receptive field mechanism and correction for retinal position for the assessment of visual discomfort. Lighting Res Technol. Prepublished Oct 17, 2017. https://doi.org/10.1177/1477153517737346 Söllner G (1965) Ein einfaches System zur Blendungsbewertung. Lichttechnik 17(5):59A–66A Stevens SS (1961) To honor Fechner and repeal his law. Science 133:80–86 Stevens SS (1975) Psychophysics: introduction to its perceptual, neural and social prospects. Wiley, New York Stone PT (2009) A model for the explanation of discomfort and pain in the eye caused by light. Lighting Res Technol 41:109–101 Tashiro T, Kawanobe S, Kimura-Minoda T, Kohko S, Ishikawa T, Ayama (2015) Discomfort glare for white LED light sources with different spatial arrangements. Lighting Res Technol 47:316–337 Tiller DK, Veitch JA (1995) Perceived room brightness: pilot study on the effect of luminance distribution. Lighting Res Technol 27(2):93–101 Tregenza PR, Romaya SM, Dawe SP, Heap LJ, Tuck B (1974) Consistency and variation in preferences for office lighting. Lighting Res Technol 6(4):205–211 Van Ooyen MHF, Van De Weijgert JAC, Begemann SHA (1987) Preferred luminances in offices. J Illum Eng Soc 16(2):152–156 Veitch JA, Newsham GR (2000) Preferred luminous conditions in open-plan offices: research and practical recommendations. Lighting Res Technol 32(4):199–212 Veitch JA, Stokkermans MGM, Newsham GR (2011) Linking lighting appraisals to work behaviors. Environ Behav 20(10):1–17 Vidovszky-Nemeth A, Schanda J (2012) White light brightness-luminance relationship. Lighting Res Technol 44:55–68 Viénot F, Durand ML, Mahler E (2009) Kruithof’s rule revisited using LED illumination. J Mod Opt 56(13):1433–1446 Wade NJ, Swanston M (2013) Visual perception: an introduction, 3rd edn. Psychology Press, New York Waters CE, Mistrick RG, Bernecker CA (1995) Discomfort glare from sources of nonuniform luminance. J Illum Eng Soc 24(2):73–85 Wei M, Houser KW, Orland B, Lang DH, Sliwinski MJ, Bose M (2014) Field study of office worker responses to fluorescent lighting of different CCT and lumen output. J Environ Psychol 39:62–76 Wilkens AJ (2016) A physiological basis for visual discomfort: application in lighting design. Lighting Res Technol 48:44–54 Wilkins AJ, Nimmo-Smith I, Tait A, McManus C, Della Sala S, Tilley A, Scott S (1984) A neurological basis for visual discomfort. Brain 107(4):989–1017

136

4 Visual Satisfaction

Xia L, Tu Y, Liu L, Wang Y, Peng S, Knoop M, Heynderickx I (2011) A study of overhead glare in office lighting conditions. J SID 19(12):888–898 Xia L, Pont SC, Heynderickx I (2017a) Light diffuseness metric part 1: theory. Lighting Res Technol 49:411–427 Xia L, Pont SC, Heynderickx I (2017b) Light diffuseness metric part 2: describing, measuring and visualising the light flow and diffuseness in three-dimensional spaces. Lighting Res Technol 49:428–445 Yakahashi H, Irikura T, Chamnongthai K (2013) Study of ethnic differences in subjective evaluation of interior lighting. In: Proceedings LuxPacifica, Bangkok, pp 47–50 Yang Y, Luo MR, Huang WJ (2017a) Assessing glare. Part 1: comparing uniform and non-uniform LED luminaires. Lighting Res Technol 49:195–210 Yang Y, Luo MR, Huang WJ (2017b) Assessing glare. Part 2: modifying unified glare rating for uniform and non-uniform LED luminaires. Lighting Res Technol 49:727–742 Yang Y, Luo MR, Huang WJ (2018a) Assessing glare. Part 3: glare sources having different colours. Lighting Res Technol 50:596–615 Yang Y, Luo MR, Huang WJ (2018b) Assessing glare. Part 4: generic models predicting discomfort glare of light-emitting diodes. Lighting Res Technol 50:739–756 Yaodong C, Zhe C, Luoxi H (2014) Chinese preferences for living room lighting in relation to CCT and color metrics. China Illum Eng J 9(1):11–17 Zhang J, Tu Y, Liu L, Wang L, Peng S (2013) Effect of the correlated color temperature of light on overhead glare in offices. In: Conference proceedings display week of SID, Vancouver, pp 1096–1098

Chapter 5

Non-visual Biological Mechanism

Abstract Daily (circadian) bodily rhythms, a fundamental property of human life, are synchronised by the natural 24-h dark-light rhythm. This entrainment by light is one of the non-visual biological effects of light. In particular, the rhythms of the hormones cortisol, supplying energy to the body, and melatonin, facilitating sleep, are important. A relatively new type of photoreceptor discovered in 2002, the photosensitive retinal ganglion cell pRGC, connects with the suprachiasmatic nucleus SCN, a structure within the brain that acts as a master biological clock. The SCN, in its turn, has pathways to the pineal gland, where melatonin is produced, and to the adrenal cortex responsible for the production of cortisol. Light may, apart from effects on circadian rhythms, also have direct, acute photobiological effects that influence alertness and performance. The spectral sensitivity of the pRGCs, given by their photopigment melanopsin, is different from that of rods and the three types of cones. Its sensitivity peaks in the blue part of the wavelength range. Rods and cones have a neural connection with ganglion cells, and consequently their signals interplay with the signal obtained from the pRGC itself. Much of this neural wiring is as yet unknown. Primarily because of this, it is impossible to define a single spectral sensitivity function or action spectrum for all non-visual effects of light. The correlated colour temperature can be used only as a rough indication for the characterisation of the spectrum of lamps for non-visual biological use. The spectrally weighted irradiances for the five human photopigments (α-opic irradiances) are at this moment the best characterisation.

Visual effects of light have been seriously studied for more than 500 years. Only some decades ago medical and biological researchers started to learn that light entering the eye also has non-visual biological effects. These effects influence the way our body “operates” and therefore influence our health and well-being. Given the relatively short period that non-visual biological effects of light are being studied, it may not come as a surprise that there still are many questions that need answers. Today, all over the world scientists from disciplines such as biology, chronobiology, medicine and lighting are making studies and carrying out © Springer Nature Switzerland AG 2019 W. van Bommel, Interior Lighting, https://doi.org/10.1007/978-3-030-17195-7_5

137

138

5 Non-visual Biological Mechanism

Fig. 5.1 Visual and non-visual biological effects of lighting and benefits of good lighting

visual effects

visual performance and comfort

work performance less errors

lighting

safety less accidents biological effects

health and well-being

less absence

experiments. A large number of publications in many professional journals increase, more or less on a daily basis, the knowledge of fundamental and application aspects of non-visual biological effects of lighting. Only if we succeed in designing and installing lighting that results in both suitable visual and non-visual biological effects can we expect optimised visual performance and comfort, health and well-being. In the context of lighting for work, this, in turn, propagates better work performance, fewer errors, improved safety and lower absenteeism. Figure 5.1 sketches the relationship between the different effects of lighting and shows the benefits to be expected from good lighting. An image-forming process in the eye produces visual effects. Other processes produce non-visual biological effects of lighting. They are, therefore, sometimes referred to as “non-image-forming effects of light” (in short, NIF effects of light). Lighting that takes both the visual and non-visual biological effects into account is often referred to as “human-centric lighting” (in short HCL). CIE defined in a draft international standard the term “integrative lighting” with the same meaning as human-centric lighting: “lighting specifically designed to produce a beneficial physiological and/or psychological effect upon humans, including visual and non-visual effects” (CIE 2016).

5.1 5.1.1

Circadian Rhythms The Principle

All living species live under a 24-h dark-light rhythm caused by the rotation of the earth around its axis in exactly 24 h. A 24-h rhythm is called a circadian rhythm after the Latin expression “circa diem” for “around a day”. Since the 1960s scientists agree that circadian rhythmicity is a fundamental property of all plant, animal and human life (Chovnick 1960). Table 5.1 gives examples of circadian rhythms that exist in humans. In this context, it is also interesting to note that some medical interventions have shown a better chance of success if they are carried out at specific times. Coordinating the timing of intake of medication with the body’s circadian rhythm (referred to as chronotherapy) often improves the positive effect of the medication, permitting

5.1 Circadian Rhythms

139

Table 5.1 Examples of human circadian rhythms Time 06:00–07:00 07:00–08:00 08:00–09:00 09:00–10:00 10:00–11:00 11:00–12:00 12:00–13:00 13:00–14:00 14:00–15:00 15:00–16:00 16:00–17:00 17:00–18:00

Bodily aspect Hay fever symptoms most intense Testosterone peaks Cortisol peaks Heart/cerebral infarcts peak Uric acid peaks Cholesterol peaks Adrenaline peaks Best reaction time Heart rate peaks Muscle strength peaks Urinary flow peaks

Time 18:00–19:00 19:00–20:00 20:00–21:00 21:00–22:00 22:00–23:00 23:00–24:00 24:00–01:00 01:00–02:00 02:00–03:00 03:00–04:00 04:00–05:00 05:00–06:00

Bodily aspect Blood pressure peaks Body temperature peaks Melatonin secretion starts

Bowel movements suppress Cortisol lowest Growth hormone peaks Melatonin peaks Store info in the brain Asthma symptoms most intense Body temperature lowest

lower dosages and reducing adverse effects (Kaur et al. 2013). Burns that happen during daytime heal approximately 60% faster than those that happen during nighttime (Hoyle et al. 2017). As early as 1729, a French astronomer, Jean-Jacques de Mairan, observed that leaf movements of mimosa plants placed in a dark cupboard continued to show rhythmic behaviour (De Mairan 1729). This marked the first suggestion for an internal clock mechanism in living organisms that generate rhythms. About 100 years later the Swiss botanist Augustin De Candolle showed that the rhythmic mimosa leaf movements under constant illumination have a self-sustained period 1–2 h shorter than they have under the natural light-dark period of 24 h (De Candolle 1832). It represents the first hint for the natural 24-h light-dark rhythm to act as a synchronising aid for the internal clock mechanism. The sunflower’s name refers to its appearance with the sun (Fig. 5.2). Sunflowers still have another connection with the sun. Sunflowers turn with the sun from east to west during the day and turn back during the night to face east again to restart the same turning cycle at dawn. For this reason, the French name for sunflower is “tournesol” (turning to the sun). Only in 2016 scientists from the universities of California and Virginia could prove that an internal circadian oscillator (i.e. a clock mechanism) coordinates this process (Atamian et al. 2016). A famous series of human circadian experiments have been carried out between 1964 and 1989 by the chronobiology pioneer Jurgen Aschoff of the Max Planck Institute in Germany (Aschoff 1965; Zulley and Knab 2000). Altogether, somewhat more than 200 test persons participated in these tests. Many test persons were students who studied for an examination. Aschoff used for his experiments a sound and lightproof underground bunker near Munich in Germany. Here each

140

5 Non-visual Biological Mechanism

Fig. 5.2 Sunflowers, with the appearance of the sun, tracking the sun controlled by an internal clock mechanism

Fig. 5.3 Wake-sleep rhythm in the case of external environmental cues (left) and an example of a free-running rhythm in the case without external environmental cues (right). In the example, the daily shift is 1 h

day 1 2 3 4 5 6 7 8 16:00

24:00

08:00

24:00 16:00

time

24:00

08:00

24:00

time

test person lived in complete isolation for up to some 4 weeks in a comfortable bed-living room with a shower and small kitchen attached. The test person was asked to behave according to his subjective feeling of the time of day or night. The test person prepared his own three meals per subjective day and turned the light off when he or she thought it was bedtime and on again when he or she thought it was morning and time to rise. The test person was free to do what he or she wished. For physical exercises, a home trainer bike was available. Body temperature, activity and sleep patterns were measured by sensors. Communication with the outside world was done only by letters. The daily rhythm of most test persons drifted away from the outside 24-h rhythm. The rhythm without any external environmental clues is usually called the free-running biological rhythm. Figure 5.3 shows on the left a constant wakesleep rhythm, for 8 days, with external environmental cues. The right part of the picture shows an example of the daily shift to a later time for the free-running situation without external cues. After their stay of some 3 weeks, many test persons had, to their surprise, “lost” half to one and a half day: that means the calendar date was more advanced than they thought (Fig. 5.4). One of the test persons said: “did something bad happen, because it is only October 15 and according to plan I would be released on

5.1 Circadian Rhythms Fig. 5.4 Difference between real date and apparent date after a stay of 24 days in an environment without external clues. Example of a person with a 25-h free-running rhythm (figure not to scale) Fig. 5.5 Typical example of the daily rhythm of core body temperature shown on a 2-day scale (based on Boivin and Czeisler 1998; CIE 2004)

141

date: 24 hr synchronized free-running apparent date:

1

2

24

24

3

25

24

24 hr

25

25

25

25 hr

1

2

3

24

Body temp (° C) 37.5 37.0 36.5

6

12

18

24

6

12

18

24

6

Clock time October 16” (Zulley and Knab 2000). The free-running biological rhythm of all test persons was close to 24.5 h with variations from slightly less than 24 h to slightly more than 25 h.

5.1.2

Bodily Rhythms

The human rhythms studied in the experiments in Aschoff’s bunker were focused on sleep, eat and performance patterns. The core body temperature was one of the body properties that were monitored. We now know quite accurately how in healthy persons the body temperature changes during the day and night (Fig. 5.5). Early in the morning, the body temperature reaches its minimum value and from that moment onwards it gradually increases until it reaches its maximum in the early evening, approximately 0.5
C higher. There are many more body properties with a circadian rhythm. Examples are the heart rate, blood pressure, liver function, generation of new cells and production of many hormones (Foster and Kreitzman 2004). More of such properties are shown in Table 5.1. The hormones melatonin and cortisol are of particular importance because they govern sleep and alertness. Cortisol increases glucose (blood sugar) to give the body energy and enhances the immune system. It is sometimes, popularly, referred to as the energy hormone. If cortisol levels are too high over an extended period, the bodily system becomes exhausted and inefficient. Melatonin slows down some bodily processes and evokes sleep. It is, therefore, often referred to as the sleep hormone. Figure 5.6 shows the daily rhythm of cortisol and melatonin hormones in healthy persons.

142

5 Non-visual Biological Mechanism

Fig. 5.6 Typical example of daily rhythms of the hormones cortisol and melatonin, shown on a 2-day scale (based on Boivin and Czeisler 1998; CIE 2004)

Cortisol and Melatonin (%) 100 Cortisol

75 50

Melatonin

25 0

6

12

18

24

6

12

18

24

6

Clock time Cortisol levels increase in the morning and prepare the body for the coming day’s activity. They then decrease gradually but remain at a high enough level to give sufficient blood sugar, and thus energy, throughout the bright day, falling finally to a minimum around midnight. The quantity of the sleep hormone melatonin drops in the morning to nearly zero, reducing sleepiness. Healthy persons have during the daytime hardly any melatonin in their body. It gradually rises again early in the evening until it reaches its maximum level some 2–3 h after midnight. The composition of cortisol and melatonin levels permits for activity and alertness during daytime: high levels of cortisol and no melatonin. During night-time, however, they allow for a good sleeping quality: low levels of cortisol and high levels of melatonin.

5.1.3

Biological Clock

The previous section described the effects of a sort of biological clock system. But, has a biological clock actually been discovered? In 1957, Pittendrigh introduced the possibility of a pacemaker organ in animals and humans (Pittendrigh and Bruce 1957). In 1972, two different groups of researchers did brain lesion experiments of the suprachiasmatic nucleus, a tiny organ located in the brain area called hypothalamus. In lesion experiments, the effect of disabling organs is studied. They found after lesion of the suprachiasmatic nucleus of rats that the animals lost their circadian rhythmicity of drinking, eating and wheel-running activity (Moore and Eichler 1972; Stephan and Zucker 1972). These 1972 experiments represent the discovery of the suprachiasmatic nucleus (SCN) being the pacemaker organ (biological clock) in mammals. In fact, the nucleus consists of two side-by-side-lying nuclei. In humans, each of the tiny two suprachiasmatic nuclei contains a few tens of thousands of nerve cells (neurons) transmitting information through electrical and chemical signals (Cassone et al. 1988). Circadian rhythms have a genetic aspect. The 2017 Nobel Prize for “physiology or medicine” was won by the American researchers Hall, Rosbach and Young for their pioneering work in the 1980s about how clock genes play a role in the molecular mechanism that produces circadian oscillation in the biological clock.

5.1 Circadian Rhythms

143

They isolated a gene in fruit flies that controls the daily biological rhythm and uncovered the mechanism governing the clockwork inside the cell. Key publications of these researchers are Bargiello et al. (1984) and Emery et al. (1998). Still much more research is needed to understand the molecular mechanism of the biological clock completely. Today it is acknowledged that the human body has a multitude of biological clocks. Peripheral clocks are, for example, located in the stomach, liver, kidneys, adrenal gland, lungs, muscle tissue and even the retina itself (Foster and Kreitzman 2004). The SCN in the brain acts as the central pacemaker that synchronises the “peripheral clocks”: a master-slave system (Balsalobre et al. 2000). The identification of the suprachiasmatic nucleus SCN as the master biological clock has been an important step in understanding how the circadian clock optimises our bodily processes. The SCN is not only a system that generates rhythms. It is also a sensor to collect environmental information that is used to synchronise the body with the 24-h rhythm (Roenneberg et al. 2013).

5.1.4

Chronotypes

Figures 5.4 and 5.5 did show a body rhythm with a period of 24 h. The bunker experiments of Aschoff, however, already showed that the free-running human rhythm is, with most people, not exactly 24 h. In fact, the period may be a little more than 1.5 h different between individuals. The average rhythmic period of a large group of people is 24 h and 15–30 min (Czeisler et al. 1999; Brown et al. 2005). The actual rhythm of a person’s biological clock determines his or her chronotype, i.e. is he a morning type of person or an evening type. Morning types of persons have a relatively short free-running body rhythm, which can be as short as 23.5 h. The moment their melatonin starts rising, called the dim-light melatonin onset (DLMO), is earlier than with evening chronotypes. Consequently, melatonin levels are earlier in the evening high, making them sleepy and wanting to go to bed early. The free-running rhythm of the biological clock of evening chronotypes may even be slightly more than 25 h. Their melatonin level is still relatively low late in the evening. They do not yet feel sleepy and do not yet like to go to bed. Questionnaires about the timing of sleep during the workweek and time spent in daylight can be used to indicate the chronotype of a person (Roenneberg et al. 2003). Chronotype has an age and gender dependency. The longest free-running rhythm occurs in the puberty until circa 21 years when it starts shortening again. Females tend to have a slightly shorter rhythm than men (Roenneberg et al. 2004; Duffy et al. 2011). Fortunately, body rhythms are normally each day resynchronised to the 24-h rhythm, a process called entrainment.

144

5 Non-visual Biological Mechanism

5.1.5

Entrainment

5.1.5.1

Entrainment by Light-Dark Transitions

Because the free-running rhythm of the biological clock is with most people different from the 24-h solar day, external cues which have an exact 24-h rhythm have to synchronise our biological clock with the 24-h rhythm. Examples of cues are lightdark and dark-light transitions, temperature changes, feeding and changes in the locomotor activity of the body (moving and exercising). In mammals, light is by far the most important cue (Yamazaki et al. 2000; Berson 2003; Roenneberg et al. 2013). Under normal situations, the natural 24-h dark-light rhythm synchronises the free-running rhythm with the 24-h rhythm. This process is called “entrainment”. Aschoff used the German word “zeitgeber” for this signalling process (Aschoff 1954). In chronobiology, this word, meaning “time giver”, has become the usual expression also in English language literature. Especially morning light is a powerful zeitgeber (Rosenthal et al. 1990; Terman et al. 1995; Revell et al. 2012).

5.1.5.2

Entrainment by Colour Transitions

The rotation of the earth around its axis does cause not only a 24-h dark-light rhythm but also a 24-h colour rhythm. During twilight, when the sun is under the horizon, skylight contains shorter wavelengths (blue light) than when the sun is over the horizon. When the sun is under the horizon, direct sunlight is screened by the earth’s surface. Only light radiated upwards from the sun into the atmosphere and scattered at microscopic particles can be seen. Short wavelengths scatter more at these particles than long wavelengths (Rayleigh scattering). It results in a high blue content (Fig. 5.7, left). When the sun climbs above the horizon, the sky gets a warmer colour because of the addition of the more yellowish direct sunlight (Fig. 5.7, right).

Fig. 5.7 Colour change of the sky at dawn; left: the sun still under the horizon and right: just above the horizon

5.1 Circadian Rhythms

145

Recently Walmsley et al. (2015) demonstrated by comparing normal and genetically modified mice that the colour transition at twilight can contribute to the entrainment of biological rhythms. Other recent studies confirm that changes in the colour of light indeed can be a cue for synchronising circadian activity (Pauers et al. 2012; Van Diepen et al. 2015). Further studies are needed to determine the practical meaning of colour transition signals for synchronising human biological rhythms.

5.1.5.3

Non-photic Entrainment

Apart from light, other cues or zeitgebers may help in entraining the circadian rhythm. Examples are environmental temperature, regular food intake and physical activity (Roenneberg et al. 2007; Refinetti 2015). It explains that a relatively large percentage of totally blind people, without a functional retina, can, to a certain extent, entrain their biological clock (Klerman et al. 1998).

5.1.6

Phase Shifting

Light and dark can shift the biological clock. For example, after an intercontinental flight through some time zones, it is the rhythm of light and dark at the new location that shifts the biological rhythm into the rhythm of the new time zone. With a larger time difference more days and nights at the new location are needed for a complete shift. The new light-dark rhythm after changing shift may shift a shift worker’s biological clock. So-called social jetlag may occur when the social time (school and work time) is not in line with the biological clock. It occurs mainly with late chronotypes. During working days they rise when their body according to their biological clock is still in the sleeping state. During the weekend they compensate for the built-up sleep deprivation by sleeping late in the morning. The consequence is that their biological clock shifts back and forth. However, shifting goes not fast enough so that their biological clock is always out of phase with their daily routine. Without special measures, the time needed for a total shift of the clock to a new time zone may take from some 2 days to more than 4 days, depending on the actual time difference. Shifting the biological rhythm is called phase shifting. How fast the phase of the rhythm shifts and in which direction depends on the timing of light and dark exposure and to light’s intensity and colour. The curve of Fig. 5.8 shows how light at different times shifts the phase of the clock. The curve is called the phase– response curve (PRC curve). The changeover point from “delay” to “advance” lies at or slightly after the moment of occurrence of the body’s minimum temperature which, in turn, is dependent on chronotype. The phase–response curve of morning chronotypes is consequently shifted to a somewhat earlier time relative to the curve of evening chronotypes. This is shown in Fig. 5.8 by the thin dotted curve (morning chronotype with a free-running rhythm of 23.5 h) and the thin dashed curve (evening type with a free-running rhythm of 25 h), respectively.

146

5 Non-visual Biological Mechanism

Phase +3 advance (hrs)

+2 +1 0

Tmin 12

18

24

06

12 hrs

+ 6 hrs 18

24

06

12

18 hrs

- 4 hrs 08

14

20

02

08 hrs

-1

Phase delay

-2

(hrs)

-3

Fig. 5.8 Phase–response curve. Light exposure of 3500 lux at eye level for 2 h; Tmin is the moment of minimum body temperature. Thin dotted line: morning chronotype (23.5 h); thin dashed line: evening chronotype (25 h). For two different time zones (+6 and 4 h), the scales at the bottom indicate when light (open arrows) and when darkness (black arrows) is needed at the new destination to shift the clock towards the new time zone. Curve drawn from data of Eastman and Burgess (2009) determined on the basis of body temperature

In the middle of the day, roughly from 11:00 to 18:00 h, light has no phaseshifting response. It is the period where light provides stable entrainment of the biological rhythm. In the middle of the night and early morning, light results in large phase-shifting responses. Light in the early night delays the clock. For morning chronotype of persons, with a fast-running clock (free-running rhythm less than 24 h), the clock has to be delayed. Late evening/early night light is doing that. Light in the late night and early morning advances the clock. Slow-running clocks (freerunning rhythm more than 24 h) have to be advanced and morning light is doing that. Since the majority of humans have a free-running rhythm of more than 24 h (evening chronotypes) bright light in the morning is beneficial for most people. As an example of how to use the phase–response curve in the practice of “fighting” jetlag, the two scales at the bottom of Fig. 5.8 indicate for a new time zone of +6 and 4 h, respectively, when light (white arrows) and when darkness (black arrows) is needed at the new destination to shift the clock quickly in the right direction. In the case of the advanced time zone of +6 h, the phase of the clock has to be advanced by light in the morning and darkness in the evening. The delayed time zone of 4 h requires light in the evening after approx. 21:00 h and darkness just after midnight and early morning.

5.2 Third Type of Photoreceptor

147

Phase shifts can relatively easily be determined from the measurement of the shift in the time of minimum body temperature. More accurate methods require laboratory measurements of melatonin (from samples of saliva taken with a cotton swab) to determine the time of melatonin onset. A low-cost estimate of phase shift without imposing any restrictions on the test persons is possible by collecting light and activity data from test persons wearing an action watch (Woelders et al. 2017). An action watch is a watch that measures movements of the test person. For the purpose of relating activity to light, the watch also contains a light sensor. Early research suggested that different types of biological rhythms, like body temperature, sleep-wake and melatonin rhythms, react differently to light exposure as far as the phase shift is concerned (Winget et al. 1978). More recent research suggests that the rhythms respond more similarly. Therefore, the phase shift determined for one aspect can be treated as being valid for other aspects as well (Dunlap et al. 1995).

5.2 5.2.1

Third Type of Photoreceptor Photosensitive Retinal Ganglion Cells (pRGCs)

Already in 1967, it was found from experiments with rats and monkeys that a pathway connects the biological clock (SCN) with the retina (Richter 1967). However, in the 1990s, it was shown that mice without any rods and cones keep their regular entrainment by light-dark transitions (Foster et al. 1991; Freedman et al. 1999). So, rods and cones in the retina do apparently not connect with the SCN. Therefore, a third type of photoreceptor, in addition to the rod and cone types, must exist in the retina which relates to non-visual biological effects. In 2002, Dave Berson and co-workers from the Brown University in the United States indeed discovered such a third type of photoreceptor. A small part of the already longknown ganglion cells in the retina (Fig. 5.9) appear to be sensitive for light incident on them (Berson et al. 2002). This light-sensitive cell is called “photosensitive retinal Fig. 5.9 Part of the retina with photoreceptor cells, including photosensitive retinal ganglion cell, pRGC (in purple). White arrows represent signals to the brain as a result of the transformation of light incident on cones and rods, the blue arrow of light incident on pRGC

to brain

light

pRGC ganglion cells

collector cells

cones and rods

148

5 Non-visual Biological Mechanism

ganglion cell”: pRGC. (In literature it is also referred to as ipRGC which stands for intrinsically photosensitive retinal ganglion cell.) This particular type of ganglion cell transmits signals to the SCN in the brain. The photoreceptors cones and rods got their name from their cone and rod shape. In this context, pRGC photoreceptors could be thought of as tiny “spheres”. Only some 2% of all ganglion cells are photosensitive (Hattar et al. 2002). The photopigment or opsin in pRGC cells is different from the photopigments in rods and cones. The density of photopigments in pRGC cells is some 10,000 times smaller than that of the rod and cone pigments (Do et al. 2009). All this means that relatively bright light is needed to obtain an effective neural signal from the total of the pRGC cells. The pRGC cells contain the pigment melanopsin (melan-opsin) (Provencio et al. 2000; Hattar et al. 2002; Kumbalasiri and Provencio 2005). As photopigments are responsible for the spectral sensitivity of photosensitive cells, the pRGC cells have their own specific (intrinsic) spectral sensitivity. Section 5.5.2 shows spectral sensitivity curves.

5.2.2

Retinal Neural Wiring

Like “normal” ganglion cells, pRGCs receive signals from cones and rods and transmit these signals to the area of the brain responsible for creating visual sensations. Figure 5.4 illustrates this with the white arrows. It was already explained in some detail in Chap. 1. The same pRGC cells also transform the light incident on their surface into a signal that is transmitted to the location in the brain where the SCN is located. So, they have a “double” function. The neural wiring in the retina results in some interaction between rod, cone and pRGC light-converted electric signals. It explains why modified animals in which only the melanopsin photopigment of the pRGCs is removed without completely removing the cell itself still do show some photobiological action (Hattar et al. 2003; Lucas et al. 2003; Panda et al. 2003). In this situation, the intrinsic photosensitivity of the pRGCs is absent, but they still can transmit signals to the SCN that pass through them from cones and rods. Genetically modified animals in which the pRGC cells are completely removed show no photobiological actions. The interaction between pRGCs, cones and rods plays a role in which non-visual effects of light are evoked under different types of light. Section 5.5 discusses this in more detail. Quite some aspects of these interactions are, however, not yet known.

5.2.3

Spatial Distribution of pRGCs

The distribution of cones and rods in the retina is well known as described in Chap. 1. The distribution of the photosensitive retinal ganglion cells, however, is still being studied. There is evidence that their distribution is not evenly distributed

5.3 Pineal Gland and Adrenal Cortex

149

over the retina. It could mean that light arriving from different directions at the eye, and thus illuminating different parts of the retina, has different non-visual biological effects. Nasal light exposure has been shown to be more effective than temporal exposure, and exposure of the inferior part of the retina more effective than exposure of the superior part (Lasko et al. 1999; Visser et al. 1999; Rüger et al. 2005; Glickman et al. 2003; Xu and Van Bommel 2011; Piazena et al. 2014). The latter would mean that light arriving from above is more effective for non-visual biological effects than light arriving from below. The effect of a non-even distribution of pRGCs is perhaps combined with the effect of a possibly different individual sensitivity of pRGCs (Broszio et al. 2017; Knoop 2018). More studies are needed before the consequences can be quantified accurately enough.

5.2.4

Field of View

To evaluate the amount of light on the retina for non-visual effects of light, often the vertical illuminance at the position of the eye is used. It is usually a suitable method for a general evaluation of lighting installations as far as non-visual lighting effects are concerned. However, for research purposes, this is often not good enough. The actual field of view of a person is determined by his or her direction of view. If, for example, a study is done with test persons carrying out a visual task on a desk, the plane of the eye is not vertical but tilted towards the desk. Moreover, the upper eyelid and the nose block some of the light directed towards the eye. This limits the actual field of view. CIE proposes to take as limits for the vertical field of view 50
above and 70
below the line of view. For situations where viewing is done with both eyes a 180
horizontal field of view can be used (CIE 2018). The illuminance at the plane of the eye for the actual field of view can be measured with an illuminance meter provided with baffles (screens). Note that outdoors the upper eyelid lowers itself dependent on the outdoor brightness. Consequently, the upper vertical limit is much lower outdoors.

5.3

Pineal Gland and Adrenal Cortex

Around the 1980s research into the pathways between the retina and different structures within the brain and other parts of the body started to provide results (Aschoff 1981; Klein et al. 1991; Hattar et al. 2002). The neural pathway between the retina and the biological clock, SCN has been established (Moore et al. 1995) and is shown in Fig. 5.10 with a light-blue line. The SCN itself connects to different other regions. A neural pathway descends from the SCN through the spinal cord and then ascends again to connect with the pineal gland (see Fig. 5.10 again) (Moore et al. 1967; Morin 1994; TeclemariamMesbah et al. 1999). Already in 1958, the American Alan Lerner discovered the

150

5 Non-visual Biological Mechanism

PVN

hypothalamus visual cortex

SCN time

pineal gland melatonin

retina

pituitary gland

light-dark

superior cervical ganglion

spinal cord

adrenal cortex cortisol

Fig. 5.10 Pathways between the retina and different areas in the brain and body. In red: the visual pathway towards the visual cortex. In blue the pathway towards the biological clock (SCN) and onwards, through the spinal cord, to the pineal gland which secretes melatonin. The dashed green line shows the pathway towards the adrenal cortex (on top of the kidneys) which secretes cortisol. Drawing not to scale

hormone melatonin (Lerner et al. 1958). The pineal gland produces and secretes melatonin according to a pronounced circadian rhythm. Now we have a connection between light received by the retina, timing obtained from the SCN and melatonin hormone levels produced by the pineal gland. This connection is essential for the proper functioning of our body. Another pathway extends from the SCN towards the adrenal cortex, the outer part of the adrenal glands, sitting on top of the kidneys (adrenal stands for the Latin “ad renes” which means “near kidneys”) (Kalsbeek and Buijs 2002). Only the part of this pathway from the SCN towards the PVN (paraventricular nucleus of the

5.4 Direct Photobiological Effects

151

hypothalamus) is an “ordinary” neural pathway (Fig. 5.10). The PVN, controlled by signals from the SCN, secretes hormones that act as a chemical messenger for the pituitary gland to secrete another type of hormone which, in turn, alerts the adrenal cortex to produce hormones. This part of the pathway that does not use neurons as messengers but hormones is called an endocrine pathway. It is indicated in Fig. 5.10 with a dashed line. The pituitary gland is also known, from its Latin expression, as “hypophysis”. Recent research suggests that, apart from this pathway, additional pathways connect the SCN with the adrenal cortex (Dickmeis 2009). The adrenal cortex produces and secretes the hormone cortisol. As with the secretion of the hormone melatonin also cortisol secretion follows a clear circadian pattern. So, here we have a connection between light received by the retina, timing obtained from the SCN and cortisol levels produced by the adrenal cortex necessary for proper energy management of our body. Apart from the two pathways described above and shown in Fig. 5.10, there are more pathways. In Chap. 1, it was, for example, described that there exists a pathway between some pRGC cells and a nucleus in the brain (OPN) that controls the pupillary reflex (Lucas et al. 2003; Chen et al. 2011; Takahashi et al. 2011). Of course, the actual pupil size directly influences the amount of light reaching the retina and thus influences both visual and non-visual effects of lighting.

5.4

Direct Photobiological Effects

In the previous sections, the focus has been on long-term effects of light and lighting on circadian rhythms. Light received up to a few days before can cause these effects. Light can, however, also have short-term, direct effects within a short time after being exposed to light. The effects usually disappear shortly after the light exposure stops. The pupillary reflex mentioned before is an example of such direct effect, as well as the direct influence of light on the heart rate. Immediate partly melatonin suppression by light during the night is yet another example of a direct effect. After the short-term night-time light exposure stops, the melatonin level is in less than 15 min restored to the normal night-time level (McIntyre et al. 1989a, b; Rea 2002). Therefore, short-term light exposure during the night has no important consequences for the normal melatonin rhythm. However, the short-term direct effect of brightlight exposure, for example, while going to the toilet, may make falling asleep again a little difficult. Gradually increasing light during awakening from a so-called light alarm clock has shown to increase cortisol levels directly after awakening accompanied by a higher arousal level (Thorn et al. 2004). Also this is a result of direct photobiological effects. Light, both at day and night, can acutely decrease sleepiness and improve alertness (Phipps-Nelson et al. 2003; Lowden et al. 2004; Vanderwalle et al. 2006; Lockley et al. 2006; Cajochen 2007). The increase of alertness, in turn, has positive consequences for performance. Chapter 6 discusses in more detail the alerting effects of light and its effect on performance.

152

5 Non-visual Biological Mechanism

Fig. 5.11 Functional MRI scans at different moments of light exposure; green dot: activity in SCN, yellow dot: in the brainstem, blue dots: in the thalamus, red dots: widespread activity in many different areas (Vanderwalle et al. 2009)

Functional brain scans (fMRI) can demonstrate direct effects of light. They can show brain activity related to tasks being carried out at different moments after light exposure. As an illustration, Fig. 5.11 shows an example of a set of fMRIs recorded throughout 20 min of light exposure while the test person carries out an auditory task (Vanderwalle et al. 2009). At the start of light exposure, activity is recorded in the SCN (green dot). Some tens of seconds later, the brainstem also shows activity (yellow dot). After 16–20 min, many brain areas show activity (red dots) as a direct result of light exposure. The thalamus is activated during the whole period of light exposure (blue dots). The mechanism underlying direct non-visual biological effects of light is less clear than the circadian effects. The cerebral cortex (the crumpled-looking top part of the brain) plays an important part in it (Jones 2003; Saper et al. 2005; Vanderwalle et al. 2009; Chellappa et al. 2011). Rautkylä et al. (2012) proposed a tentative and hypothetical model with connections between the retina and brain regions in the cerebral cortex.

5.5

Spectral Sensitivity

5.5.1

Action Spectra

5.5.1.1

Melatonin Suppression by Monochromatic Light

Chapter 1 discussed that spectral sensitivities for vision are dependent on the type of “visual action” of the light (creating brightness, enabling detection, enabling small detail vision). For spectral sensitivities for non-visual biological effects, it is the non-visual biological action that is decisive. Such actions by light include melatonin suppression, change of body temperature, change of alertness, change in sleeping quality, phase shifting of the biological rhythm and change of pupil size. Melatonin levels in human test persons can easily be measured from a sample of saliva taken with a cotton swab. Therefore, the non-visual action of melatonin

5.5 Spectral Sensitivity

153

suppression is often used in research studies with human test subjects. A more intrusive method, using blood samples, gives more accurate results. Two groups of researchers, one from the United States and one from the UK, studied, independently from each other, spectral sensitivities for non-visual effects of light. Both groups used as basis melatonin suppression by light incident on the eye. They also used the same measure for it: the time of light exposure required at night to lower the melatonin level in test persons to 50% of the maximum melatonin level of that same person in complete darkness. The results of both groups were published in the same year (Brainard et al. 2001; Thapan et al. 2001). To illustrate the procedure of these types of investigations, Fig. 5.12 shows the setup of the experiments of the US group. Because melatonin is almost completely suppressed during daytime in healthy persons, all experiments were done during the night. To ensure a same “light history” only test persons that kept a regular sleep-wake rhythm were accepted. Each test person waited before the actual test started from midnight to 2.00 a.m. blindfolded in the laboratory. After that, the subject viewed for 90 min, head rested and fixed in a holder, the interior of a sphere which was uniformly illuminated with monochromatic light from a monochromator. Before and after the test, a blood sample was taken from which melatonin suppression was determined in a laboratory. Blood samples were used here because of its larger accuracy compared with measurements from saliva samples. Test persons came back several nights to repeat the test with different light levels for a total of nine different monochromatic wavelengths. Brainard’s tests were carried out with 72 persons. From this description, it should not come as a surprise that completing all experiments required some years. Figure 5.13 shows the measurement results in terms of relative spectral sensitivity for both research groups. Short-wavelength light (blue-green) is far more effective for melatonin suppression than long-wavelength light (yellow-reddish). Melatonin suppression-based

Fig. 5.12 Laboratory set-up for research on melatonin suppression by light of different wavelengths (drawing: Brainard et al. 2001)

154

5 Non-visual Biological Mechanism

Fig. 5.13 Measured relative sensitivities of monochromatic light of different monochromatic wavelengths based on melatonin suppression during the night shown together with V(λ) curve for photopic vision (source: Brainard et al. 2001; Thapan et al. 2001)

Relative Sensitivity (%)

Brainard

100

Thapan photopic

80 V(λ)

60 40 20 0 400

450

500

550

600

650

700

λ (nm)

sensitivities are largely different from those corresponding to the photopic V(λ) curve that is also shown in Fig. 5.13. It is proof that photometric V(λ)-based units are unsuitable for predicting melatonin suppression, and, more generally, non-visual biological effects by light.

5.5.1.2

Melatonin Suppression by Polychromatic Light

The tests described in the previous section concern monochromatic light. Most coloured LEDs have such a narrow spectrum (half-maximum width smaller than some 50 nm) that, in this context, they can often be considered as monochromatic. White-light sources, of course, are not monochromatic but polychromatic. For visual effects of light, the so-called law of Abney of additivity of luminances is valid. It says that the luminance of a composition of different wavelengths (polychromatic light) is the same as the sum of the individual monochromatic luminances contained in that composition. This is why the photopic V(λ) spectral sensitivity curve can be applied, for vision, for both mono- and polychromatic light. Unfortunately, this is not the case for non-visual biological effects of light (Rea et al. 2004; Mure et al. 2007; Van Gelder and Mawad 2007). It has been shown that longer wavelengths contained in polychromatic light can, to some extent, counteract the melatonin suppression effect of a shorter wavelength. It means that the monochromatic research results of Brainard and Thapan described in the previous section cannot directly be used as a basis for melatonin suppression by polychromatic light and therefore also not for white light. The fundamental reason for this discrepancy between mono- and polychromatic light is retinal neural wiring as described in Sect. 5.2.2. The rods and cones are connected through the neural wiring with the pRGC cells. Neural wiring results in a contribution of rod and cone signals to the (dominating) signal created in the pRGC itself. Rea et al. (2005, 2010, 2012) published a theoretical circadian phototransduction model based on a possible mechanism of interaction of cones and rods with pRGCs

5.5 Spectral Sensitivity Fig. 5.14 Melatonin suppression action spectra based on the hypothetical model of Rea et al. (2010) for monochromatic and polychromatic light together with test results for monochromatic light of Brainard et al. (2001) and Thapan et al. (2001)

155

Relative Sensitivity (%)

Monochromatic

100

Polychromatic Brainard

80

Thapan 60 40 20 0 -20 400

450

500

550

600

650

700

λ (nm)

that fits well with the monochromatic test results of Brainard and Thapan. The term phototransduction stands for converting light into electrical signals. From that model, two separate action spectrum curves for melatonin suppression are derived: one for monochromatic and another for polychromatic light. Figure 5.14 shows them together with the monochromatic test results of Brainard and Thapan (adapted by Rea et al. (2005) for a 35% suppression criterion). As is evident by comparing the monochromatic curve of the model with the test results of Brainard and Thapan, the model predicts well the sharp discontinuity around 510 nm. The opposing effect of long wavelengths in polychromatic light on melatonin suppression is evident from the fact that part of the polychromatic curve has negative values. Nocturnal melatonin suppression by polychromatic light of two different spectra was consistent with the prediction from the model (Figueiro et al. 2006). A next step proposed by the authors is converting photopic lighting levels into values of circadian stimulus, CS. Circadian stimulus is a measure of the effectiveness of retinal light for the human circadian system from threshold (CS ¼ 0.1) to saturation (CS ¼ 0.7). Figure 5.15 shows calculated circadian stimulus values (CS) based on the model, depending on photopic illuminance values (on eye level) for daylight (D65); white phosphor LEDs with correlated colour temperatures of approximately 2800, 4000 and 6500 K; and a blue and red LED. With daylight or LED light with a CCT of 6500 K, the upper knee of the S-shaped curve corresponds roughly with an illuminance level on the eye between 500 and 1000 lux. For the CCT range of 2700–4000 K, the corresponding illuminances are some 1000–2000 lux. This range represents, therefore, the minimum lighting level needed for an adequate circadian stimulus. Surprisingly, the white light source with a CCT of 4030 K performance is less effective in creating circadian stimuli than the LED with a lower CCT of 2790 K. All white LEDs and fluorescent lamp light sources with a correlated colour temperature of approximately 3500–4500 K perform less effectively than sources with a lower colour temperature of approx. 2800–3000 K. The reason may be that the model makes a distinction between

156

5 Non-visual Biological Mechanism

Circadian Stimulus 0.7 0.6 LED Blue

0.5

LED 6470 K LED 4030 K

0.5

LED 2790 K

D65

0.3 0.2 LED Red

0.1 0 10

20

50

100

200

500

1000

2000

5000

10000

Eeye (lux) Fig. 5.15 Calculated circadian stimulus CS as a function of photopic illuminance for daylight (D65), white phosphor LED sources of three different correlated colour temperatures (all with a Ra value of at least 80) and a red and blue LED source (peak wavelength 630 nm and 460 nm, respectively). Calculated from the phototransduction model of Rea et al. (2005, 2010, 2012) with their institution’s circadian stimulus calculator

“warm” and “cool” light sources. It could lead to a discontinuity. Possibly, such a distinction should not only be made on the basis of the spectrum but also on the basis of lighting level. As discussed in Chap. 1, photosensitive cells in the retina have threshold values of light exposure below or above which they are active. There is not yet a consensus about the general correctness of the model. The institution of the authors (Lighting Research Centre at Rensselaer Polytechnic Institute) made a free Excel sheet available (Circadian Stimulus Calculator) for the calculation of CS for user-supplied light source spectra in dependence of illuminance at eye level.

5.5.1.3

Single Non-visual Biological Action Spectrum?

As stated before, different non-visual biological actions do respond differently to light. That means, for example, that an action spectrum for nocturnal melatonin suppression cannot be used for accurate predictions of other actions such as cortisol changes by light or for phase-shifting effects of light in jetlag or shift work situations. As an illustration, Fig. 5.16 shows a research result of the effect of blue and red light at night on melatonin and cortisol concentration in test persons (Figueiro and Rea 2010). The baseline measurements of darkness during the whole night are presented in the columns “dark”. Blue (LED) light lowers melatonin concentrations, as to be expected from the spectral sensitivities for melatonin suppression given earlier. Red (LED) light has hardly any effect on melatonin suppression, again as expected. As

5.5 Spectral Sensitivity Fig. 5.16 Effects of blue and of red light at night (40 lux at the eye during 1 h after 3 h of darkness) on levels of melatonin and cortisol (mean normalised concentration). Source: Figueiro and Rea (2010)

157

12

12

Melatonin

Cortisol

8

8

4

4

0

dark blue red

0

dark blue red

far as cortisol production is concerned, blue light also has a biological action: it increases the concentration of cortisol. Red light has a significant biological effect too regarding cortisol production (an increase of cortisol concentration), although it has no or hardly any biological effect regarding melatonin suppression. The use of orange-white light (white light from which the blue part is filtered) during the night shift has also been studied, as will be discussed in more detail in Chap. 7. These studies show that this orange-white light does not or hardly suppress melatonin at night and it does not have a significant phase-shifting effect. The same studies also showed that night-time alertness with such light was mostly higher than in the dim-light situation. These studies suggest that the melatonin suppression action of light has a different spectral sensitivity than the cortisol action effect, again an indication that a single spectral sensitivity function relevant for all non-visual biological effects of light does not exist. There are still other reasons for this phenomenon. The amount, duration and even previous history of light exposure affect the spectral and absolute sensitivity for many non-visual biological actions (Chang et al. 2011; CIE 2015a; Van Diepen et al. 2013). The level of light may influence the respective contributions of the different photoreceptor types. At low lighting levels, the relative contribution of rods may, for example, be larger than at higher levels. Some biological responses may require light over a longer time than other types of responses. The light condition experienced over the past day and even during the past hours influences the actual response to light. Brighter light or more bluish-white light in the morning or early evening lowers the response to light during the night, for example, regarding melatonin suppression or phase shift (Hébert et al. 2002; Jasser et al. 2006; Chang et al. 2011; Zeitzer et al. 2011; Kozaki et al. 2016).

5.5.2

Spectral Sensitivity of Photopigments

Since the early 2000s, it is recognised that melanopsin is the photopigment that gives the pRGC cell its intrinsic photosensitivity (Gooley et al. 2001; Hankins and Lucas 2002; Hattar et al. 2002; Newman et al. 2003; Brainard and Hanifin 2004). It is an opsin type of photopigment different from the rod opsin as well as different from the

158

5 Non-visual Biological Mechanism

three types of cone opsins. All these different opsin types of pigments absorb different wavelengths. They, therefore, result in different spectral sensitivities for the different kinds of photocells. The absorption spectra of the cones and rods are already long time known. Today, also the absorption spectrum of melanopsin has become available (Dacey et al. 2005; Peirson and Foster 2006; Gamlin et al. 2007; Bailes and Lucas 2013; Lucas et al. 2014). CIE defines the spectral melanopic sensitivity based on measurements of the absorption spectrum of the photopigment melanopsin in its international standard 026:2018 (CIE 2018). It determines the intrinsic sensitivity of the pRGC cells without the interaction effect of rods and cones through neural wiring. Figure 5.17 shows the melanopic spectral sensitivity curve of CIE standard 026:2018 (full drawn, light blue curve). The maximum melanopic sensitivity is reached at a wavelength of approximately 480 nm. Figure 5.17 also shows the sensitivity curves for the S-, M- and L-cones and rods, based on their specific photopigments (dashed curves, taken from Fig. 1.5 of Chap. 1). With these curves, it is possible to calculate for a specific lamp spectrum and a given amount of light on the outer surface of the eye the spectrally weighted irradiance for the five human photopigments. Irradiance is the power of radiation received by a surface divided by the area of that surface. It is expressed in W/m2. Fig. 5.17 Relative spectral sensitivity curve for the intrinsic photosensitivity of pRGCs for light at the outer surface of the eye (light blue drawn melanopic curve) together with the curves of rods, S-, M- and L-cone basis: 32-year, 10
standard observer. Bottom figure in logarithmic units. Source: CIE (2018)

Relative Sensitivity

pRGC S -cone

1.0

M-cone

0.8

L -cone 0.6

rod

0.4 0.2 0 400

450

500

550

600

650

450

500

550

600

650

700

λ (nm)

1

0.1

0.01 400

700

λ (nm)

5.5 Spectral Sensitivity

159

Table 5.2 Names of the five types of α-opic irradiances and the photopigments after which they are named (CIE 2015a) Photoreceptor pRGC Rod S-cone M-cone L-cone

Photopigment Melanopsin Rhodopsin Photopsin cyanolabe Photopsin chlorolabe Photopsin erythrolabe

α-Opic irradiance Melanopic irradiance Rhodopic irradiance Cyanopic irradiance Chloropic irradiance Erythropic irradiance

Symbol Ee,mel Ee,rod Ee,sc Ee,mc Ee,lc

Unit W/m2 W/m2 W/m2 W/m2 W/m2

The symbol for irradiance, Ee, is written with the subscript “e” to avoid confusion with the symbol E used for the photometric unit illuminance (i.e. V(λ) weighted irradiance expressed in lux). The spectrally weighted irradiance, for the example of melanopsin, is obtained from the formula Z E e, mel ¼

E e ðλÞ Smel ðλÞ dðλÞ

where Ee,mel ¼ melanopsin-weighted irradiance in W/m2 Ee(λ) ¼ irradiance for wavelength λ in W/m2 (available from the lamp spectrum) Smel(λ) ¼ relative melanopsin sensitivity for wavelength λ, where Smel(λ) is normalised to the value of 1 at its peak (given in Fig. 5.17) The collective name for the five spectral weighted irradiances is “α-opic irradiance”. Note that the symbol “α” is often used as the symbol for the absorption coefficient. As such, the symbol “α” is appropriate as the prefix for the spectral weighted irradiances resulting from the absorption of light by the five different photoreceptor pigments. Each of the five individual α-opic irradiances is named after its photopigment name, melanopic (pRGC), rhodopic (rod), cyanopic (S-cone), chloropic (M-cone) and erythropic (L-cone) irradiances, as indicated in Table 5.2. Often the cone irradiances are simply referred to as S-, M- and L-cone irradiances. Figure 5.18 gives for some typical lamp types the calculated α-opic irradiances for the five photoreceptors, for the condition 1000 lux at the outer surface of the eye. The values are calculated based on the formula given above for the example of melanopic irradiance. For the condition of 1000 lux at the outer surface of the eye the general formula is Z E e, α per 1000 lux ¼

Z Ee ðλÞ Sα ðλÞ dðλÞ  1000=683

E e ðλÞ V ðλÞ d ðλÞ

where Ee,α ¼ α-opic-weighted irradiance in W/m2; α to be specified as melanopic, rhodopic, S-cone, M-cone or L-cone-opic

160

5 Non-visual Biological Mechanism

Incandescent 2810 K Ra= 100

Fluorescent 4000 K Ra= 83

Daylight CIE D65 6500 K

100

100

100

%

%

%

50

50

50

0

0

0

400 450 500 550 600 650 700 λ (nm)

400 450 500 550 600 650 700

Irradiance per 1000 lux

Irradiance per 1000 lux

(μW/cm2)

(μW/cm2)

200

200

DLeq 0.49

150

λ (nm)

150

100

100

50

50

50

0

0

Mel Rod

M

0

S

L

Mel Rod

M

L

S

LED 4030 K Ra= 80

LED 2790 K Ra= 82 100

100

%

%

%

50

50

50

0

0

0

400 450 500 550 600 650 700

400 450 500 550 600 650 700

λ (nm)

(μW/cm2)

(μW/cm2)

DLeq 0.40

200 150

λ (nm)

Irradiance per 1000 lux

150

150

100

100

50

50

0

0

M

0

S

L

Red LED peak 630 nm

Mel Rod

M

L

S

100

100

%

%

%

50

50

50

0

0

0

400 450 500 550 600 650 700

400 450 500 550 600 650 700

λ (nm)

(μW/cm2)

200 150

(μW/cm2)

DLeq 0.00

λ (nm)

Irradiance per 1000 lux 200 150

1500

100

1000

50

500

0

0

M

L

λ (nm)

DLeq 10.6

2000

50

Mel Rod

L

400 450 500 550 600 650 700

100

S

M

Irradiance per 1000 lux (μW/cm2)

DLeq 1.11

Mel Rod

Blue LED peak 460 nm

Green LED peak 520 nm

100

Irradiance per 1000 lux

λ (nm)

DLeq 1.01

200

50

Mel Rod

L

400 450 500 550 600 650 700

Irradiance per 1000 lux

100

S

M

(μW/cm2)

DLeq 0.58

200

Mel Rod

LED 6470 K Ra= 81

100

Irradiance per 1000 lux

λ (nm)

DLeq 1.00

200

100

S

550 600 650 700

(μW/cm2)

DLeq 0.56

150

400 450 500

Irradiance per 1000 lux

200

0

S

Mel Rod

M

L

S

Mel Rod

M

L

Fig. 5.18 Alpha-opic irradiances of different lamp types for 1000 lux at the outer surface of the eye. S (S-cone) is cyanopic irradiance, Mel (melanopsin) is melanopic irradiance, Rod is rhodopic irradiance, M (M-cone) is chloropic irradiance and L (L-cone) is erythropic irradiance. DLeq is the melanopic equivalent daylight ratio (CIE D65). Bars for melanopic irradiances are shown somewhat broader to make them stand out

5.5 Spectral Sensitivity

161

Ee(λ) ¼ irradiance for wavelength λ in W/m2 (available from the lamp spectrum) Sα(λ) ¼ relative α-opic sensitivity for wavelength λ, where Sα(λ) is normalised to the value of 1 at its peak (given in Fig. 5.17) V(λ) ¼ standardised relative spectral sensitivity curves for photopic vision normalised to the value of 1 at its peak (given in Fig. 1.7) These five “irradiances per 1000 lux” give a good insight into the effectiveness of a lamp in evoking a reaction in each of the five photoreceptors. For non-visual biological lighting applications, it may often be interesting to compare the melanopic irradiance of a particular light source with the melanopic irradiance by daylight (of the same lighting level at the eye). For this purpose, Fig. 5.17 also gives, for each lamp type, the “melanopic equivalent daylight ratio DLeq”. It is the ratio of the melanopic irradiance of the lamp to the melanopic irradiance of 6500 K daylight (CIE standard D65 sky). This ratio is also referred to as melanopic daylight (D65) efficacy ratio (CIE 2018). The first two rows of Fig. 5.18 give the results for white-light sources, all with a colour-rendering Ra of at least 80. The first column shows light sources with a correlated colour temperature CCT of ca. 2800 K (incandescent lamp and LED), the second column of ca. 4000 K (fluorescent and LED) and the last column of ca. 6500 K (D65 daylight and LED). When comparing differences in irradiances for the same CCT value the largest difference occurs for the 2800 K case. Contrary to what is often, wrongly, expected, the incandescent lamp has a larger melanopic irradiance (in this example 63 μW/cm2) than the LED lamp with a closely same CCT value (52 μW/cm2 or 17% less). The conclusion of this example is in line with other studies (van Bommel 2010; Kobav and Bizjak 2012). This is, provided that the illuminance condition at the eye is the same (1000 lux in the above example). When comparing the melanopic irradiances for the different correlated colour temperature ranges, it is clear that with higher CCT values, there is a clear trend for higher melanopic irradiances. Popularly phrased: A larger blue content in white light evokes a larger effect in the melanopsin photopigment. The last row of Fig. 5.18 shows the results for coloured, narrow-spectrum LEDs, respectively, for red, green and blue. Note that the scale of the blue LED is different: it ranges from 0 to 2000 μW/cm2 while for all other lamps the range is from 0 to 200 μW/cm2. The blue LED results in much higher irradiances for all photoreceptors. This is, in particular, the case for melanopic and cyanopic (S-cone) irradiances. The red LED provides no cyanopic and hardly any melanopic irradiance. These largely contrasting properties of especially red and blue LEDs are important in therapeutic light applications. At first sight, it may seem strange that the blue LED also results in M- and S-cone irradiances. The very reason is the fact that the M and S-cone spectral sensitivities, having their peak in the green and red parts of the wavelength range, respectively, also reach into the blue part of the wavelength range as is evident from Fig. 5.17. The melanopic equivalent daylight ratio (D65) of a light source can also be used to express any photometric quantity (as, e.g., luminous flux, luminous intensity, luminance or illuminance) in a daylight (D65)-equivalent quantity. Examples for the 2790, 4030 and 6470 K white LED lamps of Fig. 5.18 could look like the following:

162

5 Non-visual Biological Mechanism

• A 2790 K LED lamp with a melanopic equivalent daylight (D65) ratio of 0.40 and a luminous flux of 500 lm has a melanopic equivalent daylight (D65) luminous flux of 0.40  500 ¼ 200 lm. • A 4030 K LED lighting installation with a melanopic equivalent daylight (D65) ratio of 0.58, providing an illuminance of 600 lux, has a melanopic equivalent daylight (D65) illuminance (Emel,D65eq) of 0.58  600 ¼ 348 lux. • A 6470 K LED lighting installation with a melanopic equivalent daylight (D65) ratio of 1.01, providing an illuminance of 600 lux, has a melanopic equivalent daylight (D65) illuminance of 1.01  600 ¼ 606 lux.

5.5.3

Spectral Characterisation of Lighting Installations

In some publications about non-visual biological effects of lighting, lighting installations are sometimes described by illuminances only. This is insufficient as is illustrated in Fig. 5.19, left, with the example of melatonin suppression (Giménez et al. 2016). Melatonin suppression percentages obtained with a large number of different lamp spectra are shown. The graph on the left gives the suppression values as a function of the photopic illuminance level (lux values). It is evident that there hardly exists a correlation. In contrast with this, the graph on the right, where for the same lamp spectra the suppression values are given as a function of melanopic irradiances (μW/cm2 values), shows a much better correlation. This indeed demonstrates that characterising lighting installations with illuminances only is not sufficient. The absolute minimum requirement is to use together with illuminance levels also the correlated colour temperature of the light source. It is much better to use instead of Melatonin suppression

Melatonin suppression

100

100

%

%

80

80

60

60

40

40

20

20

0

0

-20 0.1

1

100

Illuminance (lux)

-20 10000

0.01

0.1

10

1000

Melanopic irradiance (μW/cm2)

Fig. 5.19 Melatonin suppression for a large number of different lamps as a function of photopic illuminance (left) and melanopic irradiance right. Source: Giménez et al. (2016)

References

163

the colour temperature the melanopic equivalent daylight ratio (D65). An alternative is to use melanopic equivalent daylight (D65) photometric illuminance as described at the end of the previous section. For research purposes it is strongly recommended to record with the illuminance values also all five α-opic irradiances. They are a good indication of which photoreceptors are evoked to what extent by the actual lamp type. When perhaps in the future more becomes known about the effect of the neural wiring of the photoreceptors, and thus about the interplay of the different photoreceptor types, this information will only become more meaningful. The five irradiances can be recorded as five values or depicted in graphs such as shown in Fig. 5.18. Light source manufacturers should be able to supply alpha-opic irradiances for their lamps per, for example, 1000 lux. CIE has produced an Excel toolbox, freely downloadable, that easily allows for the calculation of these irradiances from the spectral intensity distribution of a lamp (CIE 2015b).

References Aschoff J (1954) Zeitgeber der tierischen Tagesperiodik. Naturwissenschaften 41:49–56 Aschoff J (1965) Circadian rhythms in man—a self-sustained oscillator with an inherent frequency underlies human 24-hour periodicity. Science 148:1427–1432 Aschoff J (1981) Handbook of behavioral neurobiology, Biological rhythms. Plenum Press, New York Atamian HS, Creux NM, Brown EA, Garner AG, Blackman BK, Harmer SL (2016) Circadian regulation of sunflower heliotropism, floral orientation, and pollinator visits. Science 353:587–590 Bailes HJ, Lucas RJ (2013) Human melanopsin forms a pigment maximally sensitive to blue light (lambdamax {approx} 479 nm) supporting activation of Gq/11 and Gi/o signalling cascades. Proc R Soc B Biol Sci 280(1759):20122987 Balsalobre A, Brown SA, Marcacci L, Tronche F, Kellendonk C, Reichardt HM, Schütz G, Schible U (2000) Resetting of circadian time in peripheral tissues by glucocorticoid signalling. Science 289:2344–2347 Bargiello TA, Jackson FR, Young MW (1984) Restoration of circadian behavioural rhythms by gene transfer in Drosophila. Nature 312:752–754 Berson DM (2003) Strange vision: ganglion cells as circadian photoreceptors. Trends Neurosci 26:314–320 Berson DM, Dunn FA, Takao M (2002) Phototransduction by retinal ganglion cells that set the circadian clock. Science 295:1070–1073 Boivin DB, Czeisler CA (1998) Resetting of circadian melatonin and cortisol rhythms in humans by ordinary room light. NeuroReport 9:779–782 Brainard GC, Hanifin JP (2004) The effects of light on human health and behavior: relevance to architectural lighting. In: CIE x027:2004 Proceedings of CIE symposium ’04 light and health: non-visual effects, pp 2–16 Brainard GC, Hanifin JP, Greeson JM, Byrne B, Glickman G, Gerner E, Rollag MD (2001) Action spectrum for melatonin regulation in humans: evidence for a novel circadian photoreceptor. J Neurosci 21(16):6405–6412 Broszio K, Knoop M, Niedling M, Völker S (2017) Effective radiant flux for non-image forming effects—is the illuminance and the melanopic irradiance at the eye really the right measure? In: Proceedings Lux Europe, Ljubljana, pp 18–20

164

5 Non-visual Biological Mechanism

Brown SA, Fleury-Olela F, Nagoshi E, Hauser C, Juge C, Meier CA, Chicheportiche R, Dayer JM, Albrecht U, Schibler U (2005) The period length of fibroblast circadian gene expression varies widely among human individuals. PLoS Biol 3(10):e338 Cajochen C (2007) Alerting effects of light. Sleep Med Rev 11:453–464 Cassone VM, Speh JC, Card JP, Moore RY (1988) Comparative anatomy of the mammalian hypothalamic suprachiasmatic nucleus. J Biol Rhythms 3:71–91 Chang AM, Scheer FAJL, Czeisler CA (2011) The human circadian system adapts to prior photic history. J Physiol 589(5):1095–1102 Chellappa SL, Gordijn MCM, Cajochen C (2011) Can light make us bright? Effects of light on cognition sleep. In: Kerkhof G, van Dongen HPA (eds) Progress in brain research. Elsevier, Amsterdam, p 190 Chen SK, Badea TC, Hattar S (2011) Photoentrainment and pupillary light reflex are mediated by distinct populations of ipRGCs. Nature 476:92–95 Chovnick A (1960) Proceedings of Cold Spring Harbor symposium on quantitative biology. In: Biological clocks 25:117–514 CIE (2004) International Commission on Illumination CIE Publication 158:2004, Ocular lighting effects on human physiology and behaviour. Vienna CIE (2015a) International Commission on Illumination CIE Technical Note 003:2015, Report on the first international workshop on circadian and neurophysiological photometry, 2013. Vienna CIE (2015b) Irradiance toolbox. http://files.cie.co.at/784_TN003_Toolbox.xls CIE (2016) International Commission on Illumination CIE DIS 017:2016 (term 17-29-030). CIE Draft international standard, ILV: International lighting vocabulary. Vienna CIE (2018) International Commission on Illumination CIE International Standard CIE 026:2018. CIE system for metrology of optical radiation for ipRGC-influenced responses to light. Vienna Czeisler CA, Duffy JF, Shanahan TL, Brown EN, Mitchell JF, Rimmer DW, Ronda JM, Silva EJ, Allan JS, Emens JS, Dijk DJ, Kronauer RE (1999) Stability, precision, and near-24-hour period of the human circadian pacemaker. Science 284(5423):2177–2181 Dacey DM, Liao HW, Peterson BB, Robinson FR, Smith VC, Pokorny J, Yau KW, Gamlin PD (2005) Melanopsin-expressing ganglion cells in primate retina signal colour and irradiance and project to the LGN. Nature 433(7027):749–754 De Candolle A (1832) La Physiologie végégetal. Béchet Jeune, Paris De Mairan JJO (1729) Observation botanique. Hist. de l’Acad. Royal Sciences, Paris, p 1 Dickmeis T (2009) Glucocorticoids and the circadian clock. J Endocrinol 200:3–22 Do MTH, Kang SH, Xue T, Zhong H, Liao HW, Bergles DE, Yau KW (2009) Photon capture and signalling by melanopsin retinal ganglion cells. Nature 457:281–287 Duffy JF, Cain SW, Chang AM, Phillips AJK, Münch MY, Gronfier C, Wyatt JK, Dijk DJ, Wright KP, Czeisler CA (2011) Sex difference in the near-24-hour intrinsic period of the human circadian timing system. Proc Natl Acad Sci USA 108(Suppl 3):15602–15608 Dunlap JC, Loros JJ, Aronson BD, Merrow M, Crosthwaite S, Bell-Pedersen D, Johnson K, Lindgren K, Garceau NY (1995) The genetic basis of the circadian clock identification of frq and FRQ as clock components. In: Circadian clocks and their adjustment, Ciba Foundation Symposium 183. Wiley, Chichester Eastman CI, Burgess HJ (2009) How to travel the world without jet lag. Sleep Med Clin 4 (2):241–255 Emery P, So WV, Kaneko M, Hall JC, Rosbach M (1998) CRY, a Drosophila clock and lightregulated cryptochrome, is a major contributor to circadian rhythm resetting and photosensitivity. Cell 95:669–679 Figueiro MG, Rea MS (2010) The effects of red and blue lights on circadian variations in cortisol, alpha amylase, and melatonin. Int J Endocrinol 2010:1–9 Figueiro MG, Rea MS, Bullough JD (2006) Circadian effectiveness of two polychromatic lights in suppressing human nocturnal melatonin. Neurosci Lett 406:293–297 Foster R, Kreitzman L (2004) Rhythms of life. The biological clocks that control the daily lives of every living thing. Profile Books Ltd., London Foster RG, Provencio I, Hudson D, Fiske S, De Grip W, Menaker M (1991) Circadian photoreception in the retinally degenerate mouse (rd/rd). J Comp Physiol A 169(1):39–50

References

165

Freedman MS, Lucas RJ, Soni B, von Schantz M, Munoz M, David-Gray Z, Foster R (1999) Regulation of mammalian circadian behavior by non-rod, non-cone, ocular photoreceptors. Science 284:502–504 Gamlin PD, McDougal DH, Pokorny J, Smith VC, Yau KW, Dacey DM (2007) Human and macaque pupil responses driven by melanopsin-containing retinal ganglion cells. Vision Res 47 (7):946–954 Giménez M, Schlangen L, Lang D, Beersma D, Novotny P, Plischke H, Wulff K, Linek M, Cajochen C, Löffler J, Lasauskaite R, Bhusal P, Halonen L (2016) D3.7 Report on metric to quantify biological light exposure doses. Accelerate SSL Innovation for Europe. SSL-erate Consortium Glickman G, Hanifin JP, Rollag MD, Wang H, Cooper H, Brainard GC (2003) Inferior retinal light exposure is more effective than superior retinal exposure in suppressing melatonin in humans. J Biol Rhythms 18:71–79 Gooley JJ, Lu J, Chou TC, Scammell TE, Saper CB (2001) Melanopsin in cells of origin of the retinohypothalamic tract. Nat Neurosci 4(12):1165 Hankins MW, Lucas RJ (2002) The primary visual pathway in humans is regulated according to long-term light exposure through the action of a nonclassical photopigment. Curr Biol 12 (3):191–198 Hattar S, Liao HW, Takao M, Berson DM, Yau KW (2002) Melanopsin-containing retinal ganglion cells: architecture, projections, and intrinsic photosensitivity. Science 295:1065–1070 Hattar S, Lucas RJ, Mrosovsky N, Thompson S, Douglas RH, Hankins MW, Lem J, Biel M, Hofmann F, Foster RG, Yau KW (2003) Melanopsin and rod–cone photoreceptive systems account for all major accessory visual functions in mice. Nature 424(6944):75–81 Hébert M, Martin SK, Lee C, Eastman CI (2002) The effects of prior light history on the suppression of melatonin by light in humans. J Pineal Res 33(4):198–203 Hoyle NP, Seinkmane E, Putker M, Feeney KA, Krogager TP, Chesham JE, Bray LK, Thomas JM, Dunn K, Blaikley J, O’Neil JS (2017) Circadian actin dynamics drive rhythmic fibroblast mobilization during wound healing. Sci Transl Med 9(415):eaal2774 Jasser SA, Hanifin JP, Rollag MD, Brainard GC (2006) Dim light adaptation attenuates acute melatonin suppression in humans. J Biol Rhythms 21(5):394–404 Jones BE (2003) Arousal systems. Front Biosci 8:S438–S451 Kalsbeek A, Buijs RM (2002) Output pathways of the mammalian suprachiasmatic nucleus: coding circadian time by transmitter selection and specific targeting. Cell Tissue Res 309:109–118 Kaur G, Phillips C, Wong K, Saini B (2013) Timing is important in medication administration: a timely review of chronotherapy research. Int J Clin Pharm 35(3):344–358 Klein DC, Moore RY, Reppert SM (1991) Suprachiasmatic nucleus: the mind’s clock. Oxford University Press, Oxford Klerman EB, Rimmer DW, Dijk D-J, Kronauer RE, Rizzo JF III, Czeisler CA (1998) Nonphotic entrainment of the human circadian pacemaker. Am J Physiol 274:R991–R996 Knoop M (2018) Opinion: studies on non-image forming effects—lighting cold cases? Lighting Res Technol 50:496 Kobav MB, Bizjak G (2012) LED spectra and melatonin suppression action function. Light Eng 20 (3):15–22 Kozaki T, Kubokawa A, Taketomi R, Hatae K (2016) Light-induced melatonin suppression at night after exposure to different wavelength composition of morning light. Neurosci Lett 616:1–4 Kumbalasiri T, Provencio I (2005) Melanopsin and other novel mammalian opsins. Exp Eye Res 81:368–375 Lasko TA, Kripke DF, Elliot JA (1999) Melatonin suppression by illumination of upper and lower visual fields. J Biol Rhythms 14:122–125 Lerner AB, Case JD, Takahashi Y, Lee TH, Mori W (1958) Isolation of melatonin, the pineal gland factor that lightens melanocytes. J Am Chem Soc 80(10):2587

166

5 Non-visual Biological Mechanism

Lockley SW, Evans EE, Scheer FAJL, Brainard GC, Czeisler CA et al (2006) Short-wavelength sensitivity for the direct effects of light on alertness, vigilance, and the waking electroencephalogram in humans. Sleep 29:161–168 Lowden A, Äkerstedt T, Wibom R (2004) Suppression of sleepiness and melatonin by bright light exposure during breaks in night work. J Sleep Res 13:37–43 Lucas RJ, Hattar S, Takao M, Berson DM, Foster RG, Yau KW (2003) Diminished pupillary light reflex at high irradiances in melanopsin-knockout mice. Science 299(5604):245–247 Lucas RJ, Peirson SN, Berson SN, Brown TM, Cooper HM, Czeisler CA, Figueiro MG, Gamlin PD, Lockley SW, O’Hagan JB, Price LLA, Provencio I, Skene DJ, Brainard GC (2014) Measuring and using light in the melanopsin age. Trends Neurosci 37(1):1–9 McIntyre IM, Norman TR, Burrows GD, Armstrong SM (1989a) Human melatonin suppression by light is intensity dependent. J Pineal Res 6(2):149–156 McIntyre IM, Norman TR, Burrows GD, Armstrong SM (1989b) Quantal melatonin suppression by exposure to low intensity light in man. Life Sci 45(4):327–332 Moore RY, Eichler VB (1972) Loss of a circadian adrenal corticosterone rhythm following suprachiasmatic lesions in the rat. Brain Res 42:201–206 Moore RY, Heller A, Wurtman RJ, Axelrod J (1967) Visual pathway mediating pineal response to environmental light. Science 155(3759):220–223 Moore RY, Speh JC, Card JP (1995) The retinohypothalamic tract originates from a distinct subset of retinal ganglion cells. J Comp Neurol 352(3):351–366 Morin LP (1994) The circadian visual system. Brain Res Rev 19:102–127 Mure LS, Rieux C, Hattar S, Cooper HM (2007) Melanopsin-dependent nonvisual responses: evidence for photopigment bistability in vivo. J Biol Rhythms 5:411–424 Newman LA, Walker MT, Brown RL, Cronin TW, Robinson PR (2003) Melanopsin forms a functional short wavelength photopigment. Biochemistry 42:12734–12738 Panda S, Provencio I, Tu DC, Pires SS, Rollag MD, Castrucci AM, Pletcher MT, Sato TK, Wiltshire T, Andahazy M, Kay SA, Van Gelder RN, Hogenesch JB (2003) Melanopsin Is Required for Non-Image-Forming Photic Responses in Blind Mice. Science 301(5632):525–527 Pauers MJ, Kuchenbecker JA, Neitz M, Neitz J (2012) Changes in the colour of light cue circadian activity. Anim Behav 83:1143–1151 Peirson S, Foster RG (2006) Melanopsin: another way of signalling light. Neuron 49(3):331–339 Phipps-Nelson J, Redman JR, Dijk DJ, Rajaratnam SM (2003) Daytime exposure to bright light, as compared to dim light, decreases sleepiness and improves psychomotor, vigilance performance. Sleep 26:695–700 Piazena H, Franke L, Thomsen B, Kamenzky I, Uebelhack R, Völker S (2014) Melatoninsuppression mit Weiβlicht-LEDs—erste Ergebnisse. In: Proceedings of the 8th symposium Licht und Gesundheit, Berlin, pp 39–52 Pittendrigh CS, Bruce VG (1957) An oscillator model for biological clocks. In: Rudnick D (ed) Rhythmic and synthetic processes in growth. Princeton University Press, Princeton, NJ, pp 239–268 Provencio I, Rodriguez IR, Jiang GS, Hayes WP, Moreira EF, Rollag MD (2000) A novel human opsin in the inner retina. J Neurosci 20:600–605 Rautkylä E, Puolakka M, Halonen L (2012) Alerting effects of daytime light exposure—a proposed link between light exposure and brain mechanisms. Lighting Res Technol 44:238–252 Rea MS (2002) Light—much more then vision. In: Light and human health: EPRI/LRO 5th international lighting research symposium, Palo Alto, pp 1–15 Rea MS, Bullough JD, Figueiro MG, Bierman A (2004) Spectral opponency in human circadian phototransduction: implications for lighting practice. In: Proceedings of CIE symposium lighting & health, Vienna, pp 111–115 Rea MS, Figueiro MG, Bullough JD, Bierman A (2005) A model of phototransduction by the human circadian system. Brain Res Rev 50:213–228 Rea MS, Figueiro MG, Bierman A, Bullough JD (2010) Circadian light. J Circ Rhythms 8(2):1–10

References

167

Rea MS, Figueiro MG, Bierman A, Hamner MS (2012) Modelling the spectral sensitivity of the human circadian system. Lighting Res Technol 44:386–396 Refinetti R (2015) Comparison of light, food, and temperature as environmental synchronizers of the circadian rhythm of activity in mice. J Physiol Sci 65:359–366 Revell VL, Molina TA, Eastman CI (2012) Human phase response curve to intermittent blue light using a commercially available device. J Physiol 590(19):4859–4868 Richter C (1967) Psychopathology of periodic behavior in animals and man. In: Zubin J, Hunt HF (eds) Comparative psychopathology, vol 205. Grune & Stratton, New York, p 227 Roenneberg T, Wirz-Justice A, Merrow M (2003) Life between clocks: daily temporal patterns of human chronotypes. J Biol Rhythms 18(1):80–90 Roenneberg T, Kuehnle T, Pramstaller PP, Ricken J, Havel M, Guth A, Merrow M (2004) A marker for the end of adolescence. Curr Biol 14:R1038-R Roenneberg T, Kuehnle T, Juda M, Kantermann T, Allebrandt K, Gordijn M, Merrow M (2007) Epidemiology of the human circadian clock. Sleep Med Rev 11:429–438 Roenneberg T, Juda M, Kramer A, Merrow M, Vetter C, Allebrandt KV (2013) Light and the human circadian clock. In: Kramer A, Merrow M (eds) Circadian clocks. Springer, Berlin Rosenthal NE, Joseph-Vanderpool JR, Levendosky AA, Johnston SH, Allen R, Kelly KA, Souetre E, Schultz PM, Starz KE (1990) Phase-shifting effects of bright morning light as treatment for delayed sleep phase syndrome. Sleep 13(4):354–361 Rüger M, Gordijn MCM, Beersma DGM, De Vries B, Daan S (2005) Nasal versus temporal illumination of the human retina: effects on core body temperature, melatonin and circadian phase. J Biol Rhythms 14:122–125 Saper CB, Scammell TE, Lu J (2005) Hypothalamic regulation of sleep and circadian rhythms. Nature 437:1257–1263 Stephan FK, Zucker I (1972) Circadian rhythm in drinking behavior and locomotor activity of rats are eliminated by hypothalamic lesions. Proc Natl Acad Sci USA 69:1583–1586 Takahashi Y, Katsuura T, Shimomura Y, Iwanago K (2011) Prediction model of light-induced melatonin suppression. J Light Vis Environ 35(2):123–135 Teclemariam-Mesbah R, Ter Horst GJ, Postema F, Wortel J, Buijs RM (1999) Anatomical demonstration of the suprachiasmatic nucleus–pineal pathway. J Comp Neurol 406:171–182 Terman M, Lewy AJ, Dijk DJ, Boulos Z, Eastman CI, Campbell SS (1995) Light treatment for sleep disorders: consensus report. IV. Sleep phase and duration disturbances. J Biol Rhythms 10 (2):135–147 Thapan K, Arendt J, Skene DJ (2001) An action spectrum for melatonin suppression: evidence for a novel non-rod, non-cone photoreceptor system in humans. J Physiol 535(1):261–267 Thorn L, Hucklebridge F, Esgate A, Evans P, Clow A (2004) The effect of dawn simulation on the cortisol response to awakening in healthy participants. Psychoneuroendocrinology 29:925–930 Van Bommel WJM (2010) Incandescent replacement lamps and health. LED Prof Rev 19:32–35 Van Diepen HC, Ramkisoensing A, Peirson SN, Foster RG, Meijer JH (2013) Irradiance encoding in the suprachiasmatic nuclei by rod and cone photoreceptors. FASEB J 27:4204–4212 Van Diepen HC, Foster RG, Meijer JH (2015) A colourful clock. PLoS Biol 1002160:1–5 Van Gelder RN, Mawad K (2007) Illuminating the mysteries of melanopsin and circadian photoreception. J Biol Rhythms 5:394–395 Vanderwalle G, Maquet P, Dijk DJ (2009) Light as a modulator of cognitive brain function. Trends Cogn Sci 13(10):429–438 Vandewalle G, Balteau E, Phillips C, Degueldre C, Moreau V, Sterpenich V, Albouy G, Darsaud A, Dessailles M, Dang-Vu TT, Peigneux P, Luxen A, Dijk DJ, Maquet P (2006) Daytime light exposure dynamically enhances brain responses. Curr Biol 16:1616–1621 Visser EK, Beersma DGM, Daan S (1999) Melatonin suppression by light in humans is maximal when the nasal part of the retina is illuminated. J Biol Rhythms 14:116–121 Walmsley L, Hanna L, Mouland J, Martial F, West A, Smedley AR, Bechtold DA, Webb AR, Lucas RJ, Brown TM (2015) Colour as signal for entraining the mammalian circadian clock. PLoS Biol 1002127:1–20

168

5 Non-visual Biological Mechanism

Winget CM, Hughes L, LaDou J (1978) Physiological effects of rotational work shifting: a review. J Occup Med 20(3):204–210 Woelders T, Beersma DGM, Gordijn MCM, Hut RA, Wams EJ (2017) Daily light exposure patterns reveal phase and period of the human circadian clock. J Biol Rhythms 32(3):274–286 Xu W, Van Bommel WJM (2011) Inferior verso superior: inferior retinal light exposure is more effective in pupil contraction in humans. Light Eng 19(2):14–18 Yamazaki S, Numano R, Abe M, Hida A, Takahashi R, Ueda M, Block GD, Sakaki Y, Menaker M, Tei H (2000) Resetting central and peripheral circadian oscillators in transgenic rats. Science 288:682–685 Zeitzer JM, Friedman L, Yesavage JA (2011) Effectiveness of evening phototherapy for insomnia is reduced by bright daytime light exposure. Sleep Med 12(8):805–807 Zulley J, Knab B (2000) Unsere Innere Uhr. Herder, Freiburg im Breisgau

Chapter 6

Light, Sleep, Alertness and Performance

Abstract A classical sleep model is based on an interaction of two different processes. A homeostatic one is characterised by increasing and decreasing sleep pressure after waking up and while asleep, respectively. The other process is a circadian one which provides the possibility to sleep: the sleep window. Light and darkness at the appropriate times strongly influence the latter process. Daytime light influences sleep possibility during the night. Here, both the level and the spectrum of light play a role. Cooler white light is more effective than warmer white light. Sleep quality during the night, of course, also influences alertness and performance during the subsequent day. On top of this effect on alertness and performance, there is also a direct photobiological effect of light on alertness and performance. A sufficient high light level for this second route towards alertness and performance is essential. There are contradictory research results on the role of the spectrum in this respect. On the basis of the research discussed in this chapter, a dynamic lighting scenario for daytime workplaces is proposed which dynamically changes both the lighting level and colour. It optimises between energy requirements on the one hand and requirements of visual and non-visual effects of lighting on the other hand.

6.1

Sleep

Just as air, food and water, sleep is a prime requirement for life. It saves bodily energy by slowing down many processes, and it has a restorative function. It is well established that insufficient sleep and sleep of low quality relate to all-cause mortality (Cappuccio et al. 2010a; Dijk 2012; Odds 2015; Yin et al. 2017). Health problems such as obesity, diabetes, cardiovascular diseases and depressive disorders can be linked with insufficient sleep or bad sleep quality (Cappuccio et al. 2008, 2010b, 2011; Costa et al. 2013; Kim et al. 2013). A recent study with 49 medical doctors, part of whom were on an overnight on-site scheme, demonstrated that disrupted sleep is associated with DNA damage (Cheung et al. 2019). This DNA damage may be causally connected with the increased risk for many of the diseases mentioned above.

© Springer Nature Switzerland AG 2019 W. van Bommel, Interior Lighting, https://doi.org/10.1007/978-3-030-17195-7_6

169

170

6 Light, Sleep, Alertness and Performance

The restorative function of sleep is not only fundamental for our health, but also needed for adequate alertness and performance during the time that we don’t sleep. Sleep also plays a role in storing new information into our long-term memory (Rasch and Born 2013).

6.1.1

Sleep Mechanism

In the early 80s of the last century, cooperation between the universities of Zurich, Basel (Switzerland), and Groningen (the Netherlands) has resulted in a classical model of how sleep is regulated (Borbély and Wirz-Justice 1982; Daan et al. 1984). It is still today the prevalent one (Borbély et al. 2016). The model is based on an interaction of two different processes: a homeostatic and a circadian one. Homeostasis is a process that opposes changes. The homeostatic process in sleep-wake regulation gradually increases sleep pressure (fatigue) after waking up and decreases it during sleep. It is referred to as Process S. Sleep pressure is behaviour dependent, namely of the time being awake and of the previous sleep history. Sleep can occur when the circadian process (called Process C) provides conditions for sleep. Process C is not behaviour dependent but dependent on the state of the body regarding, for example, melatonin and cortisol levels and body temperature. As discussed in the previous chapter, the amount and timing of light and dark influence circadian processes and thus also the sleep Process C. Figure 6.1 (top) shows an example of the typical exponential increasing sleep pressure after waking up together with the steeper, and again exponential, declining sleep pressure during sleep. Figure 6.1 (bottom) shows an example of the circadian curve where point A represents a situation where the conditions are such that sleep is Fig. 6.1 Two-process model of sleep regulation. Homeostatic Process S drives sleep pressure (increasing after awakening and decreasing during sleep). Process C, the circadian process, provides the bodily conditions for sleep between points A and B: the sleep window

Process S

Sleep pressure

8:00

Circadian condition for sleep

16:00

24:00

Process C

08:00

16:00

sleep window

A B

8:00

16:00

24:00

08:00

16:00

6.1 Sleep

171

possible and, further in time, point B where the conditions make sleep again difficult. The time frame between A and B represents the sleep window. The model is a simplification of complicated behaviour and bodily processes of which still much is unknown. A physiological confirmation of the homeostatic process has not yet been identified. Nevertheless, the model permits for the prediction of effects of sleep-wake scenarios and ratings of sleepiness, alertness and performance (Åkerstedt and Folkard 1996; Beersma and Gordijn 2007). Relatively recent research indicates that the two processes influence each other (Achermann and Borbély 2003; Dijk and Archer 2010). Sleep pressure appears to reduce the circadian amplitude, and after sleep deprivation (not having had enough sleep) the biological clock SCN appears to show a reduced response to light (Borbély et al. 2016).

6.1.2

Sleep Measures

Sleep and sleep quality can be objectively evaluated with polysomnography (PSG). PSG measures many body functions such as brain activity (EEG), heart rhythm (ECG), eye movements and muscle activity. The measurements are often used in sleep laboratories, but are difficult to use in field or home experiments. Further in this chapter, one novel study making such measurements in the field is described. Subjective sleep quality can be measured and evaluated with questionnaires. The result of a filled-in questionnaire is a value on a “sleeping” scale that ranges from poor to good sleep quality. Examples are the Pittsburgh Sleep Quality Index, PSQI (Buysse et al. 1989), and the Groningen Sleep Quality Scale, GSQS (Meijman et al. 1988). For the appropriate circumstances for which these measures have been developed, they correspond reasonably well with the results of objective PSG measurements (Westerlund et al. 2016). They, therefore, provide a useful substitute for objective measurements and are often used in studies on the relationship between light and sleep quality. Appendix G gives as an example the Groningen Sleep Quality Scale questionnaire.

6.1.3

Daytime Light

The circadian rhythm, or in the context of the two-process model of sleep, Process C, plays an essential role in regulating the sleep-wake cycle. Sufficient light during daytime and good darkness during night-time are, therefore, prerequisites for a sufficient night-time sleep of good quality. A recent study with 20 test persons showed that light exposure measured throughout the 24 h preceding the sleep period correlates with the objective sleep quality (Wams et al. 2017). This study made, for the first time, use of a novel, mobile PSG unit. It was combined with an action watch with a light sensor. Light levels varied roughly between 10 lux and 30,000 lux. The

172

6 Light, Sleep, Alertness and Performance

Fig. 6.2 Time of the firsttime exposure to the rising daytime lighting in the morning (expressed as the moment of exposure to more than 10 lux) and the average number of awakenings during the subsequent nighttime sleep (Wams et al. 2017)

Awakenings (number per hour)

1.5

1.0

0.5

06:00

10:00

14:00

Time of 1st exposure to > 10 lux results of this investigation suggest not only a fundamental influence of light exposure on the circadian aspect of sleep but also an influence of light exposure on the homeostatic sleep pressure aspect. As the authors of this study state: “ It could change the current understanding of the regulation of sleep”. A clear relationship was found between daytime lighting and subsequently, objectively measured, nighttime sleep quality. As an example Fig. 6.2 shows the relationship between the number of sleep-time awakenings per hour and the time of the first-time exposure in the morning to a lighting level larger than 10 lux. The earlier someone experiences the rising daytime lighting level, the lower the number of subsequent night-time awakenings. This finding fits well with the fact that most people have on average a free-running rhythm of their biological clock of more than 24 h (Sect. 5.1.4). Therefore, their biological clock has to be advanced, what morning light indeed can do (Sect. 5.1.6). Daylight is even on cloudy days, with “only” some 5000 lux in the horizontal field, more than sufficient to keep the biological clock entrained. However, in many climate zones, it is part of the year still dark in the early or mid-morning and again dark in the mid- or late afternoon. Under these circumstances, the electrical lighting at the workplace has to give its beneficial effects.

6.1.3.1

Light Level

Different studies carried out under interior daytime-lighting conditions in working interiors have shown that lighting levels that fulfil the requirements for visual performance and satisfaction do often not fulfil the requirements for sleep (Aries 2005; Viola et al. 2008; Hubalek et al. 2010; Boubekri et al. 2014; Canazei et al. 2014; Figueiro et al. 2017). A study of Aries (2005) illustrates this convincingly. She

6.1 Sleep Fig. 6.3 Subjective nighttime sleep quality as a function of the average daytime vertical illuminance during work hours at the working place. Electrical lighting used 3000 and 4000 K fluorescent tubes (Aries 2005)

173

Subjective Sleep Quality Good

Bad 0

500

1000

1500

2000

Evert, av (lux)

studied the relationship between daytime vertical illuminance at the eye and nighttime sleep quality of 42 office workers in April and May in the Netherlands. She recorded the average vertical illuminance, from daylight and electrical light, during working hours at the workstations of these office workers, spread over ten different office buildings. She determined the subjective sleep quality of each test person over the test period using the Groningen Sleep Quality Scale questionnaire. The results, given in Fig. 6.3, show that lighting during daytime correlates positively with nighttime sleep quality. Figure 6.3 shows that the midpoint of the sleep quality scale corresponds to a value of 400 lux vertical illuminance: the bare minimum required for acceptable sleep quality. With the kind of light distributions and layout of luminaires typically used for office lighting, the average horizontal illuminance at working plane level corresponds to roughly three times the amount of vertical illuminance at eye height. A vertical illuminance of 400 lux thus corresponds to some 1200 lux horizontally, distinctly more than the values required for visual performance and satisfaction as discussed in Chaps. 3 and 4. Hubalek et al. (2010) carried out a similar study. They measured the lighting to which their test persons (daytime office workers) were exposed by having them wearing spectacle frames, with light sensors. In this way, the illuminance at eye level was recorded. Light exposure was measured during the whole period that the test person was not in bed. The participants assessed their sleep quality by answering a simple question about their sleep quality on a 5-point scale. Each of the 23 test persons (average age 38) participated for seven consecutive days, so that also days not spent in the office were included in the tests. The tests were carried out in the months of April and June in Germany. Regression analyses showed that daily light dose (expressed in luxhour), time spent above 1000 lux and time spent above 2500 lux best predicted the outcome measures, among which was the

174

6 Light, Sleep, Alertness and Performance

Fig. 6.4 Subjective sleep quality as a function of total light dose, duration of exposure to more than 1000 lux and more than 2500 lux (on eye level), respectively. Light exposure relates to the total time not spent in bed. Black circles: office day, open circle: non-office day. Graphs are drawn from data from Hubalek et al. (2010)

Subjective Sleep Quality good

bad

10

102 103

1

10 102 103

1

10 102 103

Dose

Duration

Duration

(Luxhr)

> 1000 Lux (min)

> 2500 Lux (min)

subjective sleeping quality. Figure 6.4 shows these results for the average sleep quality of all test persons. Since light exposure until bedtime is included in the analysis, the sleep quality of some participants may have negatively been affected by late evening light. Nevertheless, the graphs show that, overall, with more light and received during a longer time, night-time sleep quality improves considerably. All three measures seem to have an asymptotic value. For the duration of exposure of more than 1000 lux on eye level, this value lies between 500 and 1000 min (8–16 h). For the same measure, a duration of slightly more than 300 min (5 h) corresponds to the midpoint on the subjective sleep quality scale (thin broken line).

6.1.3.2

Light Spectrum

The studies discussed so far used the photopic illuminance as a parameter. Therefore, these studies do not include the effects of the spectrum of the light. Figueiro et al. (2017) used the measure circadian stimulus (CS) as a parameter. As discussed in Sect. 5.5.1 of the previous chapter, circadian stimulus takes the spectral sensitivity based on night-time melatonin suppression into account. Figure 5.15 of Chap. 5 gives for different light sources the relationship between CS and illuminance level on the eye. The participants in Figueiro’s study, all daytime office workers working in five different office buildings, participated both in the summer and winter periods. During the tests, which lasted for seven consecutive days, they did wear when not in bed a so-called daysimeter as a pendant. It has light and activity sensors. The daysimeter is calibrated so that circadian stimulus CS can be determined by postprocessing of the data. The daysimeter worn as a pendant does not measure light at eye level. CS exposures may be some 25% higher at eye level than the levels measured and presented here (Figueiro et al. 2013). While in bed, the daysimeter was worn on the wrist to monitor the sleep-wake activity pattern. The test persons kept a sleep and mood log in which they also recorded the answers to the Pittsburgh Sleep Quality Index (PSQI) questionnaire. The average circadian stimulus was

6.1 Sleep Fig. 6.5 Subjective sleep quality as a function of morning circadian light (between the time of arrival in the office and 12:00 h). We converted the values of the PSQI scale (used in this research) into values corresponding to the 5-point scale used earlier in this book (Figueiro et al. 2017)

175

Subjective Sleep Quality good

R2 = 0.059

bad 0

0.1

0.2

0.3

0.4

0.5

≤ 0.15

≥ 0.30

Morning circadian light (CS) Table 6.1 Illuminances at the eye corresponding to a circadian stimulus value CS ¼ 0.3 for different light sources determined from Fig. 5.15 Light source Daylight D65 White phosphor LED 6500 K White phosphor LED 3500 K

Illuminance (lux) corresponding to CS ¼ 0.3 175 170 400

calculated over the total working hours in the office and, separately, over the morning hours only. The hypothesis that morning light is a better predictor of sleep quality than the whole working-day light was confirmed. Figure 6.5, left, shows the effect of morning CS on subjective sleep quality. The PSQI scale used by the 173 test persons has been converted in the 5-point subjective sleep quality scale used previously in this book. The trend line shows an increase in subjective sleep quality with an increase in morning circadian stimulus. Higher CS values also associate with shorter sleep-onset latency (time in bed before falling asleep). The participants also reported less sleep disturbance with higher CS values. The results show a large spread which is, of course, related to large differences in individual sleep quality. Differences in late evening light exposure between the participants may also have played a role in this. Figure 6.5, right, shows the results for the participants grouped according to lowand high-received circadian stimulus. The participants that received morning CS values higher than 0.3 were grouped in the high group and those with CS values lower than 0.15 in the low group. The subjective sleep quality of the group with CS values larger than 0.3 scores was statistically significantly better. For guidance of the practical meaning of a CS value of 0.3, Table 6.1 gives the corresponding illuminance values at the eye for different light sources. The above studies are examples of proof of the relationship between daytime light exposure and night-time sleep and sleep quality. They also give some information about the required illuminance level at the eye and the effect of the spectrum of the light source to support an acceptable sleep quality. Less quantitative information is

176

6 Light, Sleep, Alertness and Performance

available on the effect of the duration of light exposure. In general, much more research is still needed for a more detailed quantification of the effects of light on sleep quality. With the light sources of today, the minimum requirement for morning light is at least 400 lux (at the eye) for sources with a correlated colour temperature of around 3500 K. With light sources with CCT values of around 6500 K this value can be considerably lower, down to slightly less than 50%.

6.2 6.2.1

Alertness and Performance Alertness and Performance Measures

Sleepiness and alertness can objectively be determined from the brain activity with electroencephalography (EEG). The lower the activity of the frontal brain EEG in the delta (1–4 Hz) and theta (4–8 Hz) ranges and the higher in the alpha (7–14 Hz) range, the higher the alertness of the tested person (Cajochen et al. 1999). Slowrolling eye movements and changes in the eye blink pattern that can be measured with electrodes placed above the eyebrow provide other possibilities of objective measurement of sleepiness and alertness (Cajochen et al. 1999). The result of such measurements is an electro-oculogram (EOG). All these neurophysiological methods are difficult to use in normal working environments, are difficult to employ for a larger group of test persons and may interfere with the way of behaviour of test persons. Alertness can be measured relatively easily by standardised subjective rating scales using questionnaires. Such measurements have shown to give reliable results (Cajochen 2007). One such questionnaire measures mental activity or arousal (Mehrabian and Russell 1974; Bakker et al. 2014). It uses six different categories, each with expressions of opponent feelings, like stimulated–relaxed, excited–calm and wide awake–sleepy. From this, a measure for alertness can be calculated in a 24 to +24 scale. Sleepiness is the reverse of alertness. Therefore alertness can also be measured with questionnaires that ask questions about subjective sleepiness. The 1–9-point Karolinska Sleepiness Scale, KSS, mentioned after the Karolinska medical university in Stockholm, is often used for this purpose. KSS measures the subjective level of sleepiness experienced in the previous 10 min (Åkerstedt and Gillberg 1990). It is a 9-point scale with 1 defined as “extremely alert” and 10 “extremely sleepy, fighting sleep”. KSS correlates well with objective EEG measurements that identify sleepiness (Kaida et al. 2006a). Subjective alertness is also studied by personal assessment of the momentary state of vitality, i.e. the feeling of being energetic. The extremes of vitality can, for example, be described by “feeling extremely energetic” and “feeling completely drained”. Different questionnaires for vitality, using some sub-questions, have been developed and validated by correlating them with more objective methods. Examples are the Lee Visual Assessment Scale for Energy, VAS-e (Lee et al. 1990) and Subjective Vitality Scale, SVS (Ryan and Frederick 1997).

6.2 Alertness and Performance

177

Different types of vigilance tests can be used to measure performance objectively. The Psychomotor Vigilance Test (PVT) is often used. The term psychomotor relates to the fact that the response of the test involves both brain and movement (motor). The test measures how good attention is sustained while giving a physical reaction (like pushing a button) to the appearance of visual stimuli. The stimulus can be as simple as a coloured dot on a dark computer screen. Instead of visual PVT tasks also auditory PVT tasks are used where the test person has to respond to sound stimuli. Often, somewhat more complicated visual tasks that test the mental acuity have been used (Turnage et al. 1992). They include logical reasoning tests where a letter pair, like AB, is presented on a computer screen together with a statement as “A is not followed by B”. The test person has to respond by typing T (for true) or F (for false). The number of correct answers within a defined time frame is the measure of performance. Short-term memory tests are also used. For example, a set of four digits is shown for a short moment on the computer screen followed by a sequence of five digits. The test person has to respond by typing T if the actual digit was part of the original five-digit set or F if it was not. The visual component of the tasks must always be far above the threshold of perception to ensure that reaction time is only influenced by the alertness condition and not by visual effects. Performance tests may last from some 5 min to more than half an hour. A complicating factor with objective performance tests is that different performance tests may have a different sensitivity for differences in lighting, resulting in a different relationship between lighting and measured performance (Boyce et al. 1997; Huiberts et al. 2015a).

6.2.2

Daytime Light

Light affects sleepiness, alertness and performance during daytime through two different routes (Fig. 6.6). Route 1 represents the circadian route and route 2 the direct photobiological one. Route 1 starts at the moment of daytime light on the previous day (“yesterday”). Daytime alertness and performance will obviously be better after a night with sufficient sleep of good quality. Daytime light of yesterday affects the night-time sleep quality of yesterday as discussed in the previous chapter. Therefore, the daytime lighting of yesterday influences sleepiness, alertness and performance of today. Route 2, the direct photobiological route, concerns light that invokes an acute activating effect because of direct photobiological processes as described in Sect. 5.5. Figueiro et al. (2018) provide an extensive literature review on the relation between light, circadian entrainment, acute activation and alertness. Only alertness and performance experiments carried out throughout at least three consecutive days can measure the combined effects of the two routes of the influence of light on alertness and performance. Unfortunately, hardly any such multi-day studies have been made. Quite many one-day studies have been and are being carried out. One-day studies into the effects of daytime lighting on alertness and

178

6 Light, Sleep, Alertness and Performance

Fig. 6.6 The two routes through which daytime light may influence daytime alertness and performance

Yesterday Route 1 (circadian effect) Daytime Light yesterday

Circadian Process

Sleep Pressure Process

Night-time Sleep yesterday

Today Route 2 (direct effect) Daytime Light today

Direct Photobiological effects

Alertness and Performance today

performance that compare the effects of different lighting situations measure the beneficial effects of the second route only, i.e. direct photobiological effects. Their results have to be considered together with the results of dedicated sleep studies described already in Sect. 6.1. Most of the one-day type of investigations on alertness and performance are of the laboratory type that makes use of simulated office environments. They indeed show that bright daytime light exposure affects daytime subjective alertness positively during regular working hours (Gifford et al. 1997; Phipps-Nelson et al. 2003; Kaida et al. 2006b; Rüger et al. 2006; Vandewalle et al. 2006; Leger et al. 2007; Smolders et al. 2012; Canazei et al. 2014; Huiberts et al. 2015b, 2017).

6.2.2.1

Light Level

Küller and Wetterberg (1993) measured objectively, by EEG, the alertness of 33 test persons (age between 19 and 48) in a laboratory made to look like an office environment. Half of the test persons did administrative work under 450 lux desk

6.2 Alertness and Performance Fig. 6.7 Average EEG delta activity of 33 test persons in a simulated office environment with an average desk illuminance of 450 and 1700 lux, respectively. Afternoon measurements. Average for lighting of 3500 and 5500 K (Küller and Wetterberg 1993)

179

EEG delta activity (%) 100 450 lux

1700 lux

450 lux 1700 lux

50

0 left hemisphere

right hemisphere

illuminance (85 cd/m2 wall luminance). They did this one day under lighting of 3500 K CCT, and another day, a week later, of 5500 K. The other half of the group did work under 1700 lux (450 cd/m2 wall luminance) again one day under 3500 K and another day under 5500 K. At both CCT values, the EEG delta activity, measured in the afternoon, was considerably lower under the high lighting level of 1750 lux compared to the low level of 450 lux. As mentioned earlier, delta activity of the EEG is an indicator for sleepiness. It means that the higher lighting level had a pronounced alerting effect. Figure 6.7 shows the averaged relative delta activity values averaged for the two CCT values for both the left and right hemispheres of the brain. Separate values of the delta activity for the two CCT situations are not available from the publication. At the low lighting level of 450 lux, the alpha activity was significantly higher for the 5500 K situation than it was for the 3500 K situation. As mentioned before, a higher alpha activity associates with better alertness. As an example of studies on the relationship between lighting, subjective alertness and objectively measured performance, we take a laboratory study in a simulated office by Smolders et al. (2012). Thirty-two students (age between 18 and 35 years) participated in the study under two different lighting conditions, 200 and 1000 lux at the eye, either in the morning or in the afternoon. Correlated colour temperature CCT was 4000 K. In sessions of 90 min, each participant did different performance tests and completed different questionnaires at different moments of the session. Subjective alertness was obtained from the Karolinska Sleepiness Scale (KKS) questionnaire, subjective vitality from the SVS vitality questionnaire and objective performance from a 5-min auditory Psychomotor Vigilance Test (PVT). Figure 6.8 shows part of the results obtained from this study for the morning and afternoon sessions, respectively, for both the 200 and 1000 lux conditions. Daytime lighting of 1000 lux at the eye has a statistically significant positive effect on subjective alertness and vitality, both in the morning and the afternoon, compared to 200 lux. This also holds for performance (PVT speed) in the morning. These experimental results are in agreement with most of the studies referred to at

180

6 Light, Sleep, Alertness and Performance

Vitality

Alertness 7

Performance

6

200 lux

1000 lux

1000 lux

(PVT speed, 1/sec)

3.7

4

3.6

200 lux

3 5

1000 200 lux lux

1000 lux

1000 lux 200 lux

3.5

3.4

4 2 3

200 lux 1000 lux

200 lux

3.3

2

1 Morning

Afternoon

3.2 Morning

Afternoon

Morning

Afternoon

Fig. 6.8 Subjective alertness (reversal of KKS sleepiness), vitality and reaction speed of a PVT auditory performance test, for a lighting level of 200 lux at the eye and 1000 lux, respectively. CCT ¼ 4000 K. Adapted from Smolders et al. (2012)

the beginning of this section. The afternoon alertness results are also in line with the objective sleepiness EEG measurement results shown in Fig. 6.7. Performance in the afternoon showed to be slightly lower in the 1000 lux situation relative to the 200 lux situation (the difference being statistically not significant). A 2017 study, investigating the effect of season, found similar performance measurement results for the autumn/winter period (Huiberts et al. 2017). However, in spring hardly any difference was found between the 1000 and 200 lux situations, both in the morning and afternoon situations. At the location of this test the daylight length (sunrise to sunset) at the first day of spring, summer, autumn and winter is 12:25, 16:40, 12:15 and 7:50 (h:min), respectively. The seasonal results of this study seem to be an effect of previous lighting history (because of daylight exposure during the periods spent outside the test room; the room itself had no daylight admission).

6.2.2.2

Light Spectrum

Multi-day studies on the effect of blue-rich daytime light on daytime alertness and performance (through the combination of routes 1 and 2 of Fig. 6.6) are hardly carried out. As mentioned before, one-day studies measure only the direct photobiological effects (through route 2 of Fig. 6.6). During daytime, melatonin levels in healthy persons are near zero; therefore, of all direct photobiological aspects discussed in Chap. 5 only the non-melatonin ones can play a role during daytime.

6.3 Dynamic Daytime Lighting Scenario

181

Quite some one-day studies into a possible relation between daytime blue-rich white light and daytime alertness and performance have been carried out. Many show a positive effect of blue-rich light (Noguchi and Sakaguchi 1999; Mills et al. 2007; Viola et al. 2008; Shi et al. 2009; Iskra-Golec et al. 2012; Ferlazzo et al. 2014; Colau and Fotios 2015; Ishii et al. 2018; Ye et al. 2018). A few studies show no or mixed effects (Knez and Kers 2000; Gornicka 2008; Santhi et al. 2013; Smolders and De Kort 2017). Differences in experimental conditions may be reasons for the contradictory results. Think of conditions like the duration and timing of light exposure, timing of the alertness and performance measurement, type of performance task and range of light levels and spectra studied. With regard to light level, it may well be that direct photobiological effects of blue-rich white light show saturation at higher lighting levels. It would mean that blue-rich light has positive effects at lower lighting levels but not at higher lighting levels. The fact remains that with the present state of knowledge it is still uncertain whether or not blue-rich light has positive direct photobiological effects on alertness and performance. Chapter 7 will show that blue light and blue-rich white light at night do have a clear and substantial positive effect on night-time alertness and performance.

6.3

Dynamic Daytime Lighting Scenario

The state of knowledge of today as described in the previous sections indicates that for entrainment of the biological clock and for direct photobiological effects (influencing alertness and performance) the bare minimum illuminance at the eye is some 400–500 lux. This holds for light sources with a correlated colour temperature between approximately 3000 and 4000 K. For higher colour temperatures the required value is lower. For values of around 6500 K, the required value can be halved: some 200–250 lux. Lamps of 3000–4000 K have a melanopic equivalent daylight (D65) factor of some 0.4–0.6 and lamps of 6500 K approximately 1.0 (Fig. 5.18). These eye illuminances, therefore, correspond to melanopic equivalent daylight (D65) illuminances (Eeye,mel,D65eq) of approximately 200–300 lux. The non-visual biological effect of light is not directly governed by the illuminance on the working plane, but by light entering the eye. For the lighting designer, however, it is more convenient to work with illuminances on the horizontal working plane. Converting required illuminances on the eye (for all locations in the space) into an approximate average horizontal illuminance value of the horizontal working plane is thus needed. With the kind of light distributions and layout of luminaires typically used in office and industrial spaces, the average horizontal illuminance at working plane level corresponds to roughly three times the amount of vertical illuminance at eye height. On this basis, Table 6.2 gives the minimum required average horizontal illuminance values on the working plane. In most offices and industrial spaces, these lighting levels are not available from daylight entering the premises for those not working close to windows, especially not in winter time outside the tropical latitudes (Aries 2005; Hubalek et al. 2010;

182

6 Light, Sleep, Alertness and Performance

Table 6.2 Minimum horizontal illuminance on the working plane of office type of spaces CCT 3000–4000 K ca. 6500 K

Eeye (lux) 500 250

Eeye,mel,D65eq (lux) 250 250

Ehor,av (lux) 1500 750

Converted from illuminances at the eye

Ehor (lux) 750 cool white warm white

500

0 08:00

12:30

17:00

Fig. 6.9 Example of a dynamic lighting scenario for human-centric lighting in offices (daytime use). Gradually changing lighting level from 750 to 500 lux (horizontally) and light colour from cool white (approx. 6500 K) to warm white (3000–4000 K) to facilitate both adequate visual and non-visual biological effects (i.e. entrainment of the biological clock, night-time sleep, daytime alertness and performance). The value of 750 lux horizontal illuminance for 6500 K corresponds to roughly 250 lux melanopic equivalent daylight (D65) illuminance. The value of 500 lux for 3000–4000 K corresponds to roughly 85 lux melanopic equivalent (D65) illuminance. Adapted from Van Bommel (2006)

Smolders et al. 2013; Figueiro and Rea 2016). In these situations, artificial lighting should take care of the extra lighting required. A lighting scenario which dynamically changes both the lighting level and colour enables optimisation between energy requirements on the one hand and requirements of visual and non-visual effects on the other hand. Figure 6.9 sketches an example of such a daytime dynamic lighting scenario for a working day starting at 8:00 in the morning and finishing at 17:00. The morning lighting begins with a horizontal illuminance level of 750 lux and a cool white colour (approx. 6500 K). This situation corresponds to roughly 250 lux melanopic equivalent daylight (D65) illuminance. It helps to entrain the biological clock and evokes direct photobiological effects providing alertness and performance potential. This lighting condition lasts for some 2.5–3 h corresponding to the duration of the phase advance peak of the phase–response curve (shown in Fig. 5.8 of Chap. 5). After that, the lighting level gradually decreases, and the colour gradually turns warmer. By the beginning of the afternoon, around lunchtime, the lighting level reaches its lowest level of 500 lux while the colour has its warmest tint (around 3000–4000 K). This situation corresponds to roughly 85 lux melanopic

References

183

equivalent (D65) illuminance. The horizontal lighting level is never lower than 500 lux, being the minimum level recommended for visual performance and comfort as discussed in Chaps. 3 and 4. The midday situation of low lighting level and warm colour supports relaxation around lunchtime. It also does not work against short naps for those who want to have them. Lunchtime power naps have shown to help to feel revitalised and to improve performance for some 2–3 h in the afternoon (Stampi 1992; Takahashi et al. 2004; Mednick and Ehrman 2006; Ficca et al. 2010). These multiple short naps should last for some 3–20 min and be taken well before 15:00 h, the latter in order not to diminish the sleep pressure which is already building up and required for sleep during the night. A short bright-light exposure after lunchtime also has a revitalising effect (Kaida et al. 2006b, 2013). Therefore, a short-lasting (some 30 min) post-lunch rise in light level combined with higher colour temperature is included in the scenario of Fig. 6.9. As discussed, high-level lighting has a lower alerting and performance-improving effect in the afternoon than in the morning. Whether, in the afternoon, high CCT lighting has such an effect at all is doubtful. Therefore, after the post-lunch rise, the lighting level decreases again while the colour temperature gradually returns to 3000–4000 K. The absence of bright cool light in the afternoon also avoids interference with the gradual rise of melatonin later in the afternoon or early evening. The dynamic daytime scenario illustrated in Fig. 6.9 is based on the bare minimum requirements for everyday office tasks. For difficult visual tasks, the minimum lighting level has to be higher than 500 lux according to the relationship between visual performance, size and contrast of that visual task, as discussed in Chap. 3. For challenging tasks with severe consequences in the case of errors, alertness and performance requirements can be extremely high. This, of course, also holds for potentially hazardous conditions in industrial environments. For such tasks or conditions, the lighting level of 750 lux should be increased, and higher correlated colour temperatures than 6500 K could be considered.

References Achermann P, Borbély AA (2003) Mathematical models of sleep regulation. Front Biosci (Landmark eds) 8:S683–S693 Åkerstedt T, Folkard S (1996) Predicting duration of sleep from the three process model of regulation of alertness. Occup Environ Med 53(2):136–141 Åkerstedt T, Gillberg M (1990) Subjective and objective sleepiness in the active individual. Int J Neurosci 52:29–37 Aries M (2005) Human lighting demands, healthy lighting in an office environment. Public presentation of PhD thesis, University of Technology Eindhoven, Eindhoven Bakker I, Van Der Voordt T, Vink P, De Boon J (2014) Pleasure, arousal dominance: Mehrabian and Russell revisited. Curr Psychol 33(3):405–421 Beersma DGM, Gordijn MCM (2007) Circadian control of sleep-wake cycle. Physiol Behav 90:190–195 Borbély AA, Wirz-Justice A (1982) Sleep, sleep deprivation and depression. A hypothesis derived from a model of sleep regulation. Hum Neurobiol 1:205–210

184

6 Light, Sleep, Alertness and Performance

Borbély AA, Daan S, Wirz-Justice A, DeBoer T (2016) The two-process model of sleep regulation: a reappraisal. J Sleep Res 25:131–143 Boubekri M, Cheung IN, Reid JR, Wang C-H, Zee PC (2014) Impact of windows and daylight exposures on overall health and sleep quality of office workers: a case-control pilot study. J Clin Sleep Med 10(6):603–611 Boyce PR, Beckstead JW, Eklund NH, Strobel RW, Rea MS (1997) Lighting the graveyard-shift: the influence of a daylight-simulating skylight on the task performance and mood of night-shift. Light Res Technol 29:105–134 Buysse DJ, Reynolds CF, Monk TH, Berman SR, Kupfer DJ (1989) The Pittsburgh sleep quality index: a new instrument for psychiatric practice and research. Psychiatry Res 28(2):193–213 Cajochen C (2007) Alerting effects of light. Sleep Med Rev 11:453–464 Cajochen C, Khalsa SBS, Wyatt JK, Czeisler CA, Dijk D-J (1999) EEG and ocular correlates of circadian melatonin phase and human performance decrements during sleep loss. Am J Physiol Regul Integr Comp Physiol 277:R640–R649 Canazei M, Dehoff P, Staggl S, Pohl W (2014) Effects of dynamic ambient lighting on female permanent morning shift workers. Lighting Res Technol 46:140:156 Cappuccio FP, Taggart FM, Kandala N-B, Currie A, Peile E, Stranges S, Miller MA (2008) Metaanalysis of short sleep duration and obesity in children and adults. Sleep 31(5):619–626 Cappuccio FP, D’Elia L, Strazzullo P, Miller MA (2010a) Sleep duration and all-cause mortality: a systematic review and meta-analysis of prospective studies. Sleep 33(5):585–592 Cappuccio FP, D’Elia L, Strazzullo P, Miller MA (2010b) Quantity and quality of sleep and incidence of type 2 diabetes. A systematic review and meta-analysis. Diabetes Care 33:414–420 Cappuccio FP, Cooper D, D’Elia L, Strazzullo P, Miller MA (2011) Sleep duration predicts cardiovascular outcomes: a systematic review and meta-analysis of prospective studies. Eur Heart J 32:1484–1492 Cheung V, Yuen VM, Wong GTC, Choi SW (2019) The effect of sleep deprivation and disruption on DNA damage and health of doctors. Anaesthesia 2019:1–7. https://doi.org/10.1111/anae. 14533 Colau A, Fotios S (2015) Using lighting to improve concentration in the classroom. In: Proceedings of 28th CIE Session, Manchester Costa IC, Carvalho HN, Fernandes L (2013) Aging, circadian rhythms and depressive disorders: a review. Am J Neurodegener Dis 2(4):228–246 Daan S, Beersma DG, Borbély AA (1984) Timing of human sleep: recovery process gated by a circadian pacemaker. Am J Phys 246:R161–R178 Dijk D-J (2012) Sleep and health: beyond sleep duration and sleepiness? J Sleep Res 21:355–356 Dijk DJ, Archer SN (2010) PERIOD3, circadian phenotypes, and sleep homeostasis. Sleep Med Rev 14:151–160 Ferlazzo F, Piccardi L, Burattini C, Barbalace M, Giannini AM, Bisegna F (2014) Effects of new light sources on task switching and mental rotation performance. J Environ Psychol 39:92–100 Ficca G, Axelsson J, Mollicone DJ, Muto V, Vitiell MV (2010) Naps, cognition and performance. Sleep Med Rev 14:249–158 Figueiro MG, Rea MS (2016) Office lighting and personal light exposures in two seasons: impact on sleep and mood. Lighting Res Technol 48:52–364 Figueiro MG, Hamner R, Bierman A, Rea MS (2013) Comparison of three practical field devices used to measure personal light exposures and activity levels. Light Res Technol 45(4):421–434 Figueiro MG, Steverson B, Heerwagen J, Kampschroer K, Hunter CM, Gonzales K, Plitnick B, Rea MS (2017) The impact of daytime light exposures on sleep and mood in office workers. Sleep Health 3:204–215 Figueiro MG, Nagare R, Price LL (2018) Non-visual effects of light: how to use light to promote circadian entrainment and elicit alertness. Lighting Res Technol 50:38–62 Gifford R, Hine DW, Veitch JA (1997) Meta-analysis for environment-behavior and design research, illuminated with a study of lighting level effects on office task performance. In: Moore GT, Marans RW (eds) Advances in environment, behavior, and design. Plenum Press, New York, pp 223–253

References

185

Gornicka GB (2008) Lighting at work: environmental study of direct effects of lighting level and spectrum on psychophysiological variables. PhD thesis, Eindhoven University of Technology, Eindhoven Hubalek S, Brink M, Schierz C (2010) Office workers’ daily exposure to light and its influence on sleep quality and mood. Lighting Res Technol 42:33–50 Huiberts LM, Smolders KCHJ, De Kort YAW (2015a) Shining light on memory: effects of bright light on working memory performance. Behav Brain Res 294:234–245 Huiberts LM, Smolders KCHJ, De Kort YAW (2015b) Shining light on memory: effects of bright light on working performance. Behav Brain Res 194:234–245 Huiberts LM, Smolders KCHJ, De Kort YAW (2017) Seasonal and time-of-day variations in acute non-image forming effects of illuminance level on performance, physiology, and subjective Well-being. Chronobiol Int 34(7):827–844 Ishii H, Kanagawa H, Shimamura Y, Uchiyama K, Miyagi K, Obayashi F, Shimoda H (2018) Intellectual productivity under task ambient lighting. Lighting Res Technol 50:237–252 Iskra-Golec I, Wazna A, Smith L (2012) Effects of blue-enriched light on the daily course of mood, sleepiness and light perception: a field experiment. Lighting Res Technol 44:506–513 Kaida K, Takahashi M, Åkerstedt T, Nakata A, Otsuka Y, Haratani T, Fukasawa K (2006a) Validation of the Karolinska sleepiness scale against performance and EEG variables. Clin Neurophysiol 117:1574–1581 Kaida K, Takahashi M, Haratani T, Otsuka Y, Fukasawa K, Nakata A (2006b) Indoor exposure to natural bright light prevents afternoon sleepiness. Sleep 29:462–469 Kaida K, Takeda Y, Tsuzuki K (2013) The effects of short afternoon nap and bright light on task switching performance and error-related negativity. Sleep Biol Rhythms 11:125–134 Kim Y, Wilkens LR, Schembre SM, Henderson BE, Kolonel LN, Goodman MT (2013) Insufficient and excessive amounts of sleep increase the risk of premature death from cardiovascular and other diseases: the multi-ethnic cohort study. Prev Med 57:377–385 Knez I, Kers C (2000) Effects of indoor lighting, gender, and age on mood and cognitive performance. Environ Behav 32:817–831 Küller R, Wetterberg L (1993) Melatonin, cortisol, EEG, ECG and subjective comfort in healthy humans: Impact of two fluorescent lamp types at two light intensities. Lighting Res. Technol 25(2):71–80 Lee KA, Hicks G, Nini-Murcia G (1990) Validity and reliability of a scale to assess fatigue. Psychiatry Res 36:291–298 Leger D, Bayon V, Elbas M, Philip P, Choudat D (2007) Underexposure to light at work and its association to insomnia and sleepiness. J Psychosom Res 70:29–36 Mednick SC, Ehrman M (2006) Take a nap! Change your life. Workman Publishing Company, New York Mehrabian A, Russell JA (1974) An approach to environmental psychology. MIT Press, Cambridge, MA Meijman TF, De Vries-Griever AH, De Vries G, Kampman R (1988) The evaluation of the Groningen sleep quality scale. Heymans Bulletins Psychologische Instituten, Groningen, University of Groningen, Groningen Mills PM, Tomkins SC, Schlangen LJM (2007) The effect of high correlated colour temperature office lighting on employee wellbeing and work performance. J Circadian Rhythms 5:2–10 Noguchi H, Sakaguchi T (1999) Effect of illuminance and color temperature on lowering of physiological activity. Appl Hum Sci 18:117–123 Odds W (ed) (2015) Sleep, circadian rhythms, and metabolism: the rhythm of life. Apple Academic Press, Inc., Oakville, ON Phipps-Nelson J, Redman JR, Dijk D-J, Rajaratman SMW (2003) Daytime exposure to bright light, as compared to dim light, decreases sleepiness and improves psychomotor vigilance performance. Sleep 26:695–700 Rasch B, Born J (2013) About sleep’s role in memory. Physiol Rev 93:681–766 Rüger M, Gordijn MCM, Beersma DG, De Vries B, Daan S (2006) Time-of-day-dependent effects of bright light exposure on human psychophysiology: comparison of daytime and nighttime exposure. Am J Physiol Regul Integr Comp Physiol 290(5):R1413–RR142

186

6 Light, Sleep, Alertness and Performance

Ryan RM, Frederick C (1997) On energy, personality, and health: subjective vitality as a dynamic reflection of well-being. J Pers 65:529–565 Santhi N, Groeger JA, Archer SN, Giminez M, Schlangen LJM, Dijk D-J (2013) Morning sleep inertia in alertness and performance: effect of cognitive domain and white light condition. PLoS One 8:e79688 Shi L, Katsuura T, Shimomura Y, Iwanaga K (2009) Effects of different light source color temperature during physical exercise in human EEG and subjective evaluation. J Human Environ Syst 12:27–34 Smolders KCHJ, De Kort YAW (2017) Investigating daytime effects of correlated colour temperature on experiences, performance and arousal. J Environ Psychol 50:80–93 Smolders KCHJ, De Kort YAW, Cluitmans PJM (2012) A higher illuminance induces alertness even during office hours: findings on subjective measures, task performance and heart rate measures. Physiol Behav 107:7–16 Smolders KCHJ, De Kort YAW, Van Den Berg SM (2013) Daytime light exposure and feelings of vitality: results of a field study during regular weekdays. J Environ Psychol 36:270–279 Stampi C (ed) (1992) Why we nap; evolution, chronobiological functions of polyphasic and ultrashort sleep. Springer Science+Business Media, New York Takahashi M, Nakata A, Haratani T, Ogawa Y, Arito H (2004) Post-lunch nap as a worksite intervention to promote alertness on the job. Ergonomics 47:1003–1013 Turnage JJ, Kennedy RS, Smith MG, Baltzley DR, Lane NE (1992) Development of microcomputer-based mental acuity tests. Ergonomics 35(10):1271–1295 Van Bommel WJM (2006) Non-visual biological effect of lighting and the practical meaning for lighting for work. Appl Ergon 37:461–466 Vandewalle G, Balteau E, Phillips C, Degueldre C, Moreau V, Sterpenich V, Albouy G, Darsaud A, Desseilles M, Dang-Vu TT, Peigneux P, Luxen A, Dijk D-J, Maquet P (2006) Daytime light exposure dynamically enhances brain responses. Curr Biol 16:1616–1621 Viola AU, James LM, Schlangen LJM, Dijk D-J (2008) Blue-rich white light in the workplace improves self-reported alertness, performance and sleep quality. Scand J Work Environ Health 34:297–306 Wams EJ, Woelders T, Marring I, van Rosmalen L, Beersma DGM, Gordijn MCM, Hut RA (2017) Linking light exposure and subsequent sleep: a field polysomnography study in humans. Sleep 40:zsx165 Westerlund A, Lagerros YT, Kecklund G, Axelsson J, Åkerstedt T (2016) Relationships between questionnaire ratings of sleep quality and polysomnography in healthy adults. Behav Sleep Med 14(2):185–199 Ye M, Zheng SQ, Wang ML, Luo MR (2018) The effect of dynamic correlated colour temperature changes on alertness and performance. Lighting Res Technol 50:1070–1081 Yin J, Jin X, Shan Z, Li S, Huang H, Li P, Peng X, Peng Z, Yu K, Bao W, Yang W, Chen X, Liu L (2017) Relationship of sleep duration with all-cause mortality and cardiovascular events: a systematic review and dose-response meta-analysis of prospective cohort studies. J Am Heart Assoc 6(9):e005947

Chapter 7

Shift Work, Light, Sleep and Performance

Abstract More than 15% of the working force works in shift work. Shift work often leads to a mismatch of the body circadian rhythm and the work-sleep rhythm. The phase shift and thus misalignment between the body circadian rhythm and the worksleep rhythm have adverse effects on the health of the shift worker. It also affects sleep, alertness and performance adversely. Specifically designed shift work lighting can help reduce these problems. The night shift is considered to be the most disruptive one, and therefore this chapter concentrates on night work. Depending on the duration, timing and rotating frequency of shifts and of the risk of work, the objective of shift work lighting is different. For permanent night-shift work and slow-rotating shifts, the goal should be a complete resetting of the circadian rhythm. For fast-rotating shifts, with changeover periods of some 3–7 days, usually partial or compromise phase shifting offers an adequate possibility that also allows the shift worker to have a relatively normal social life. The circadian rhythm of workers in single-night shifts or very-rapid-rotating shifts should preferably not be phase shifted. These objectives can only be obtained with different lighting schedules. Recent research results are available as a basis for such schedules. Some lighting schedules use bright white light of gradually changing colour temperature; others use intermittent very bright light pulses of relatively short duration and again others, light of which the short wavelengths are filtered (short-wavelength depleted white light).

In our 24-h society, between 15% and 25% of the working force work during hours different from the regular daytime working hours (Eurofound 2017; Bureau of Labor Statistics 2005; Alterman et al. 2012). It mostly concerns shift work in which workers succeed one another at the same workplace that is around the clock in operation. Shift work may consist of long-term night shifts or of rotating shifts in which the workers change after a certain period from one schedule to another, e.g. night, morning/afternoon and afternoon/evening shift. The period after which rotating shifts change has a distinct influence on sleep and performance. Of the various shifts, the night shift is considered to be the most disruptive to sleep and performance. Therefore, this chapter focuses on night work.

© Springer Nature Switzerland AG 2019 W. van Bommel, Interior Lighting, https://doi.org/10.1007/978-3-030-17195-7_7

187

188

7.1

7 Shift Work, Light, Sleep and Performance

Circadian Misalignment

The circadian rhythm of most night-time workers who work under no extra bright light does not shift much (Eastman et al. 1995a; Weibel and Brandenberger 1998; Dumont et al. 2001; James et al. 2007; Boudreau et al. 2013). This even holds for the large majority of night-time workers who work on a fixed night-time schedule (Folkard 2008). It leads to a mismatch of the body circadian rhythm and the nighttime work–daytime sleep rhythm. The body is as far as its circadian phase is concerned in the “biological night” at the moment it has to work and in the “biological day” when it has to sleep. Table 7.1 lists the most important causes for the misalignment. One reason that the biological clock does not shift much is that the usual moderately bright lighting in working environments of around 500 lux horizontal illuminance (roughly corresponding to 150–200 lux vertically at eye height) is not sufficient for effective phase shifts. Another reason is that for phase shifting towards the new “night-time work”–“daytime sleep” rhythm, it is also necessary to have darkness during the day after the night shift. Of course, night workers sleep during daytime, but before going to bed and after having slept there is often still full daylight for quite a while. That keeps the biological clock entrained, to a fairly high degree, with the natural day-night rhythm and not with the new “night-time work”–“daytime sleep” rhythm. Yet, another cause is that most night workers switch to a normal day-night rhythm during the weekend or other days off in order not to be socially out of phase with society. The misalignment between the body circadian rhythm and the work-sleep rhythm of the night worker has adverse effects on health in the form of mental, gastrointestinal, cardiovascular and metabolic (among which are obesity and type 2 diabetes) disorders and cancer (Frost et al. 2009; Wang et al. 2011; Archer and Oster 2015; Costa 2016; Vetter et al. 2018). Night-time lighting may play a role in this, as will be discussed in Chap. 10. Further sections of this chapter show that circadian misalignment also affects sleep and performance adversely. A radical remedy against the three causes of circadian misalignment mentioned in Table 7.1 is a complete artificial reversal of day and night as far as lighting and way of living are concerned. Here it should be understood that complete resetting of the circadian rhythm to the new work-sleep rhythm is only beneficial for workers who are on a permanent night shift or a long-lasting night shift of some weeks. It is not desirable to completely reset the biological clock for workers in fast-rotating shifts Table 7.1 Three causes for circadian misalignment in shift workers Night-time lighting: Daytime lighting before and after daytime sleep: Days off:

Horizontal illuminance of 500 lux not effective in phase shifting Bright daylight helps to stay entrained to the natural day-night rhythm Switch back from “night-time work” and “daytime sleep” to the natural day-night rhythm

7.2 Sleep, Alertness and Performance

189

with changeover periods of some 3–7 days. This is because full readjustment takes at least 3 days. With fast-rotating shifts, it means that once readjustment is complete, immediately a new readjustment to the new shift has to start. The consequence is that the fast-rotating shift workers would continuously phase shift back and forth which would put them in constant desynchrony. Partial circadian adjustment can have advantages in such fast-rotating shift situations because it facilitates the switch to a new shift or to a daytime schedule during days off. Smith et al. (2009) proposed a “compromise shift” of 3-h delay for permanent night workers. It brings the phase of the circadian rhythm at which people are most sleepy at around 10:00 in the morning. This is early in the daytime sleep period after a night shift and late in the night-time sleep period on days off. Their study showed that with such 3-h-delayed phase shift, maintained over the night-shift and days-off periods, sleep during the daytime after night shifts and during the late night during days off was sufficient. Compromise shifting also has the advantage that the shift workers can have a reasonably normal social life during days off. For persons working in a single-night shift followed by regular day shifts on the next days (as, for example, medical doctors being active during a single night for an emergency) any phase shifting is undesirable. It would disturb the normal sleepwake possibilities during the days and nights following the single-night shift.

7.2 7.2.1

Sleep, Alertness and Performance Sleep

Not-phase-shifted or only partially phase-shifted night-time workers sleep during the day when it is according to their circadian rhythm (i.e. biologically) daytime. Their sleep is shorter and of lower quality than the sleep of daytime workers that takes place during their biological night (Härmä et al. 1998; Cermakian and Boivin 2003; Dijk and Von Schantz 2005; Paech et al. 2010; Åkerstedt et al. 2010; James et al. 2017). This, of course, is caused by the above-discussed circadian misalignment. Further causes can be daytime environmental noise, and a not-sufficiently-darkened sleeping room during the daytime sleep. Shift work sleep disorder (SWSD) is identified as a circadian sleep disorder in the International Classification of Sleep Disorders (Schwartz and Roth 2006; AASM 2014). Insomnia and excessive sleepiness are symptoms of it. In a study into the occurrence of this disorder involving 2500 workers, it was found that some 10% of night and rotating shift workers aged between 18 and 65 have a shift work sleep disorder according to this classification (Drake et al. 2004). The study also shows that the night shift is the most disruptive one, even more than the rotating shifts used in their study. The prevalence of SWDS in their night workers is some 25% higher than in rotating shift workers.

190

7 Shift Work, Light, Sleep and Performance

Most night shift experiments concerning the use of specific lighting regimes to readjust the circadian rhythm to a new night-time work–daytime sleep schedule show that sleep duration and quality improve with partial or complete readjustment (Eastman et al. 1995b; Dawson et al. 1995; Martin and Eastman 1998; James et al. 2003; Boudreau et al. 2013).

7.2.2

Alertness and Performance

Night-time work has adverse effects on alertness and performance (Rouch et al. 2005; Folkard et al. 2005; Basner et al. 2008; Kazemi et al. 2016). The adverse effects are stronger at lower night-time lighting levels. A study by Boyce et al. (1997) illustrates this well. They studied in a simulated control room the effects of different lighting conditions on night-time workers. Here we describe two lighting conditions providing uniform lighting (CCT of 4000 K) from a luminous ceiling of widely different lighting levels. One installation provided a horizontal illuminance level of 250 lux (roughly corresponding to 80 lux vertically at eye height) and the other 2800 lux (roughly 1000 lux vertically). Sixteen test persons (average age 30.5 years) worked under this lighting, three successive nights, starting at midnight until 8:00 o’clock in the morning. The work activities consisted of filling out questionnaires and carrying out eight widely different performance tests. Subjective alertness was measured using a questionnaire that measures mental activity or arousal (discussed in Sect. 6.2.1). For ease of comparison, we have converted the results into the 9-point scale earlier used in this book for sleepiness and alertness. Figure 7.1 shows subjective arousal for the two lighting regimes averaged over the three nights as a function of hours worked after the beginning of the night-time shift. A steady decline in subjective arousal, or alertness, over the night occurs for both the high- and low-lighting-level regime. The maximum value shown on the vertical scale (4.5) is just below the midpoint of the 1–9-point alertness scale. It implies that the level of alertness of all the night-time workers right from the beginning of their shift is always below the satisfactory level. This holds for both the low- and highlighting conditions. It is also evident from Fig. 7.1 that the high-lighting-level regime always results in a significant higher alertness level, although still under the satisfactory level. In the above-described study of Boyce, the test persons did also carry out nighttime performance tests. Figure 7.2 shows the results of two of the performance tests. It concerns a logical reasoning and a short memory test (described in Sect. 6.2.1). Both performance tests showed a statistically significant difference between the two lighting conditions, with the high lighting level resulting in better performance. The difference between the two lighting conditions is much larger for the reasoning test than for the memory test. It is an illustration of what has been mentioned earlier: different performance tasks may have a different sensitivity for different lighting conditions. Some other performance tests also used in this study did not show an effect of the lighting at all.

7.2 Sleep, Alertness and Performance Fig. 7.1 Subjective alertness for a 250 and 2800 lux horizontal illuminance condition (CCT ¼ 4000 K) as a function of the time worked after midnight (average of three successive nights). Arousal scale used in the experiments is converted to a 1–9-point scale of which only the range of 3–4.5 is displayed. Adapted from Boyce et al. (1997)

191

Arousal (Alertness) 4.5 2800 lux 4.0

250 lux 3.5

3.0 0

2

4

6

8

Time (hrs. after midnight) Fig. 7.2 Performance (in the number of correct answers) of a logical reasoning and a short-term memory task for 250 and 2800 lux horizontal illuminance conditions (CCT ¼ 4000 K). Average of three nights and six measuring cycles per night. The figure is drawn from data of Boyce et al. (1997)

2800 lux

Performance 43 42 41

2800 lux

40 39

250 lux 250 lux

38 Reasoning task

Memory task

Other studies confirm that alertness and performance during night shifts improve with bright light schedules (Crowley et al. 2004; Santhi et al. 2008; Smith et al. 2009; Chang et al. 2012; Chapdelaine et al. 2012; Boudreau et al. 2013). As for the reason for the effect of night-time light on night-time alertness and performance, there is still some uncertainty. The direct photobiological alerting effects valid for daytime light exposure probably also play a role at night. As mentioned earlier, melatonin levels in healthy persons are near zero during daytime and reach their maximum at night-time. Until recently, it was generally thought that the most important alerting and performance-improving effect of light at night is caused by direct (acute) suppression of melatonin (Wright 1997; Cajochen et al. 1998; Daurat

192

7 Shift Work, Light, Sleep and Performance

et al. 2000; Cajochen 2007). Three recent, independent, studies revealed that it is questionable whether indeed melatonin suppression is a route towards the demonstrated alertness and performance improvements by light at night (Rüger et al. 2006; Figueiro et al. 2009, 2014; Regente et al. 2017).

7.3 7.3.1

Night-Time Lighting Strategies Bright Light

A complete reversal of day and night can be obtained by combining a couple of measures: • The night is brightened with artificial light to a level that permits for rapid resetting of the biological clock. • The night workers stay in relative darkness during daytime (for example, with dark goggles with a transmission of approx. 20%). • The night workers sleep in a completely darkened bedroom. • They follow this pattern also during their days off. Of course, such a lifestyle will often not be acceptable. The question, therefore, is whether it is possible with only part of these measures to get substantial improvements in resetting the circadian rhythm for night-time workers. The combination of extra bright light at night, wearing dark goggles while travelling home from the workplace and sleeping in a suitably darkened bedroom, has been investigated in simulated laboratory and field studies. The conditions of these tests were quite different regarding the lighting levels applied and the number of consecutive nights taken into account. The spectrum of the lighting was often only specified in terms of “cool white light”, or “fluorescent tubes were used”. It may be assumed that the colour temperature of the lamps was between 4000 and 6000 K. These studies usually applied high lighting levels of between 3000 and 10,000 lux (measured vertically at eye height) and indeed did show that it is possible to completely or nearly completely reset the clock with the bright light of these levels within approximately three nights (phase shift of 9–12 h). This worked when dark goggles were used to reduce the bright morning light (Eastman 1987; Czeisler et al. 1990; Horowitz et al. 2001; Burgess et al. 2002; Crowley et al. 2003; Sack et al. 2007; Santhi et al. 2008). A recent study showed that three nights with 5000–6500 lux vertically does completely reset not only the central biological clock (SCN) but also peripheral biological clocks (Cuesta et al. 2017). Also here, the spectrum of the light sources is not given. Only a few studies compared the effectiveness of different light levels for resetting the biological clock. These studies show that night-time lighting levels (CCT around 4000 K) down to some 500–600 lux (vertically at eye height) can result in a phase shift of the circadian rhythm of up to 5 h after 3 days (Zeitzer et al. 2000; Santhi et al. 2008). At these relatively low light exposure values, the timing of the

7.3 Night-Time Lighting Strategies

193

Phase Shift 07:30

90

lux

night work 600

23:00

90

07:00

13:00

free

90

lux

23:00

90 07:00

afternoon sleep

14:00

08:00

600 night work 90

23:00

+ 2.3 hr (± 0.6)

22:00 16:00

morning sleep

90

90

16:30

20:00

free

23:00 90

-5.0 hr (± 0.4)

22:00

Fig. 7.3 Light exposure illuminance schedules used in a simulated night-time work shift situation to study the effect of morning and evening sleep time of night-shift workers exposed to 600 lux (vertically at eye height) during the half part of the night shift (CCT 4100 K). Resulting phase shifts after three nights shown on the right. Adapted from Santhi et al. (2008)

daytime sleep influences the phase shift considerably. This is illustrated in Fig. 7.3 which shows the light schedule and phase shifts obtained with the study of Santhi et al. (2008). In this study, 35 persons (average age 28.4) participated in a 10-day shift work simulation protocol. Night-time light exposure of 600 lux vertically (CCT of 4100 K) combined with “afternoon” sleep (14:00–22:00) results in a phase shift after three nights of +2.3 h (Fig. 7.3, top) and “morning” sleep (08:00–16:00) in 5 h (Fig. 7.3, bottom). Note that the light exposure during the night has two different lighting levels. The sequence of these different levels is adapted to the timing of the daytime sleep. Afternoon sleep combines with high-level light at the second half of the night shift, advancing the phase (Fig. 7.3, top), while morning sleep combines with high-level light in the first half, delaying the phase (Fig. 7.3, bottom). Night workers usually prefer to go to bed relatively soon after the end of their night shift (i.e. morning sleep). Free time means that the test persons were free to engage in activities at locations to their wish. In the Santhi et al. (2008) study, the test persons were also asked to do vigilance tests and fill out (KKS) sleepiness questionnaires during their night shift. Figure 7.4 shows the subjective alertness results, as a reversal of KKS sleepiness values, of three consecutive nights compared with the baseline situation (daytime work). The schedule, aimed at complete resetting of the biological clock for permanent or long-lasting night shifts, improves the alertness gradually over the first three nights. Also here it can be seen that bright night-time lighting helps considerably to improve alertness in night-shift work, but it remains under the day-work situation and also, for all but one condition, under the 4.5 midpoint value of the 1–9 alertness scale. The alertness at the third night of the afternoon sleepers recovered to near-daywork level. In line with these results, the phase shift with the afternoon sleepers was indeed larger than with the morning sleepers (see again Fig. 7.3). The visual psychomotor vigilance test (PVT) results, of the same study, are not reported here but did show the same trend as the alertness results.

194 Fig. 7.4 Subjective alertness as determined in three consecutive night shifts with the bright light schedule for afternoon and morning sleep as illustrated in Fig. 7.3 (four night-time hours of 600 lux vertically and four night-time hours of 90 lux vertically) compared with the baseline previous daytime work result. Adapted from Santhi et al. (2008)

7 Shift Work, Light, Sleep and Performance

Alertness 5.0 4.5 4.0 3.5 3.0 2.5

1

2.0 day

2

3

night

Afternoon sleep

1 day

2

3

night

Morning sleep

Santhi et al. (2008) further showed that bright light schedules with moderate brightness induce daytime sleep in night workers of the same quality as the nighttime sleep of daytime workers. The study compared for this purpose the daytime sleep for the schedules of Fig. 7.3 with that for a baseline situation of regular nighttime sleep (22:00–06:00 h) of the previous 2 days with regular day-shift work. Sleep was polysomnographically (PSG) recorded. Daytime sleep duration and efficiency of both the morning and afternoon sleep during the night-shift days did not statistically differ from the baseline sleep on regular day-shift days. Prior history of light also has a significant effect on the lighting level needed for shifting the clock (Dumont et al. 2001; Chang et al. 2011). Dumont et al. (2009) did experiments with simulated night work under 50 lux only, to test the effect on phase shifting of a daytime lighting schedule of 150–300 lux vertically at eye height (CCT of 4100 K) before or after the 8-h daytime sleeping time. Figure 7.5 gives the three different light exposure schedules of this study with the resulting average phase shift of the circadian rhythm after three working nights without bright light. In this study, 38 persons between 20 and 35 years participated. The different lighting levels (vertical illuminances at eye height) used in the schedules are chosen to mimic different situations: • 50 lux: night-time working lighting that does not invoke any phase shifting • 1800 lux daytime lighting to mimic natural outdoor lighting during commuting home from the night shift • 400 lux to mimic natural outdoor lighting while wearing dark goggles • 300 and 500 lux moderate indoor lighting • 20 lux dim indoor lighting

7.3 Night-Time Lighting Strategies

195

Phase Shift 09:00

night work 50 lux 00:00

12:00 14:00

1800 300

150

22:00 24:00

afternoon sleep

20

+ 2.3 hr (± 1.3)

08:00 09:00

night work 50 lux 00:00

1800

17:00

morning sleep

20:00

24:00

20

150

- 1.7 hr (± 1.6)

08:00 09:00

night work 50 lux 00:00

400

17:00

morning sleep

08:00

20:00

150 150

24:00

300

20

- 4.1 hr (± 1.3)

23:00

Fig. 7.5 Three light exposure schedules (CCT 4100 K) used in a simulated night-time work shift situation to study the effect of daylight exposure to circadian phase shifting. Lighting values are vertical illuminances at eye height. Resulting phase shifts after three nights are shown on the right. Adapted from Dumont et al. (2009)

All schedules significantly shift the circadian rhythm with the schedule that includes dark goggles resulting in the largest shift: 4.1 h. Just by controlling daytime light exposure, a phase shift is realised that is of the same order as that obtained with constant bright night-time light of 600 lux horizontally used in the Santhi et al. (2008) study. In most practical situations, however, it is difficult to have night-time workers stick to such a relatively “dark” condition during their entire daytime-awake period.

7.3.2

Intermittent Bright Light

The study of Santhi et al. (2008) described in the previous section (Fig. 7.3) showed already that it is not required for (partial) phase shifting to have constant bright light during the whole work shift. Many studies revealed that intermittent light of bright periods exchanged with periods of relatively low lighting levels phase shifts the circadian system more effectively than constant bright light (Boivin and James 2002; Crowley et al. 2003; Gronfier et al. 2004; Santhi et al. 2008; Smith et al. 2009; Chang et al. 2012; St Hilaire et al. 2012). The reason may be that the phase-shifting effect of bright light depends not only on the amount of the bright light itself but also on changes in the light. Phase shifting may be extra sensitive at the beginning of a bright-light period (Burgess et al. 2002). Relatively short periods of bright light are referred to as bright-light pulses. The schedules investigated in bright-light pulse studies vary widely: light levels between 3000 and 10,000 lux (vertically), number of

196

7 Shift Work, Light, Sleep and Performance

bright-light pulses between one and six and pulse duration between 12 min and 4 h. The study of Santhi et al. (2008) represents the schedule with the lowest pulse light level: 600 lux vertically (see Fig. 7.3 again). All studies included a rather strict daytime protocol with at least the obligation to wear dark goggles (15% transmission or less) while commuting home after the night shift. Smith et al. (2009) tried a light schedule with pulses of 4100 lux vertically aimed at obtaining a continuous delayed phase shift of 3 h resulting in the compromise partial circadian resetting of which the principle is discussed in an earlier part of this section. Their strict schedule (both night and daytime) was successful in obtaining this goal during both the days of night-shift work and the weekend days off. Figure 7.6 shows the schedule for the days of night-shift work. Nineteen test persons (average age 26 years) participated in the study during three simulated night shifts, 2 days off, four more night shifts and 2 more days off. The night-shift lighting uses four bright-light pulses (5100 K, 4100 lux vertically) for 15 min, exchanged with 45 min of dim light (4100 K, 50 lux vertically). To prevent the circadian rhythm from delaying more than 3 h a “light brake” of daylight exposure (average duration 27 min) was included in the afternoon immediately after the morning daytime sleep (8:30–15:30). Sleep after the last night shift before a weekend off was shortened (8:30–13:30) to build a small amount of homeostatic sleep to facilitate the subsequent sleep period during the first day off. During the two weekend days itself, the sleep period was from 03:00 to 12:00. Note that also the Santhi et al. (2008) schedule (Fig. 7.3) includes a “light brake” after the morning sleep period. Earlier it has been shown that experiments with constant bright light gradually made clear that phase shifts can be obtained with much lower bright light levels than initially thought. This makes it likely that the effect obtained with the bright-light pulses of Smith et al. (2009) can also be obtained with considerably lower illuminance values than the 4100 lux used in the study. The schedule of the

Phase Shift 08:30

15:30 17:30

blue block

23:00

00:45

50

01:45

03:00

blue block

-3.2 hr (± 2)

50

02:45

04:00

50

4100

02:00

4100

4100

01:00

50

27 min. daylight

07:00

night work

4100

23:00

morning sleep

23:00

03:45

50 07:00

Fig. 7.6 Light exposure schedule for the night-shift work days used in a simulated night-time work shift situation to get a compromise delayed phase shift of 3 h (during night-shift work days and weekend days off) with intermittent bright-light exposure (5100 K; 4100 lux) of four 15-min pulses. Lighting values are vertical illuminances at eye height. Blue block: glasses with 15% average transmission, ranging from 0% at 400 nm to about 25% at 650 nm. The resulting phase shift is shown on the right. Adapted from Smith et al. (2009)

7.3 Night-Time Lighting Strategies

197

Santhi et al. (2008) study, sketched in Fig. 7.3, with intermittent light of one bright pulse of 600 lux vertically seems to confirm this.

7.3.3

Short-Wavelength Depleted White Light

It has already been mentioned that for persons working in a single-night shift followed by one or more regular day shifts any phase shifting is undesirable. It would disturb the normal sleep-wake possibilities during the days and nights following the single-night shift. Short wavelengths under approximately 550 nm, falling in the melanopic range in which the pRGC has high sensitivities, are in particular effective in phase shifting. Kayumov et al. (2005) were the first to test the effect of night workers, in a bright-light night-shift situation, wearing goggles that filter all wavelengths under 530 nm. These blue block goggles make the light appear orange-white and are therefore also referred to, in laymen terms, as orange glasses. The test involved a comparison of the effect of night-time lighting of 800 lux (horizontally) unfiltered and 800 lux (horizontally) filtered. The circadian rhythm with the filtered light was not shifted, and melatonin not suppressed during the night shift, contrary to the situation with normal bright night-time light. One might expect that alertness and performance would be lower with the filtered light compared to the unfiltered bright-light condition because melatonin with the filtered light was no longer suppressed. Surprisingly that is not the case: alertness and performance were about the same for the filtered and unfiltered bright-light situations. The fact that short-wavelength depleted bright light used at night does not phase shift the circadian rhythm, does not suppresses night-time melatonin and keeps night-time alertness and performance at a similar level as with unfiltered night-time light has been confirmed by other researchers (Sasseville et al. 2006, 2015; Rahman et al. 2011, 2013; Van De Werken et al. 2013; Regente et al. 2017). As will be discussed in Chap. 10, long-term suppression of melatonin may be a possible cause of increased prevalence of cancer in shift workers. It is another reason why the use of shortwavelength depleted bright light is of interest. Figure 7.7 shows the night-time melatonin suppression under dim light (less than 3 lux), bright light and short-wavelength depleted bright light as measured under simulated night-shift conditions by Rahman et al. (2011) and Van De Werken et al. (2013). The left-side curves show results for light filtered below 460 nm and 380 nm, respectively (“yellow” filter), while the right-side curves give the results for light filtered below 530 nm (“orange” filter). The vertical illuminances used in the studies are given in Fig. 7.7. Both studies show that unfiltered bright light suppresses melatonin to a great extent while short-wavelength depleted bright light keeps melatonin levels nearly at the same, much higher, level obtained with near darkness. The left part of the picture also shows that only filtering wavelengths below 460 nm instead of below 480 nm make the effect of filtering considerably less effective. In this wavelength range,

198

7 Shift Work, Light, Sleep and Performance

Relative Melatonin level

Relative Melatonin level

Dim (< 3 lux)

100

Dim (< 3 lux)

100

%

%

80

< 480 nm

80

< 530 nm (190 lux)

(440 lux)

60

60 < 460 nm

40

40

(460 lux)

Bright (255 lux)

20

20 Bright (510 lux)

0 20:00

24:00

Rahman et al. 2011

04:00

0 08:00

Time

22:30

00:30

02:30

Van De Werken et al. 2013

04:30

06:30

Time

Fig. 7.7 Relative melatonin levels with night-time exposure of dim light, bright light and bright short-wavelength filtered light. The 100% value is taken as the level where the dim light situation had its maximum value. Lux values are vertical levels at eye height. In the study depicted on the left, light exposure lasted from 20:00 to 08:00 h. The participants were not dark adapted before the test, just like in regular shift rotation. The participants of the study shown on the right were dark adapted (3 Gbits/sa ca. 10 m No sharing (each user has his or her “own light beam”) None

Limited number of licenced frequencies 2–30 GHz 100 Mbit/s to 1.5 Gbit/s

Daylight and ambient electric light High (communication contained within the room)

All electric devices Low to moderate

20–100 m Sharing (more users: lower transfer rates) High

a

Using light, for example laser LED light, that does not have the function of illumination, data transfer speeds of more than 100 Gbit/s are foreseen

15.2.2 Data Push Applications Many VLC applications concern the use of VLC as a tool for pushing information, i.e. downloading only or broadcasting. Reasons for using VLC instead of RF are no congestion as with the RF bands, lower installation cost, lower energy consumption, higher security, no electromagnetic interference and, in the near future, higher data transfer speeds. Radiofrequency interference with sensitive electronics is a concern in healthcare, laboratories, control rooms and many industrial environments. VLC offers here a secure alternative. Train stations also have problems with this kind of interference. Here, the noise complicates sound communication with the public, making crowd control, especially in emergencies, difficult. VLC communication also opens new perspectives in this respect. New applications using data pushing have emerged and will continue to emerge for situations where the installation of RF systems requires too much effort. An example is the use of VLC in shops in combination with location beacons described already in the section “Internet of Things (IoT)”. Another example is the use of the reading lamp in planes to transfer sound and video for the in-flight personal sound and movie system. Since dedicated communication cables are no longer needed, the robustness of the system increases considerably, while the fuel consumption of the plane decreases because of the eliminated weight of cables.

380

15

Light Beyond Illumination

Entirely new application possibilities of combining downloading and uploading of data by visible light also emerge (TMR 2015). In the next sections, the examples of indoor navigation, light used as a sensor and bidirectional communication (Li-Fi) will be discussed. Many more VLC applications, some of which quite revolutionary, will undoubtedly be developed. Just one revolutionary example from outside the discipline of illumination of rooms will be mentioned here for the medical profession. Studies are ongoing to use visible light that penetrates the human body to communicate with implanted medical devices for monitoring of several functions of the human body or to adjust the settings of the devices (Faria et al. 2017). VLC communication here avoids the problems associated with the use of RF communication in healthcare environments.

15.2.3 Indoor Navigation A GPS (global positioning system) uses the signals received from satellites at different positions to determine the location of the GPS. The satellite signals are not strong enough to penetrate buildings. Consequently, a GPS can only be used outdoors. With a VLC system, it is possible to make the LED luminaires used for the lighting in a building, acting similar to the satellites in a GPS system, so enabling indoor navigation. Such a system is called either a visible light positioning system (VLP) or an indoor positioning system (IPS). The satellite signals received by a GPS are used to calculate the distance of the GPS from the known locations of the satellites. In an indoor VLC navigation system, the data transferred by the luminaires in the room contain an identification of each luminaire so that, with a stored plan of the luminaires, the location of each luminaire (x,y,z) can be determined from the transmitted data (Fig. 15.6). With this data, the location of the photodetector can be calculated. The accuracy obtained depends on the positioning algorithm used, which, in turn, puts requirements on the system itself (Kavehrad and Zhang 2015; Jiménez et al. 2017). A simple algorithm only uses the signal strength obtained from the luminaires and Fig. 15.6 Principle of a visible light positioning system. The smartphone app calculates the position from the locations of the luminaires which are contained in the encoded data transmitted from the luminaires

ID3 ID5 ID1

z

ID7 (x,y,z)1 (x,y,z)3

y

(x,y,z)5

x

(x,y,z)7

ID = luminaire identification

luminaire positions smartphone map + route

15.2

Visible Light Communication, VLC

381

subsequently calculates the approximate location from the four strongest signals received at the photodetector. In principle, the transmitted data can also contain a time signal so that also the time needed for the light from each luminaire to reach the detector can be determined (time of flight). It would allow for a very accurate determination of the position. The camera of a smartphone can be used as photodetector in combination with a smartphone app to carry out the calculations. The smartphone can in this way be used as an indoor GPS. The app can also contain a visualised map of the plan of the building, allowing for navigating through the building just as navigating with a GPS outdoors. Persons wanting to use VLP navigation only have to download an app made available for the VLP-equipped building concerned. VLP indoor navigation systems are useful in, for example, large hospitals, airports, museums, shops and car parks (Arnon 2015; Ghassemlooy et al. 2017). In the case of museums and shops, a VLP system can be combined with pushing additional information once a person arrives at a specific position. In a museum this could be information about the painting at that location. Of course, as with an outdoor GPS, it is also possible to mark the actual location, which may be convenient, for example, for finding a parked car in a large, covered car park.

15.2.4 Light as Sensor By adding to the luminaires of a VLC system photodetectors, it is possible to measure the time of flight, ToF, of light from the moment it leaves a luminaire until its, small, reflected component returns at the luminaire from which it originated (Fig. 15.7). In this way the VLC system can be used for scanning a room to pinpoint the location of objects in the room: light as a sensor, also referred to as time-of-flight, ToF, sensor. The coding enables the distinguishment between light which originates from a specific luminaire and light reaching the photodetector from other luminaires. By combining the results of luminaires installed at different positions, the accuracy of sensing a specific object increases. The contours of objects and even the pose and movements of persons (sitting, standing, laying or walking) can be determined with time-of-flight sensors (Karlicek et al. 2016; Bhattacharya and Radke 2016; Woodstock et al. 2016). Since all this is done in real time, sudden changes in pose from standing to laying can be recognised as a person who has fallen. It can subsequently set an alarm. This emerging feature of light as a sensor, once technically perfected, would be well suited for use in healthcare environments and homes of (older) singles living alone. Many other applications can be thought of. Two applications using outdoor lighting and two using indoor lighting will be described as examples of what is possible. In road lighting, the light of the road lighting luminaires can detect empty parking places and guide car drivers to these places by pushing that information by VLC. Parking garages could be designed with different-sized parking lots. By sensing the contour of each incoming car with the garage lighting, all cars could be VLC-guided to a

382

Light Beyond Illumination

luminaire driver

LEDs

demodulator decoder

processor

photo detector

encoded light

0 01011001 11010111101001110 10 01001100 01 01 10 001100

encoder modulator 0010 0011 110 1011 0110 01101001 1101 0100 01001011 0 0011

Fig. 15.7 Light as a sensor. The time of flight of encoded light of a luminaire leaving the luminaire until its reflected light returns at the luminaire is used to determine the location of an object in a room

15

reflected light

object

matching, empty location. By measuring the contours of passing cars on open roads, traffic lights can be arranged to let large trucks drive on. Since they need much time to accelerate after having stopped, such a measure facilitates a better traffic flow. A different application is that of using light sensing in indoor horticulture. By incorporating in some luminaires a photodetector, the height of the plants (and, also foreseen, the colour of the plants) can be sensed by the lighting to automatically adapt the illumination (level and colour) accordingly. This enables far-reaching automation of huge “grow factories”.

15.2.5 Li-Fi VLC can be extended into a bidirectional communication system with a down- and uplink. Harald Haas of the University of Edinburgh introduced at TEDGlobal in 2011 the name Li-Fi (light fidelity) for such bidirectional systems (Haas 2011). It is a much-needed alternative for, or a complement to, the more and more congested Wi-Fi wireless RF communication system. Of course, it would be disturbing if for the uplink visible light is used. Disturbing light beams would be radiating from all connected devices in the room, such as PCs, laptops and smartphones. So, while with Li-Fi the downlink indeed uses visible light, the uplink uses either invisible infrared or Wi-Fi (Fig. 15.8). Dongles for use in PCs and laptops with an integrated photodetector and infrared or RF (Wi-Fi) source have been introduced. A room connector must be used to interconnect Li-Fi networks in adjacent rooms. It sends data from one side of a wall to the other side through an optical fibre. In most data communication situations,

References

luminaire driver

LEDs

demodulator decoder

processor

photo detector

0 0010 0011 110 1011 1001 0110 0110 11011001011 100

encoder modulator

encoded light

10111101 1011011011 10101110 1110111000

Fig. 15.8 Bidirectional Li-Fi data communication network with visible light downlink and invisible infrared uplink

383

encoded invisible infrared

dongle with photodetector infrared source

much more data is downloaded than uploaded. Therefore, even by using a Li-Fi system with the upload link using not infrared but Wi-Fi, still, a substantial contribution to reducing the congestion of Wi-Fi networks is obtained (Dimitrov and Haas 2015). The comparison between VLC and RF communication given in Table 15.1 also holds for the comparison of Li-Fi and Wi-Fi. Because of the explosive use of the Internet of Things, the limits of the capacity of Wi-Fi will be reached quickly. It makes a fast roll-out of Li-Fi, with its virtual infinite bandwidth, indispensable. It may be expected that this fact, together with the energy efficiency, security and ease of installation of Li-Fi systems will be a strong drive for the fast development of the global Li-Fi market, as confirmed by international market reports (GMI 2016; IHS 2017; TMR 2018). The Li-Fi market will certainly not only exist in lighting installations with the dual function of illumination and data transfer. Dedicated communication-only installations using laser LED beams (in the infrared wavelength region) are also being developed where transfer speeds of 100 Gbits/s are foreseen (Koonen et al. 2015; Haas 2018).

References Arnon S (ed) (2015) Visible light communication. Cambridge University Press, Cambridge Azhar A, Tran T, O’Brien D (2013) A gigabit/s indoor wireless transmission using MIMO-OFDM visible-light communications. IEEE Photon Technol Lett 25(2):171–174 Bell AG (1880a) On the production and reproduction of sound by light. Am J Sci 20(118):305–324

384

15

Light Beyond Illumination

Bell AG (1880b) Apparatus for signalling and communication called “photophone”. Patent No. 235.199 United States patent office Bhattacharya I, Radke RJ (2016) Arrays of single pixel time-of-flight sensors for privacy preserving tracking and coarse pose estimation. 2016 IEEE Winter Conference on Applications of Computer Vision (WACV), pp 1–9 Caicedo D, Pandharipande A (2015) Sensor-driven lighting control with illumination and dimming constraints. IEEE Sensors J 9(9):5169–5176 Cossu G, Khalid AM, Choudhury P, Corsini R, Ciaramella E (2012) 3.4 Gbit/s visible optical wireless transmission based on RGB LED. Opt Express 20:B501–B506 Cseh T, Rajhandari S, Fekete G, Udvary E (2017) Modulation schemes, Chapter 4, pp 97–143. In: Ghassemlooy Z, Alves LN, Zvánovec S, Khalighi M-A (eds) Visible light communications, theory and applications. CRC Press, Boca Raton Dimitrov S, Haas H (2015) Chapter 4: Digital modulation-schemes. In: Principle of LED light communication; towards networked Li-Fi. Cambridge University Press, Cambridge Djordjevoic IB (2012) Coded modulation techniques for optical wireless channels. In: Arnon S, Barry JR, Karagiannidi GK, Schober R, Uysal M (eds) Advanced optical wireless communication systems. Cambridge University Press, Cambridge European Commission (2018) Open AIS D7.6 Final report of project Architectures for Intelligent solid state lighting systems Faria M, Alves LN, Sérgio de Brito André P (2017) Transdermal optical communications, Chapter 10, pp 309–336. In: Ghassemlooy Z, Alves LN, Zvánovec S, Khalighi M-A (eds) Visible light communications, theory and applications. CRC Press, Boca Raton Ghassemlooy Z, Popoola W, Rajbhandari S (2013) Chapter 4: Modulation techniques. In: Optical wireless communications. CRC Press, Boca Raton Ghassemlooy Z, Alves LN, Zvánovec S, Khalighi M-A (eds) (2017) Visible light communications, theory and applications. CRC Press, Boca Raton GMI (2016) Global market insight report GMI462: li-fi market size, share—industry forecast report 2023. Global Market Insights, Selbyville, DE Haas H (2011) Wireless data from every light bulb, TEDGlobal. http://bit.ly/tedvlc. Accessed 1 Nov 2018 Haas H (2018) LiFi is a paradigm-shifting 5G technology. Rev Phys 3:26–31 Huang C, Zhang X (2017) Impact and feasibility of dark-light LED on indoor visible light positioning system. Proceedings 2017 IEEE 17th International Conference on Ubiquitous Wireless Broadband (ICUWB), Salamanca, pp 1–5 IEEE (2011) IEEE standard for local and metropolitan area networks-part 15.7: short-range wireless optical communication using visible light IHS (2017) Bremmer B: Indoor navigation & Li-Fi report—2017. IHS Market, technology, media and telecom Islim MS, Haas H (2016) Modulation techniques for LiFi. ZTE Commun 14(2):29–40 Jiménez RP, Rabadan-Borges JA, Rufo-Torres JF, Luna-Rivera JM (2017) VLC applications for visually-impaired people, Chapter 7, pp 235–252. In: Ghassemlooy Z, Alves LN, Zvánovec S, Khalighi M-A (eds) Visible light communications, theory and applications. CRC Press, Boca Raton Karlicek RF (2018) Lighting and the internet of things. Forum for illumination research, engineering, and science. IESNA, New York Karlicek RF, Radke RJ, Little TDC, Butala PM, Jia L (2016) Sensory lighting system and method for characterizing an illumination space. US Patent No. 9363859 Kavehrad M, Zhang W (2015) Light positioning system (LPS). Chapter 4, pp 70–87. In: Arnon S (ed) Visible light communication. Cambridge University Press, Cambridge Komine T, Nakagawa M (2004) Fundamental analysis for visible-light communication system using LED lights. IEEE Transact Consumer Electron 50(1):100–107

References

385

Koonen AMJ, Oh CW, Mekonnen K, Tangdiongga E (2015) Ultra-high capacity indoor optical wireless communication using steered pencil beams. Proceedings of the International Topical Meeting on Microwave Photonics (MWP2015), Paphos, Cyprus Mathews E, Guclu SS, Liu Q, Ozcelebi T, Lukkien JJ (2017) The internet of lights: an open reference architecture and implementation for intelligent solid state lighting systems. Energies 10(1187):1–27 Minerva R, Biru A, Rotondi D (2015) Towards a definition of the Internet of Things (IoT), Revision 1. IEEE OpenAiS (2018) Open architectures for intelligent solid state lighting systems. http://openais.eu/en/ consortium. Accessed 17 Oct 2018 Pandharipande A, Newsham GR (2018) Lighting controls: evolution and revolution. Lighting Res Technol 50:115–128 Pandharipande A, Zhao M, Frimout E, Thijssen P (2018) IoT lighting: towards a connected building eco-system. IEEE 4th World Forum on Internet of Things (WF-IoT), pp 664–667 Randall (2015) The smartest building in the world, inside the connected future of architecture. https://bloomberg.com/features/2015-the-edge-the-worlds-greenest-building/. Accessed 17 Oct 2018 Tanaka Y, Komine T, Haruyama S, Nakagawa M (2003) Indoor visible light data transmission system utilizing white LED lights. IEICE Transact Commun E86-B(8):2440–2454 Tian Z, Wright K, Zhou X (2016) The dark light rises: visible light communication in the dark. Proceedings of the 22nd annual international conference on mobile computing and networking. MobiCom’16, New York, pp 2–16 TMR (2015) Market research report TMRGL 5909: visible light communication market—global industry analysis, size, share, growth, trends and forecast 2015–2022. Transparency Market Research, Albany TMR (2018) Market research report TMRGL 12164: Li-Fi—global industry analysis, size, share, growth, trends and forecast 2018–2026. Transparency Market Research, Albany Woodstock T-K, Radke RJ, Sanderson AC (2016) Sensor fusion for occupancy detection and activity recognition using time-of-flight sensors. 2016 19th International Conference on Information Fusion (FUSION), pp 1695–1701

Part III

Application

Chapter 16

Lighting Quality and Standards

Abstract The quality of an interior lighting installation must be expressed by photometric values that influence visual performance, visual comfort and non-visual biological effects. The photometric parameters that can be used for specifying, designing and measuring the quality of interior lighting installations are summarised in this chapter. They range from parameters for illuminance level and illuminance uniformity, wall and ceiling luminance, glare restriction, threedimensional object and face recognition, modelling, colour appearance and colour rendering. They are based on findings of the investigations described in Chaps. 2–8. Some of these investigations are seen reflected in standards and recommendations. However, international standards and recommendations prepared by recognised lighting standardisation bodies that specify lighting from both the point of view of visual effects and non-visual biological effects (human-centric lighting) do not yet exist. This chapter describes standards and recommendations with an international character for lighting quality seen from a visual perception and visual comfort point of view, prepared by the International Lighting Commission CIE, the European Committee for Standardisation (CEN) and the Illuminating Engineering Society of North America (IESNA).

16.1

Lighting Quality Parameters

For specification and design purposes, the quality of a lighting installation needs to be expressed in terms of values for quality parameters. Table 16.1 groups the lighting quality aspects for indoor lighting installations under the heading visual and non-visual biological aspects, respectively. For these aspects, quality parameters emerge from the research described in Chap. 3 for visual performance, Chap. 4 for visual satisfaction, Chapters 5–7 for health aspects and Chap. 2 for colour quality aspects. Different standards use different quality parameters to specify lighting quality, as will be shown in Sect. 16.2 of this chapter. Some quality parameters emerging from the research described in this book are not (yet) used in standards. They could, however, be a valuable future addition to lighting standards or be a better replacement for existing quality parameters. Table 16.2 lists all lighting quality parameters that this chapter discusses. © Springer Nature Switzerland AG 2019 W. van Bommel, Interior Lighting, https://doi.org/10.1007/978-3-030-17195-7_16

389

390

16

Lighting Quality and Standards

Table 16.1 Lighting quality aspects for indoor lighting installations Visual lighting quality aspects Lighting level On the tasks On the room surfaces Lighting uniformity Lighting direction Glare restriction Colour rendering

Non-visual biological lighting quality aspects Lighting level On the eye Time dependent Spectrum Time dependent Timing Duration

Table 16.2 Lighting quality parameters Aspect Lighting level and uniformity

Glare restriction Face recognition Modelling Colour appearance

Colour rendering

Lighting quality parameter Average horizontal illuminance

Abbreviation Ehor,av and Ehor,min/Ehor,av

Average wall illuminance

Ewall,av and Ewall,min/ Ewall,av Eceil,av and Eceil,min/Eceil,av Lav,B40 and Lmax/Lmin,B40 MRSE Ee, mel DLeq Emel,D65eq

Average ceiling illuminance Average wall luminance in B40 band Mean room surface exitance Melanopic irradiance Melanopic equivalent daylight (D65) factor Melanopic equivalent daylight (D65) illuminance Circadian stimulus Unified glare rating for a lighting installation Cylindrical illuminance Cylindrical-to-horizontal illuminance ratio Vector-to-scalar ratio Correlated colour temperature

CS UGRL Ecyl Ecyl/Ehor at a point Evector/Escalar at a point CCT

Chromaticity point distance from blackbody locus Colour-rendering index

Duv Ra

Colour fidelity index Colour gamut index

Rf Rg

16.1.1 Lighting Level and Uniformity 16.1.1.1

Horizontal Illuminance

The lighting quality parameter employed for the lighting level on the imaginary working plane is, in the case of horizontal working planes, the average horizontal illuminance, Ehor, av. For different types of tasks, different values are required. The

16.1

Lighting Quality Parameters

391

entire working room area is sometimes taken as the basis for the determination of the average illuminance. However, in cases where it is clear which areas of the workspace will be used as working areas, requirements for the average illuminance need only to be fulfilled for the defined task areas. Defining task areas can lead to energy saving relative to the situation where the entire room area is considered as task area. A prerequisite is that the task areas can be defined in the design stage of the lighting installation. The parameter mostly used for the uniformity of the horizontal illuminance is the ratio minimum to average horizontal illuminance: Ehor, min/Ehor, av. In the case of defined task areas, usually higher uniformity values are specified for the task area itself than for the remaining surrounding areas mainly used for circulation only. All lighting standards use the average horizontal illuminance together with the horizontal illuminance uniformity as a lighting parameter in their specifications. 16.1.1.2

Wall and Ceiling Illuminance

In some lighting standards values are specified for the average wall and ceiling illuminance, Ewall, av, and Eceil, av, respectively. This is done to provide a bright enough impression from the room boundary surfaces for the visual comfort and satisfaction of persons in the room, a subject extensively discussed in Chap. 4. As uniformity parameters, the ratios of minimum to average wall and ceiling illuminances are used. Of course, the final brightness effect is not only dependent on the illuminance on the surfaces but also on the reflection factors of the surfaces. Illuminance values specified are usually aimed at getting satisfying conditions for rooms with wall reflectances of at least 0.5 and ceiling reflectances of at least 0.7. 16.1.1.3

Wall and Ceiling Luminance

In Chap. 4 it was shown that the average luminance in the B40 band, Lav, B40, has the highest correlation with the factor pleasant visual lightness of a room. The lighting quality parameter Lav, B40 is illustrated in Fig. 4.6 of Chap. 4. From the sketch given there, it can be seen that depending on the size and height of the room, different parts of the walls and the ceiling play a role in this parameter. At this moment this quality parameter is seldom used in lighting standards for interior lighting. Since with LEDs it is relatively easy to distribute the light in a room in many different ways, this parameter can be a worthwhile addition to the set of lighting quality parameters used in lighting standards. As was also shown in Chap. 4, a uniformity parameter is connected with the luminance in the B40 band: Lmax/Lmin, B40. 16.1.1.4

Mean Room Surface Exitance

Mean room surface exitance, MRSE, is a measure of the amount of light, reflected from all the surfaces and objects in a space, arriving at the eye of an observer, without including the direct light on the eye from the light sources. As shown in

392

16

Lighting Quality and Standards

Chap. 4, it is equal to the mean illuminance at the eye (for all viewing directions) due to the indirect light coming from all surfaces. MRSE is used to characterise perceived brightness and adequacy of the lighting of a space. In Chap. 4 it has been pointed out that for spaces with different room reflectances, MSRE indeed describes the average quantity of reflected light on the eye, but includes no information about the differences for various viewing directions. It makes the use of MSRE as a general lighting quality parameter less suitable. 16.1.1.5

Melanopic Irradiance

It has been explained in Chap. 5 that a single spectral sensitivity function relevant for all non-visual biological effects of light does not exist. For lighting-level specifications for non-visual lighting applications (human-centric lighting), the melanopic irradiance is probably for many applications the more suitable quality parameter. At this moment there are not yet lighting standards produced by recognised lighting standardisation bodies that specify melanopic irradiances. Many standard bodies are, however, considering including such specifications in future human-centric lighting type of applications. The term “melanopic lux” has been defined as an alternative measure to melanopic irradiance (Enezi et al. 2011). However, it is a non-SI unit and in direct conflict with the guidelines produced by the International Committee on Weights and Measures regarding the International System of Units. This alternative unit is therefore not recommended. In Chap. 5 it was also explained that, for taking into account non-visual biological lighting effects, it might often have sense to compare the melanopic irradiance of a particular light source with the melanopic irradiance by daylight (of the same lighting level at the eye). For this purpose, the “melanopic daylight equivalent factor DLeq” is used. It is the ratio of the melanopic irradiance of a particular lamp to the melanopic irradiance of 6500 K daylight (CIE standard D65 sky). The melanopic daylight equivalent ratio (D65) of a light source can be used to express any photometric quantity (as, e.g., luminous flux, luminous intensity, luminance or illuminance) in an equivalent daylight (D65) quantity. Lighting level can thus be expressed as “melanopic daylight (D65) equivalent illuminance, Emel,D65eq”. For the example of a lighting installation providing 1000 lux, the melanopic lighting level can be indicated as “this lighting installation providing an illuminance of 1000 lux, has a melanopic daylight (D65) equivalent illuminance of DLeq
1000 lux”. 16.1.1.6

Circadian Stimulus

Circadian stimulus, CS, has been described in Chap. 5. It is a measure of the effectiveness of retinal light for the human circadian system based on nocturnal melatonin suppression. For lighting applications where melatonin suppression is a crucial element, it may serve well as a quality lighting parameter. However, it may not do so in general for human-centric lighting applications where many more factors than melatonin suppression play a role. Here the quality parameter of melanopic irradiance, mentioned above, is a more suitable quality parameter.

16.1

Lighting Quality Parameters

393

16.1.2 Glare Restriction 16.1.2.1

Unified Glare Rating for a Lighting Installation

As has been explained in Chap. 4 the unified glare rating (UGR) method makes it possible to predict the discomfort glare rating for each observer position in a lighted room and for each viewing direction. In this way, discomfort glare of each indoor lighting installation can be evaluated for all relevant situations. This may, for example, be important for industrial lighting installations where different working positions may be connected with different tasks, each requiring a different glare restriction. In most offices, the lighting specifier and designer are interested in a metric that predicts the overall quality of an installation with regard to discomfort glare without the need to check many different individual observer positions and viewing directions. So, a glare-restricting quality parameter for an installation as one entity is needed. In Sect. 4.4.2.2, such a parameter has been described based on reference observer conditions and a standard calculation procedure defined by CIE (1995, 2010). The single value parameter is referred to as unified glare rating for a lighting installation, UGRL. It is used in many lighting standards to specify glare restriction for indoor lighting installations.

16.1.3 Face Recognition and Modelling 16.1.3.1

Cylindrical Illuminance

Different lighting parameters for the perceptibility of the human face (and other three-dimensional objects in a room) have been used in different studies (Chap. 3). They include vertical illuminance, cylindrical illuminance, semi-cylindrical illuminance and face luminance, all of course at face height. The semi-cylindrical and cylindrical illuminances are based on a vertical cylinder. Where the lighting of a space has a high degree of symmetry, as is often the case in office type of buildings, the cylindrical illuminance, Ecyl, is easy to work with. This is because Ecyl only requires one value at each point in space to be evaluated. The other parameters require an evaluation at each point on, at least, four mutual perpendicular vertical planes. The cylindrical illuminance at a point is equal to the average of all vertical illuminances on planes around that point.

16.1.3.2

Cylindrical-to-Horizontal Illuminance Ratio

Chapter 4, Sect. 4.3, explained that the balance between the diffuse and directional components of a lighting installation determines the quality of modelling of threedimensional objects in a lighted space. The modelling effect of lighting is good when it reveals the details and texture of objects and results in an aesthetically pleasing

394

16

Lighting Quality and Standards

environment. The cylindrical-to-horizontal illuminance ratio, Ecyl/Ehor, in relevant points of the space, is used in some lighting standards as a quality parameter for the modelling aspect of the lighting installation.

16.1.3.3

Vector-to-Scalar Ratio

A more complicated but in many cases more accurate lighting parameter for modelling is the “vector-to-scalar ratio”, Evector/Escalar, described in Sect. 4.3.1. Here, the illuminance vector, Evector, characterises the direction and magnitude of the light flow at a point in the lit space. The scalar illuminance, Escalar, at a point characterises the overall illumination at that point. Where modelling, and more generally, the flow of lighting is considered to be important the vector-to-scalar ratio could serve as a relevant lighting quality parameter. It is unfortunate that this concept, already introduced in 1967, seems to have fallen into oblivion.

16.1.4 Colour Appearance 16.1.4.1

Correlated Colour Temperature

The colour appearance of a light source can exactly be given by its chromaticity coordinates in the colour chromaticity diagram (Chap. 2). For white-light light sources, it is usually more practical to use the correlated colour temperature, CCT, calculated from the chromaticity coordinates. With the value of CCT, it is relatively easy to picture the colour tint of a white-light source without actually seeing the light. Most lighting standards indeed use CCT as the quality parameter for the colour appearance of white-light light sources.

16.1.4.2

Chromaticity Distance from Blackbody Locus

As discussed in Sect. 2.3.2, a white-light source with its chromaticity point further away from the blackbody locus quickly gets an unacceptable non-whitish tint. With chromaticity points above the blackbody locus, the light gets a somewhat green-yellowish tint and with points under the blackbody locus a pinkish one. The crosswise distance of a lamp’s chromaticity point from the blackbody locus is called Duv. It can be used, in combination with the correlated colour temperature, as a colour appearance quality parameter. It is not yet used in international lighting standards.

16.2

Standards and Recommendations

395

16.1.5 Colour Rendering 16.1.5.1

Colour Fidelity

Besides the colour appearance of a white-light source, the colour rendering of surfaces lit by a white-light source is also of great importance. When the goal is “true” colour rendering, colour fidelity is the goal. Two different colour fidelity quality parameters are used in lighting standards (extensive details are given in Chap. 2). The CIE general colour-rendering index Ra is used since 1965 in lighting standards. The Illuminating Engineering Society of North America has introduced a novel colour fidelity metric, with the name colour fidelity index, Rf. This colour fidelity parameter is especially used in the North American lighting standards. CIE also recognises it for scientific use.

16.1.5.2

Colour Saturation

In some applications, a light source that shifts a particular colour or colours in a specific direction, for example towards more saturation and thus to more vivid colours, may make the visual scene more pleasant. The Illuminating Engineering Society of North America has introduced a colour saturation quality parameter that indicates if, on average, a light source shifts colours into a more or a less saturated direction, and if so how much. That metric is called gamut index, Rg, It is meant to be used in conjunction with Rf. Rg is a valuable parameter not for use in lighting standards but for use by lighting designers of special lighting installations where surface colours have to be shifted on purpose. Think of, for example, theatrical and show lighting installations.

16.2

Standards and Recommendations

Many national and international standards and recommendations related to interior lighting installations do exist. They vary in content from lighting quality aspects, building energy specifications, lamp and luminaire specifications, wired and wireless connection specifications and many more. This book concentrates on standards and recommendations with an international character for the subject lighting quality. There are not yet international standards or recommendations prepared by recognised lighting standardisation bodies that specify lighting from both the point of view of visual effects and non-visual biological effects (human-centric lighting). CIE and ISO established a joint technical committee (JTC 14 “integrative lighting”) to undertake an analysis of published scientific studies and review experience from published application studies on non-visual effects of light on humans, with the aim to provide guidance for safe and beneficial use in lighting applications beyond

396

16

Lighting Quality and Standards

illumination for vision. Experts with a lighting, medical or (chrono)biological background work together in this committee. Hopefully, the outcome of their work lays the cornerstone for future standards and recommendations for lighting installations fulfilling both visual and non-visual biological needs. This section describes the lighting standards and recommendations that specify lighting from the point of view of visual performance and visual comfort prepared by the International Lighting Commission CIE, the European Committee for Standardisation (CEN) and the Illuminating Engineering Society of North America (IESNA). For copyright reasons, tables of these standards and recommendations are reproduced only in part.

16.2.1 ISO-CIE Standard The ISO-CIE Standard “Lighting of indoor workplaces” was prepared by CIE and published in 1989 as a joint ISO-CIE international standard (ISO-CIE 1989). The second edition of this international standard was published in 2002 (ISO-CIE 2002). The standard provides quantitative specifications for the lighting of indoor workplaces that are meant to ensure visual comfort, visual performance and visual safety for the workers. The main body of the standard consists of tables with specified values for work plane illuminance, surrounding area illuminance, glare limitation and colour rendering for 31 different types of indoor workspaces each with different visual tasks. CIE is in the process of updating the ISO-CIE 2002 Standard. The first edition of the European Standard “Lighting of workplaces—part 1: Indoor workplaces, EN 12464-1” published in 2003 was largely based on the 2002 ISO-CIE Standard, while some of the definitions were refined (CEN 2003). In 2011, the second edition of this European Standard was published keeping most lighting specifications for the working plane and adjacent surrounding areas of the first edition unchanged, but adding more types of workplaces. This updated version also contains specifications for the lighting of the walls, ceiling, objects and people in the space (CEN 2011). The 2002 ISO-CIE Standard is not discussed in more detail to avoid duplications in this book. The specifications of the 2011 edition of the European Standard are dealt with in more detail in the next section. Where relevant it is pointed out which new lighting criteria the European Standard added to the ISO-CIE 2002 Standard.

16.2.2 European Standard The main body of the European Standard “EN 12464-1: Light and lighting – Lighting of workplaces – Part 1: Indoor workplaces” consists of tables with specified values for work plane illuminance, surrounding area illuminances, glare limitation and colour rendering for many different categories of interior work areas, each with

16.2

Standards and Recommendations

397

different visual tasks (CEN 2011). Table 16.3 gives all categories for which the standard gives detailed lighting specifications. Each working type area listed in this table is subdivided into areas of different tasks or activities. Table 16.4 shows the subdivision for the category “offices” (category 26).

16.2.2.1

Illuminance of the Working Plane and Its Surroundings

All illuminance level values specified in the European Standard (as well as in the ISO-CIE Standard) are maintained average illuminance values over the task area on the reference surface: Eav, maint. The maintenance factor for the selected lighting equipment, environment and maintenance schedule on which the design has to be based should be stated in the specifications and design documents. To give a similar perceptual difference between successive steps the following scale of illuminance is used (intermediate values are not used in the standard): 20–30–50–75–100–150–200–300–500–750–1000–1500–2000–3000–5000. The EN Standard (as well as the ISO-CIE Standard) defines the working plane as the reference surface at which work is usually done. Although this often is a horizontal plane it can also be vertical or inclined; think for example of tasks in industrial areas or technical hand-drawing offices. The workplace itself is divided into the task area, being the area within which the visual task is carried out, the immediate surrounding area that surrounds the task area and the background area adjacent to the immediate surrounding area (Fig. 16.1). The band of the immediate surrounding area around the task area has a width of at least 0.5 m, while the band of the background area around the immediate surrounding area has a width of at least 3 m within the limits of the space. The size and position of the task area should be stated in the specifications and the design documents. The same holds for the immediate surrounding and background area. For a workplace where the location (and size) of the task area is unknown or where maximum flexibility in the location of task areas is wanted, the whole area is treated as the task area. The required maintained average illuminance on the task area is, of course, dependent on the type of area, task or activity. As an example, Table 16.3 gives the required values for the category offices. The illuminance of the immediate surrounds and background may be lower than the illuminance on the task area. There should, however, be a relationship between the illuminance on the surrounds and the illuminance on the task area to avoid large spatial variations in illuminances in the visual field that could lead to visual stress and discomfort. Table 16.5 gives the minimum required relationship. Using the concept of gradually lower lighting levels from the task area to the immediate surrounding and background areas, of course, results in substantial energy savings relative to using the same lighting level over the whole space. There may be visual tasks that benefit from a larger immediate surrounding area and/or background area than the minimum required bandwidths specified. In such cases, larger bandwidths should be specified and documented in the design documents.

398

16

Lighting Quality and Standards

Table 16.3 Categories of indoor work areas for which the European Standard gives lighting specifications (CEN 2011) General areas inside buildings 1. Traffic zones inside buildings 2. Rest, sanitation and first-aid rooms 3. Control rooms 4. Store rooms, cold stores 5. Storage rack areas

Offices 26. Offices Places of public assembly 28. General areas 29. Restaurants and hotels 30. Theatres, concert halls, cinemas 31. Trade fairs, exhibition halls 32. Museums 33. Libraries 34. Public car parks (indoor) Healthcare premises 37. Rooms for general use 38. Staff rooms 39. Wards, maternity wards 40. Examination rooms (general) 41. Eye examination rooms 42. Ear examination rooms 43. Scanner rooms 44. Delivery rooms 45. Treatment rooms (general) 46. Operating areas 47. Intensive care units 48. Dentists 49. Laboratories and pharmacies 50. Decontamination rooms 51. Autopsy rooms and mortuaries

Industrial activities and crafts 6. Agriculture 7. Bakeries 8. Cement, cement goods, concrete, bricks 9. Ceramics, tiles, glass, glassware 10. Chemical, plastics and rubber industry 11. Electrical and electronic industry 12. Food stuff and luxury food industry 13. Foundries and metal casting 14. Hairdressers 15. Jewellery manufacturing 16. Laundries and dry cleaning 17. Leather and leather goods 18. Metal working and processing 19. Paper and paper goods 20. Power stations 21. Printers 22. Rolling mills, iron and steel works 23. Textile manufacture and processing 24. Vehicle construction and repair 25. Wood working and processing Retail premises 27 Retail premises Educational premises 35. Nursery school, play school 36. Educational buildings

Transportation areas 52. Airports 53. Railway installations

16.2

Standards and Recommendations

399

Table 16.4 Lighting specifications of the European Standard for the different subcategories of the main category “offices” (CEN 2011) Offices (category 26): type of area, task or activity Filing, copying, etc. Writing, typing, reading, data processing Technical drawing CAD work stations Conference and meeting rooms Reception desk Archives

Fig. 16.1 Example for a single-person office of task area, immediate surrounding areas and background area as used in the European Standard

Eav, maint (lux) 300 500 750 500 500 300 200

Uo 0.40 0.60 0.70 0.60 0.60 0.60 0.40

UGRL 19 19 16 19 19 22 25

Ra 80 80 80 80 80 80 80

task area immediate surrounding area background area > 0.5 m

task area

immediate surrounding area

Table 16.5 Relationship between average illuminance on the task area, average illuminance on immediate surrounding areas and average illuminance on the background areas (CEN 2011) Eav, task (lux) 750 500 300 200 150 100 50

Eav, immediate surrounding areas (lux) with Uo  0.4 500 300 200 150 Eav, task Eav, task Eav, task

Eav,background area (lux) with Uo  0.10 1/3
Eav, immediate surrounding areas 1/3
Eav, immediate surrounding areas 1/3
Eav, immediate surrounding areas 1/3
Eav, immediate surrounding areas 1/3
Eav, immediate surrounding areas 1/3
Eav, immediate surrounding areas 1/3
Eav, immediate surrounding areas

The illuminance uniformity, referred to as Uo, is specified as the ratio of the minimum to average value Uo ¼ Emin/Eav. The specified values are given for the category offices for the task area in Table 16.3 and for the immediate surrounding and background areas in Table 16.4.

400

16

Lighting Quality and Standards

Table 16.6 Specified wall and ceiling illuminance with uniformity Uo as specified in the European Standard (CEN 2011) Surface Walls Ceiling

Eav maint (lux) 50 30

Uo 0.10 0.10

Table 16.7 Specifications of the European Standard for recognition and modelling of objects and faces in the interior space (EN 2011) Type of area Activity areas Offices, meeting and teaching areas

16.2.2.2

Ecyl, av, maint (lux) 50 150

Uo, cyl – 0.10

Ecyl/Ehor 0.30–0.60 0.30–0.60

Illuminances on Room Surfaces

Unlike the ISO-CIE Standard, the European Standard also specifies illuminances on the walls and ceiling of the space. Table 16.6 shows the minimum required values for the average illuminance and the uniformity Uo.

16.2.2.3

Lighting of People and Objects in the Interior Space

The 2011 European standard does give, contrary to the 2002 ISO-CIE Standard, lighting specifications for proper recognition of objects and faces for situations where this is considered to be important. The specification is in terms of the average cylindrical illuminance in the space (the basis for the cylindrical illuminance is a vertical cylinder). For good modelling, a range for the value of the ratio of cylindrical to horizontal illuminance at the relevant points in the space is used. Table 16.7 gives the minimum required values. The height at which these values have to be evaluated can, for example, be 1.2 m for sitting people and 1.6 m for standing people.

16.2.2.4

Discomfort Glare of the Installation

Discomfort glare restriction is specified by the unified glare rating of a lighting installation, UGRL, as defined in Sect. 4.4.2.2. As an example, the values specified for the category offices are given in Table 16.4. In the European Standard (as well as in the 2002 ISO-CIE Standard) the UGR tabular method at a 1:1 spacing-to-height ratio is specified for the calculation of UGR. As discussed in Sect. 4.4.2.2, this specification could be improved by changing it into a 1:0.25 spacing-to-height ratio. Hopefully, this will be done in future standards.

16.2

Standards and Recommendations

16.2.2.5

401

Shielding Against Glare

Bright lamps or parts of bright lamps have to be shielded especially in order not to impair the vision of moving persons in the lit space (as explained in Sect. 13.1.5). The European Standard specifies minimum shielding angles in dependence of the luminance of the lamp used in the luminaire. Figure 13.8 shows how the shielding angle is defined. Table 16.8 gives the shielding angles as specified for different luminances of the lamp used in the luminaire.

16.2.2.6

Indirect Glare due to Reflections in Display Screens

For workstations with display screens which are vertical or inclined up to 15 , the standard gives limits for the average luminaire luminance at an elevation angle of 65 and above from the downward vertical (Fig. 16.2). The limits are given for different types of display screens. The screens are characterised by positive polarity (bright screen, dark characters as most office and home systems use) or negative polarity (dark screen, bright characters as most CAD systems use) and by the screen luminance (high screen luminance larger than 200 cd/m2 or medium screen luminance less than 200 cd/m2). Figure 16.2 shows two examples of screens and corresponding limiting values. For positive-polarity display screens the standard’s specification for the luminaire luminance at 65 and above is less than 3000 cd/m2. This requirement can relatively easily be fulfilled with many luminaire types, this in contrast with the requirement of

Table 16.8 Minimum shielding angles as a function of the luminance of the lamp used in the luminaire

Fig. 16.2 Limiting value of luminaire luminance for the example of a positivepolarity, high screen luminance and a negativepolarity, medium screen luminance display screen as specified in the European Standard (CEN 2011)

Shielding angle α No requirement 15 20 30

65°

Lamp luminance (kcd/m2) < 20 20 to < 50 50 to < 500  500

L ≤ 3000 cd/m2 pos. polarity Lscreen > 200 cd/m2

65°

L ≤ 1000 cd/m2 neg. polarity Lscreen ≤ 200 cd/m2

402

16

Lighting Quality and Standards

1000 cd/m2 for negative-polarity displays (CAD systems). Note that the specification should be fulfilled radially around the luminaire.

16.2.2.7

Colour

The minimum value of the colour-rendering index Ra is specified for all types of work, task or activity. In the vast majority, a minimum value of 80 is required, as is also the case in the category offices (see Table 16.3). The choice of colour appearance is considered a matter of psychology and aesthetics. The standard, therefore, does not give specifications for it.

16.2.3 North American Standard The tenth edition of the IES Lighting Handbook published by the Illuminating Engineering Society of North America contains nearly 500 pages with detailed lighting specifications and design guidelines for interior lighting applications (DiLaura et al. 2011). Table 16.9 gives the lighting application fields for which recommendations are given. As an illustration of the North American lighting recommendations, we discuss the subject lighting for offices in more detail. In the IES handbook, lighting for offices is based on the ANSI/IES Design Guide RP-1-12, Recommended Practice for Office Lighting (ANSI/IES 2012). Two addendums for this publication have been published in 2018 (ANSI/IES 2018a, b). The ANSI/IES design guide can be distinguished into two parts, a part which gives lighting specifications and another part that gives recommendations for the lighting design. Here, we concentrate on the part with the lighting specifications. The guide provides detailed lighting specifications for the office application types given in Table 16.10. Each application type is divided into a large number of different visual task categories. As an illustration, the categories for “reading and writing” (office-offices) are shown in Table 16.11. Table 16.9 Indoor lighting application fields for which lighting specifications and design guidelines are given in the IES Lighting Handbook (DiLaura et al. 2011) Lighting for art Lighting for common applications Lighting for courts and correctional facilities Lighting for education Lighting for emergency, safety and security Lighting for healthcare Lighting for hospitality and entertainment Lighting for libraries

Lighting for manufacturing Lighting for miscellaneous applications Lighting for offices Lighting for residences Lighting for retail Lighting for sports and recreation Lighting for worship

16.2

Standards and Recommendations

403

Table 16.10 Office application types for which the ANSI/IES Design Guide RP-1-12 gives lighting specifications (ANSI/IES 2012)

Administration Building entries Circulation Conferencing Drafting and design Equipment rooms Food service

Table 16.11 Visual task categories for the office application type “reading and writing” (ANSI/IES 2012)

Ereaders Facsimile Handwritten work Laptop Microforms (projected)

IT tasks Reading and writing Support spaces Toilets and locker rooms Training rooms Transition spaces

Print media VDT screen and keyboard White board Printer generated on white paper

Table 16.12 Visual task subcategories for the visual task “print media” with the specified maintained illuminance targets for different age groups as specified in the ANSI/IES Design Guide RP-1-12 (ANSI/IES 2012) Print media visual task 6-pt Font matte paper and ink 6-pt Font specular paper and ink 8-pt Font matte paper and ink 8-pt Font specular paper and ink 12-pt Font matte paper and ink 12-pt Font specular paper and ink

Etarget, maint (lux) Age < 25 250 250 150 150 100 100

Etarget, maint (lux) Age 25–65 500 500 300 300 200 200

Etarget, maint (lux) Age > 65 1000 1000 600 600 400 400

Finally, each visual task category is again divided into subcategories to fine-tune further the visual task as is illustrated for the visual task category “print media” in the left column of Table 16.12.

16.2.3.1

Illuminance Target Value System

Design Guide RP-1-12 specifies for all the visual tasks of all the application types so-called target illuminance values. Table 16.12 gives as an example the specified target illuminance values for the visual task “print media” of the application type “reading and writing”. Target illuminance values are maintained values averaged over the task areas. The designer is responsible for determining what the relevant application types and tasks are in a specific situation. Often, the illuminance has to support multiple tasks. In that case, the tasks have to be ranked by importance, prevalence or frequency to determine the commonly occurring task with the highest recommended target illuminance. Recommended target illuminance values are given for three different age groups. The plurality of occupants expected for a given space

404

16

Table 16.13 Upper limits of luminance ratios for an office room as specified in the ANSI/ IES Design Guide “RP-1-12” (ANSI/IES 2012)

Lighting Quality and Standards

Paper task and VDU screen Task and immediate surroundings Task and remote areas

Upper limits 3:1 or 1:3 3:1 or 1:3 10:1 and 1:10

should be reviewed, and the target illuminance selected appropriately. Localised, additional task lighting should be considered for the older age group population in the case that group is a minority. In some applications, task locations and areas are known, such as desk/reading surface, or non-horizontal machine surface. If task locations are known, then the recommended target illuminance values apply only to those locations. The uniformity of the illuminance over the work surface is specified by the ratio of minimum to average illuminance as Emin/Eav  0.7. Illuminance variations across an entire space may be larger. To keep resulting luminance variations in the entire room within appropriate limits, Design Guide RP-1-12 specifies the upper luminance ratio limits as given in Table 16.13. 16.2.3.2

Wall Luminance

Apart from the above-specified luminance ratios between visual task and surroundings of the task, the Design Guide also suggests a minimum average wall luminance in offices of at least 30 cd/m2. 16.2.3.3

Modelling

The Design Report does not give quantitative specifications for modelling of threedimensional objects and faces. It is stated that for modelling a combination of direct and indirect lighting is “generally desirable”. 16.2.3.4

Discomfort Glare of the Installation

In the 1950s, 1960s and 1970s American researchers developed a discomfort glare evaluation system for lighting installations consisting of regular arrays of ceilingmounted direct luminaires with flat-bottom diffusers, like prismatic lenses (EbleHankin and Waters 2004). It is called the visual comfort probability, VCP. Design Guide RP-1-12 writes that according to this evaluation system, office lighting installations should have a VCP value of 80 or larger. The report also states, however, that the VCP system cannot be applied with any reliability to other situations than regular arrays of ceiling-mounted direct luminaires with flat-bottom diffusers. For these situations, the report refers to the CIE unified glare rating system UGR (without giving further specifications).

16.2

Standards and Recommendations

16.2.3.5

405

Overhead Glare

Section 4.4.4 of this book explained that luminaires with large luminance values at high angles (relative to the horizontal) could evoke an uncomfortable sensation, referred to as overhead glare, although these angles are not taken into account in VCP or UGR. To avoid discomfort because of overhead glare, Design Guide RP-112 specifies that luminances of luminaire surfaces or exposed lamps at higher angles should be less than 8000 cd/m2. The higher angles should here be taken as 53 and higher, as 53 is the angle above which VCP does not take luminances into account (for UGR this angle is 60 ). 16.2.3.6

Indirect Glare due to Reflections in Display Screens

To avoid problems because of reflections in display screens, Design Guide RP-1-12 specifies limits for the average luminaire luminance at an elevation angle of 65 and above from the downward vertical, just as the European Standard does (Fig. 16.2). The limits themselves are more severe than those given in the European Standard. These limits are given for different types of display screens. They depend on the significance of viewing monitor screens with respect to overall work. For situations where the viewing significance is normal, as in offices, Table 16.14 gives the limits. As an alternative approach, Design Guide RP-1-12 also gives limits for the luminous intensities emitted by a luminaire at specific angles. 16.2.3.7

Colour

Design Guide RP-1-12 describes the CIE colour-rendering index Ra with a specified value of at least 80. The 2018 Addendum of the Design Guide (ANSI/IES 2018a) describes in addition to Ra also the IES-approved method, TM-30, for evaluating light source colour rendition with the colour fidelity index Rf as a significant improvement over previous methods. This method has been dealt with in some detail in Sect. 2.5.2 of this book. The addendum does not yet specify a minimum value for Rf. Just like the European Standard, Design Guide RP-1-12 does not give recommendations for the correlated colour temperature of light sources. Table 16.14 Limiting values of average initial luminaire luminance for different quality display screens for a normal screen viewing significance as specified in Design Guide RP-1-12 (ANSI/IES 2018a) Screen quality Medium–good pos. polarity Medium–good neg. polarity Poor pos. polarity Poor neg. polarity

Average initial luminaire luminance (cd/m2) at 65 and above 1500 cd/m2 1000 cd/m2 500 cd/m2 200 cd/m2

The screen qualities poor, medium and good are defined in the Design Guide

406

16

Lighting Quality and Standards

References ANSI/IES (2012) Design guide: recommended practice for office lighting. Illuminating Engineering Society of North America, New York ANSI/IES (2018a) Design guide: recommended practice for office lighting, Addendum 1. Illuminating Engineering Society of North America, New York ANSI/IES (2018b) Design guide: recommended practice for office lighting, Addendum 2. Illuminating Engineering Society of North America, New York CEN (2003) Light and lighting—lighting of work places—part 1: indoor work places. EN Standard 12464-1. European Committee for Standardization, CEN, Brussels CEN (2011) Light and lighting—lighting of work places—part 1: indoor work places. EN Standard 12464-1. European Committee for Standardization, CEN, Brussels CIE (1995) International Commission on Illumination CIE Publication 117:1995, Technical report, 1221 Discomfort glare in interior lighting, Vienna CIE (2010) International Commission on Illumination CIE Publication 190:2010, Technical report, Calculation and presentation of unified glare rating tables for indoor lighting luminaires, Vienna DiLaura DL, Houser KW, Mistrick RG, Steffy GR (2011) The lighting handbook, tenth edition: reference and application. Illuminating Engineering Society of North America, New York Eble-Hankin M, Waters C (2004) VCP and UGR glare evaluation systems: a look back and a way forward. Leukos 1(2):7–38 Enezi J, Revell V, Brown T, Wynne J, Schlangen L, Lucas RJ (2011) A "melanopic" spectral efficiency function predicts the sensitivity of melanopsin photoreceptors to polychromatic lights. J Biol Rhythm 26(4):314–323 ISO-CIE (1989) Lighting of indoor work places. International standard ISO8995 CIE S 008/E (1989) ISO-CIE (2002) Lighting of indoor work places. International standard ISO8995 CIE S 008/E, second edition (2002)

Chapter 17

Design Aspects

Abstract Five steps can be identified in the lighting design process that are common to most lighting applications: analysis of the lighting functions, determination of the quality parameters and their values, choice of the lighting and control system, choice of lamp and luminaire types and determination of the number and positions of the luminaires. The more important specific application aspects are discussed for the application fields office lighting, industrial lighting, classroom lighting, lighting for healthcare institutions and emergency lighting. Advantages of general lighting, localised lighting and their combinations are reviewed, just as the advantages of direct, indirect and combination of direct and indirect lighting. Dynamic lighting scenarios for office and industrial lighting that optimise performance, health and well-being are given. For classroom lighting, dynamic automated lighting with the possibility for the teacher to put the lighting in a concentration or relaxation mode is proposed. For wardrooms and intensive care units in hospitals, lighting is discussed that provides a robust and regular circadian rhythm for the patients with potential advantages for their recovery. Dynamic lighting in nursing homes for the elderly can provide not only a robust circadian rhythm but also a therapeutic effect for many Alzheimer’s patients with regard to their sleep-wake rhythm. The most important objective of emergency lighting is to ensure the safety of users and visitors of a building when in the case of a calamity the normal lighting fails. Based on international standards the emergency lighting quality specifications are explained.

Many books have been written about lighting design. Good examples are the extensive lighting design chapters of the SLL Lighting Handbook of the Society of Light and Lighting of the UK (SSL 2018), oriented on CIE and European Lighting Standards, and the Lighting Handbook of the Illuminating Society of North America, oriented on North American Standards (DiLaura et al. 2011). It is outside the scope of this book to give a complete design guide for the many interior lighting application fields. The basic phases of the technical lighting design aspects are described here, and application-specific aspects for office lighting, industrial © Springer Nature Switzerland AG 2019 W. van Bommel, Interior Lighting, https://doi.org/10.1007/978-3-030-17195-7_17

407

408

17

Design Aspects

lighting, classroom lighting, lighting for healthcare institutions and emergency lighting will be discussed in this chapter.

17.1

The Design Process

The lighting design process varies with the type of application and with the character of the designer. As far as the technical aspects of the lighting design process are concerned, five steps can be identified that are common to most lighting applications. They are listed, in the sequence of the design process, in Table 17.1.

17.1.1 Analysis of the Lighting Function Lighting contains scientific, technical, aesthetical and sometimes even art aspects. A consequence is that there are often conflicting requirements and that there is no one ideal solution for a particular situation. It is the task of the lighting designer to prioritise the different requirements. This can only be done by an extensive exchange of information with the client: the first phase of the design process. For this exchange being effective and for avoiding misunderstandings, the lighting designer should avoid as much as possible the use of light technical terms. Even a, for the lighting professional, simple term as “lux” may not be understood by the laymen client, or, even worse, misunderstood as standing for “luxury”. The analysis starts with determining the function or functions of the space. In the case of, for example, an office, industrial or healthcare area, the next step is analysing, for each function, the critical visual tasks of the users of the area. Information about the age distribution of the working force may be required. If non-visual biological aspects of lighting are of importance, details of the work or school schedule or schedules must be obtained. Lighting control requirements, wishes and possibilities should be discussed in depth so that they can be taken into account at an early enough phase of the design process. For this, it is needed to make a survey of all activities to be carried out by the personnel at the different locations in the building. This information forms the basis for determining what lighting sensors are needed, whether a specific area is best suited with area control or individual luminaire control. Table 17.1 The five phases of the lighting design process

1. Analysis of the lighting functions 2. Determination of the relevant lighting quality parameters and their values 3. Choice of the lighting and control system 4. Choice of the lamp and luminaire types 5. Determination of the number and positions of the luminaires

17.1

The Design Process

409

The dimensions of the spaces and, in the case of lighting for work, the floor plan of the workplaces, have to be defined. Positions of windows and doors are needed. Details of the possible mounting possibilities for the luminaires (both from a mechanical and electrical point of view) must be obtained. The (planned) interior design of the space has to be translated into reflectances of the surfaces of the space. Here, we have just one of many examples that illustrates that an early exchange of information between the lighting designer and the client and others involved in the project (such as the interior designer and architect) is essential. Explaining that the reflectances of the interior design and furniture play a role in the quantity of light needed may save considerably in costs and energy usage. Lighting designs should be based on maintained values. Therefore, in this phase of the design, information must be obtained about the factors influencing the maintenance factor, i.e. the cleanness/dirtiness of the space and the planned maintenance schedule. Also this is an example where early information can save costs and energy. To enable the needs for ambience to be satisfied, the preferences for style and aesthetics must be determined. These latter aspects are probably the most import aspects for applications where leisure is the key factor, such as restaurant, bar and hotel lounge lighting. For shops, the shop formula that the client has in mind has to be defined, not only regarding the type of merchandise, but also regarding the customer target group, price level and type of service (self-service or, at the other hand of the service scale, personalised service). Of course, the available lighting budget plays a crucial role in possible lighting solutions. The details of the budget have to be discussed in this early stage of the design process to avoid disappointments at the client and designer. Energy and sustainability aspects should be part of such a discussion.

17.1.2 Determination of Lighting Quality In this phase of the lighting design process, the lighting designer translates that what has been learned in the functional analysis phase into required values for the relevant lighting parameters. Here good lighting standards are of enormous help. The previous Chap. 16 has shown that lighting standards specify lighting values for a large number of different tasks (CEN 2011; DiLaura et al. 2011; ANSI/IES 2012). Even in the case of a task not mentioned in the standard, it is often possible, by comparing the actual task with some different but similar tasks (similar regarding size, contrast and visibility distance), to arrive at sensible proposals for further discussion with the client. As has been discussed in Chap. 16, at this moment there do not yet exist lighting standards, prepared by authoritative lighting bodies, for non-visual biological effects of lighting. Chapters 5, 6 and 7 of this book gave guidance on what illuminances can be considered for non-visual biological effects of lighting for the office and industrial lighting application fields. For the lighting application fields educational institutions, healthcare institutions and nursing homes, Sect. 17.2 of this chapter gives guidance in this respect.

410

17

Design Aspects

17.1.3 Choice of Lighting and Control System Many lighting and control systems are designed with an emphasis on creating good seeing conditions. Such lighting systems can be referred to as functional systems. Functional-only systems are often designed for offices, schools and industrial environments. To satisfy non-visual biological requirements, dynamic lighting systems (in colour and lighting level) can be part of such functional-only systems. The choice for the functional lighting can be general lighting, general plus local lighting or fully localised lighting. These systems can be executed as direct lighting, indirect lighting or a combination of these two. In industrial lighting, the choice for high-bay or low-bay lighting has to be made. The advantages and disadvantages of these systems and subsystems will be discussed for some important lighting application fields in Sect. 17.2 of this chapter. The functional lighting system can, as the only system, fulfil the lighting needs in some office and industrial situations. However, often ambience (attractiveness, atmosphere or status) plays a role as well. In such situations, an additional secondary lighting system may be needed to create the right ambience. In application fields such as restaurants, shops and museums, the functional and ambience parts of the lighting have equal weight. Of course, the two systems should be designed so that they do not interact negatively.

17.1.4 Choice of Lamp, Luminaire and Control Type Only after the phases of analysis of the lighting function and choice of lighting and control system have been completed, the lamp and luminaire type can be chosen. Chapters 11, 13 and 14 give background information helping with this. This information should, of course, be supplemented with photometric, energy and price information from the manufacturer or manufacturers.

17.1.5 Determination of Number and Positions of Luminaires Only in the last phase of the design process the number and positions of the luminaires are determined. To have an equal and honest playing field, calculations needed for this should be based on standards, such as the EN or ANSI/IES standards discussed in Chap. 16. Many different advanced commercial lighting calculation software programs are available that permit for setting the lighting standard according to which the calculations are executed. In most cases, lighting manufacturers pay for having the required photometric data of their luminaires (in digital form) included in these software programs. The advantage for the independent lighting designer is not only the availability of the photometric data of many

17.2

Application-Specific Aspects

411

luminaires from many different manufacturers, but also that the software package is for free. Most of these packages include high-level visualisations of the calculated projects.

17.2

Application-Specific Aspects

17.2.1 Office Lighting 17.2.1.1

Functional Lighting System

The functional lighting in offices can be realised as • General lighting • General plus localised lighting • Fully localised lighting All these systems can be executed as • Direct lighting • Indirect lighting • A combination of direct and indirect lighting General lighting: General lighting employs luminaires in a regular pattern to provide the required lighting level and uniformity over the entire area of the office room. In this case, the total area is considered as the task area. It permits the largest flexibility in the location of the workstations. The lighting quality has to be based on the more critical tasks. If there is a broad range of different tasks, locations with easier tasks get unnecessary high lighting levels, consequently resulting in unnecessary high energy consumption. A smart, sensor-based, lighting control system can control the lighting of workstations dependent on not only the actual occupancy and daylight contribution but also the actual task being carried out. In Chap. 12, it was explained that, because of the mostly horizontal flow of daylight through vertical windows, the luminaires providing the electric lighting should not be switched off when the daylight level reaches the specified lighting level, but at a somewhat higher lighting level of around 1.5 times the specified lighting level. If the general lighting system is executed as a colour and light-level dynamic system, it can also be made suitable to fulfil non-visual biological needs which positively influence health, wellbeing and performance of the office workers. Chapters 5–7 have extensively discussed the subject of non-visual effects of lighting in the context of optimising health and performance for both daytime and shift workers. Figure 6.9 recommends a dynamic lighting scenario for daytime office workers while Fig. 7.10 gives scenarios for shift workers with three different types of rotating schemes. General lighting plus localised lighting: In the case of general plus localised lighting, the general part of the lighting illuminates the entire area to a lighting level lower than that required for the more critical tasks. Local lighting at workstations

412

17

Design Aspects

with the more critical tasks supplements the moderate general lighting. The European lighting standard gives proper guidance as far as the required lighting levels for different activities are concerned (CEN 2011). The workstations with the critical tasks are considered as the task areas, while the remaining area is considered as the immediate surrounding area where the standard indeed specifies lower lighting levels. With a smart control system, it is possible to adapt the lighting level to the actual task carried out at each workstation. The general lighting has to provide, during a part of the morning and at all workstations, the required high lighting and CCT levels to fulfil non-visual biological needs. The local lighting could be dimmed or switched off to save energy during that period. This, of course, is only feasible with a smart lighting installation. Fully localised lighting: Fully localised lighting is lighting where the luminaires are arranged entirely with respect to the workstations. Also with localised lighting, a smart control system can adapt the lighting level to the actual task carried out at each workstation. Fully localised lighting requires the designer to choose luminaires that provide enough “spill light” to ensure a sufficiently high lighting level in the immediate surrounds of the workstations. If fully localised lighting is used in a space where also flexibility is needed, this may be a challenging endeavour. Even in the case of localised lighting with desk lights, rearranging the workstations may result in immediate surroundings with insufficient light. Direct lighting: Direct lighting systems can be obtained from ceiling-surfacemounted, ceiling-recessed-mounted and ceiling-suspended (pendant) luminaires (Fig. 17.1, top). These systems can be efficient because the lighting reaches the working plane directly. Modelling can be reasonably good, provided that the reflectances of walls and ceiling are high enough to obtain, in addition to the direct light contribution also indirect light contributions through inter-reflections. For the same reasons, the luminances of the walls and ceiling can also be sufficiently high. With the use of direct lighting from desktop luminaires, both modelling and room surface luminances may be a problem. If the light distribution of the luminaires used for the general, direct lighting is not too narrow, sudden light transitions are avoided which would be uncomfortable, particularly, for those moving around the space. Specular reflections from the workstation or materials placed on the workstation or from display units may arise. In Chap. 3 (Sects. 3.5.1 and 3.5.2) methods have been discussed to minimise this problem. Often rows of linear luminaires are applied. In that case, they should run parallel to the windows and parallel to the viewing direction of the workers. With this way of arranging the luminaires, not the whole length of the luminaire can give rise to glare. Reflections from the luminaire in the working plane only affect the small area corresponding with the crosswise area of the luminaire (Fig. 17.2, left) instead of affecting a large area corresponding with the longitudinal area of the luminaire (Fig. 17.2, right). Disturbing reflections from daylight in materials on the work desk are strongly restricted because of the crosswise daylight incidence relative to the direction of view. If the row of luminaires nearest to the windows is positioned too close to the windows, persons located in the window zone may be seen in silhouette as was illustrated in Fig. 12.10.

17.2

Application-Specific Aspects

413

Direct lighting

ceiling-recessed

ceiling-mounted

suspended

Indirect lighting

free-standing

suspended

free-standing

suspended

Direct / indirect lighting

ceiling-recessed (or mounted) plus free-standing

Fig. 17.1 Ceiling-recessed, ceiling-mounted, ceiling-suspended and freestanding luminaires used for fully direct, fully indirect and a combination of direct-indirect general lighting systems

reflection area

reflection area

Fig. 17.2 Arranging linear luminaires parallel to the viewing direction and windows limits disturbing glare from the windows and restricts the area of the working plane where disturbing reflections from the luminaires may occur

414

17

Design Aspects

Indirect lighting: Indirect lighting can be obtained from suspended and freestanding luminaires (Fig. 17.1, middle). Freestanding luminaires can be executed in the form of desktop luminaires. Fully indirect lighting has some disadvantages. First of all, the light only reaches the working plane after reflection from the ceiling and upper parts of the walls. Even in the case of a white ceiling and white walls with a reflectance of 0.70, a loss of 30% occurs. Indirect light is inherently diffuse and hardly creates shadows. Modelling, therefore, is poor. The monotone and dull lighting character can be compared with the shadow-free lighting situation outdoors on a cloudy day. An advantage of indirect lighting is that disturbing reflections from lighting in the workstations and display units are limited to a large extent. Whereas with ceiling-mounted and -recessed direct luminaires the luminaire shape and design are visually not very present, that is not the case with suspended and freestanding luminaires. In the latter case, the lighting designer must take into account that the shape and design of the luminaire play a role in the ambience of the space. It can help to give the space an own typical character appropriate for a particular target group. Freestanding luminaires and, to a lesser extent, suspended luminaires may form a visual obstruction to people sitting or working in the space. Simple three-dimensional sketches of the space with the luminaires, made from different viewpoints, help the designer to avoid annoying situations for the users of the space. In the case of freestanding luminaires, a solution has to be found for the power supply cables. Combination of direct and indirect lighting: The combination of direct and indirect lighting makes use of a combination of ceiling-mounted and freestanding luminaires, only freestanding luminaires or only suspended luminaires (Fig. 17.1 bottom). Freestanding luminaires can be of the desktop type and may also be integrated into the office furniture. It facilitates flexibility. If carefully designed the disadvantages of the fully indirect system can be largely avoided. Sufficiently high ceiling and wall luminances can be obtained while minimising the reflection problem in surfaces and display units which may occur with the fully direct lighting system. Modelling will easily be better than with the fully direct system (Veitch et al. 1996; Boyce et al. 2006). The best results are obtained if the direct lighting contributes to more than 50% of the lighting of the working plane. The energy efficiency lies proportionally, according to the direct and indirect contribution, between that of the fully direct and fully indirect system. The combined system is visually more present than the other systems. The lighting installation’s products become part of the interior design of the space. As with the fully indirect system, attention should be paid to avoid the luminaires from forming a visual obstruction.

17.2.1.2

Ambient Lighting System

As has been mentioned before, the functional lighting system can, as the only system, fulfil the lighting needs in quite some office and industrial situations. When special ambience requirements play a role as well, an additional secondary

17.2

Application-Specific Aspects

415

lighting system may be needed to enhance the room decor. This is, for example, in an office type of building often the case in conference rooms. The secondary lighting can consist of accent lighting, effect lighting and architectural lighting. Accent lighting is used to emphasise particular objects or features in the space. Effect lighting is used to create with the lighting itself an attractive feature, for example by creating specific light patterns on a wall. Dynamic lighting (in lighting level and colour) may be an interesting component of effect lighting. Architectural lighting is lighting that emphasises the interior architecture of a space. Usually, the secondary lighting is equipped with a smart dimming control system to enable a multipurpose use of the space.

17.2.2 Industrial Lighting 17.2.2.1

Work Area Lighting

The lighting for industrial work areas can be realised as • General lighting • General plus localised lighting Fully localised lighting in the, mostly, larger industrial areas is not recommended. It easily makes a room look gloomy and chaotic. General lighting: General lighting uses a regular array of overhead luminaires to produce functional lighting over the whole working area with a lighting level and uniformity that fulfils the more critical tasks being carried out in the working area. This lighting provides freedom in the placement of machinery and workplaces. Which type of light is suitable not only depends on the type of tasks being carried out but also on the height of the space. It is convenient to divide industrial interiors into two main groups: low-bay interiors with mounting heights lower than some 7 m, and high-bay interiors with heights above 7 m. High-bay interiors are typically industrial halls. Low-bay interiors come in two versions: those with lower heights up to some 3 m, often housed in multistorey buildings and having an office type of character, while those with higher heights, often single-storey buildings, have a more factory type of character. Low-bay lighting: Low-bay interiors with mounting heights lower than some 3–4 m have much in common with office buildings: white ceilings and walls of high reflectance. Lighting systems, often ceiling mounted, can be of the type also used in office environments. If the working stations are not distributed equally over the space, the concept of defining task areas and immediate surround areas (as discussed in the previous section about office lighting) can be applied to design a more economical, irregular, luminaire pattern focused on the actual task areas. Often rows of linear luminaires are applied. In that case, they should run parallel to the windows and the viewing direction of the worker (see Fig. 17.2).

416

17

Design Aspects

Low-bay interiors with heights higher than some 4 m usually use continuous rows of interconnected linear luminaires, so-called battens. If the machinery allows so, they are often suspended from the roof to hang at some 4 m above the floor. An alternative for interiors with the higher heights in this range, say above approximately 6 m, is the use of small floodlight types of luminaires in a regular or workingstation-orientated irregular pattern. If the spacing between the luminaires is too wide, harsh, disturbing shadows may occur because at some locations strong light comes from mainly one direction only. If this can be avoided, this type of lighting installation can be economical, with respect to installation, operation and maintenance. High-bay lighting: High-bay interiors (height of more than 7 m) use highintensity floodlights, usually referred to as high-bay luminaires, mounted on a rail system suspended from the roof. The high mounting height of the high-intensity luminaires means that for most activities they are far away from the usual lines of sight so that glare can be restricted. Where activities require upward-directed lines of view, the positions of high-bay luminaires must be carefully determined and specific shielding may be required. General plus localised lighting: Where the work positions are permanent, a combination of lower level general lighting with localised lighting can have significant economic advantages. The general part of the lighting illuminates the entire area to a lighting level lower than that required for the more critical tasks. Local lighting at workstations with the more critical tasks supplements the moderate general lighting. The European lighting standard gives proper guidance as far as the required lighting levels for different activities are concerned (CEN 2011). The workstations with the critical tasks are considered as the task areas, while the remaining area is considered as the immediate surrounding area where the standards specify lower lighting levels. Locally mounted luminaires should be located so as not to form visual obstructions for the workers.

17.2.2.2

Task-Related Dedicated Lighting Effects

In industry the variety of tasks is enormous. Many lighting standards specify lighting quality figures and values for many different industrial tasks. If a specific task is not given in such a standard it is still often possible to find similar tasks (for example in terms of contrast, specularity, size and moving speed), sometimes even from a different type of industry, which can be used as guidance for arriving at the required lighting quality for that task. However, there are tasks which require a dedicated analysis to find out which lighting effect enables the execution of the task. Typical examples of such tasks are visual inspection of finished products. The analysis is best done by trying to execute the task at the actual or simulated working area. Some examples of obtaining dedicated lighting effects for some typical inspection tasks are shown in Fig. 17.3.

17.2

Application-Specific Aspects

417

Prevention of veiling reflections Similar light incidence and viewing direction

Detection of specular details Same reflected light direction and viewing direction

Detection of surface irregularities Low angle of light incidence (to produce shadows)

Detection of blemishes in policed surface Large area diffuse light source

Detection of irregularities in transparant material Transmitted light from diffuse light source

Detection of contours Silhouette lighting Fig. 17.3 Examples of how, with different lighting situations, dedicated lighting effects can be obtained needed for executing inspection tasks. Adapted from de Boer and Fischer (1981)

17.2.3 Classroom Lighting 17.2.3.1

Lighting Goals

Lighting in the classrooms of educational buildings should in the first place fulfil the visual performance and comfort needs of the pupils and students. These are much the same as those of office workers. The functional lighting for offices described in a previous section can, therefore, also be used as guidance for classroom lighting.

418

17

Design Aspects

There is, however, one crucial difference: the young age of the school children and students makes carrying out the visual task for them easier. This is because the younger eye functions physiologically better than the older eye and younger people have better eyesight at shorter distances (Chap. 8). Children indeed move their head closer to the visual task, which makes their task visually easier. Most lighting standards, consequently and rightly, specify lower lighting levels for classroom lighting. However, there may be a pitfall in this. Fixed low-level lighting in classrooms, suitable for use by children and young students, may not be suitable for adult training and education in the evening. Apart from facilitating visual performance and comfort, the lighting can also positively influence night-time sleep quality, daytime alertness, relaxation and emotional state of pupils and students. As a consequence, lighting can influence learning performance. Much of the physiological background of these effects of lighting has been extensively discussed in Chapters 5 and 6. In Fig. 6.6 two routes were shown through which some of these effects are obtained. For school pupils and students, a third route can be distinguished. The three different routes are also referred to as three order effects (Novotny and Plischke 2014): • Route 1 (first-order effects): immediate, direct effects of light that directly influence alertness, attention, concentration and mood • Route 2 (second-order effects): long-term circadian effects of the light-dark rhythm which affect in particular the night-time sleeping quality, which, in turn, affects alertness, attention, concentration and mood of the pupil the day after that night • Route 3 (third-order effects): lighting that has positive first- and second-order effects may reduce the disruptive behaviour of some of the pupils, having, in turn, a positive effect on the entire classroom population. Figure 17.4 shows a scheme produced by Mott et al. (2012) that shows the relation of most of the effects of the three routes with the performance and wellbeing of the pupils. Several studies have been published on the potential of dynamic classroom lighting influencing the performance of the pupils and students. Dynamic lighting here relates to both lighting level and colour (CCT). As early as 1981, physiological and behavioural changes in an educational setting (in this case with handicapped children) were analysed under different lighting conditions (Wohlfarth and Sam 1981). This study was in 1993 replicated in a PhD study in a regular elementary classroom with 6-year-old children (Grangaard 1993). The lighting and wall colours in the classroom changed three times during the 30-day test period. Pupils’ behaviour was videotaped for 15 min at the same time each day and subsequently analysed. Blood pressure was measured each day also at the same time. In the 10 days with full spectrum lighting and blue walls, blood pressure decreased as did the off-task behaviour relative to the 10-day period with fluorescent lighting and white walls. The off-task behaviour was 22% lower. After approximately the year 2000, when the lighting community got familiarised with the fundamentals of the mechanism of non-visual biological effects of lighting, many more studies on the

17.2

Application-Specific Aspects

419

School Performance and Well-being

Visual

Circadian

Cognitive

Behaviour

visual performance visual comfort

nighttime sleep daytime alertness and energy

concentration attention

cooperative disturbing restless

Mood

ADHD

Daylight and Electric Light Fig. 17.4 The relation between lighting effects that together influence school performance and well-being of pupils and students. Adapted from Mott et al. (2012)

relation between learning performance and school lighting were carried out (Winterbottom and Wilkins 2009; Figueiro and Rea 2010; Govén et al. 2010; Rautkylä et al. 2010; Barkmann et al. 2012; Mott et al. 2012; Sleegers et al. 2013; Keis et al. 2014; Gentile et al. 2018; Morrow and Kanakri 2018). The experimental conditions of these studies differ considerably in the location on earth, the window design of the buildings, the season (and consequently daylight contribution) and the timing of the experiments. Other differences concern the artificial lighting levels in the control situation (horizontal illuminance range from 300 to 500 lux and correlated colour temperature CCT from 3000 K to 4000 K), in the experimental situation (horizontal illuminance range from 500 to 1000 lux and CCT from 4000 K to 17,000 K), the type of lighting installation and room reflectances and thus room surface luminances. Finally, also the type of test persons (children and students, respectively) and the type of tests the pupils or students had to carry out differed: from performance tests in reading, writing and mathematics to attention, concentration, motivation and mood tests. A detailed mutual comparison is therefore not possible. Although some of the results are conflicting, most of the studies show that an increase in lighting level, especially in the morning hours, increases learning performance and mood. Increase in learning performance may already be obtained by increasing the lighting level (horizontally) from 300 to 500 lux (Govén et al. 2010). The studies in which also the effect of colour temperature was studied mostly demonstrate that an increase in learning performance and mood coincides with an increase in colour temperature in both the morning and afternoon hours. The positive results are most pronounced in the darker winter months and with window designs and locations providing more daylight penetration in the classrooms.

420

17

Design Aspects

In the learning process, moments of concentration and focus have to be alternated with moments of relaxation. Lighting can be used to facilitate this process: lower level, warmer light for relaxation and higher level, cooler light for concentration and focus. One study investigated the effect on students of tuning the lighting by the teacher, based on desired activity and mood (Morrow and Kanakri 2018). The study shows that such tuning of the lighting indeed positively impacts the engagement and mood of the students.

17.2.3.2

Recommendations

From the above, it follows that classroom lighting can benefit from dynamic lighting, i.e. dynamic in lighting level and colour temperature. In the darker periods of the year and in situations with poor daylight penetration, it is recommended to increase the lighting level from the today often recommended level of 300 lux horizontal illuminance to at least 500 lux with an increase in colour temperature to some 6500 K during the first 2 h of the morning. The lighting scenario sketched in Fig. 6.9, adapted to 500 lux, can serve as guidance. For adult education, the scenario should be adapted to 750 lux (as shown in Fig. 6.9). To give the teacher the possibility to adjust the lighting level and colour temperature to the desired activity and mood, at least two dedicated scenes should be made available to the teacher. One “relaxation” scene providing a lighting level and colour temperature reduction and another “concentration scene” providing a lighting level and colour temperature increase. Automatic daylight-dependent lighting control should, of course, be incorporated in a smart, dynamic lighting installation.

17.2.4 Lighting for Healthcare Institutions Healthcare institutions include both hospitals and nursing homes for disabled and elderly persons. Lighting requirements differ widely for medical staff and patients. The dedicated lighting for the different categories of the medical staff is outside the scope of this book. This section concentrates on the lighting for the patients in hospital wardrooms and intensive care units and elderly persons in nursing homes.

17.2.4.1

Wardrooms

Daytime lighting: Patients may be in recumbent, semi-recumbent or sitting positions. For all these positions the lighting should be free from glare. Indirect lighting is, therefore, a good solution for the general lighting in wardrooms. A disadvantage is the poor modelling of such systems. It may make the patients look bad. An alternative solution which can provide better modelling is low-luminance direct lighting. Most standards specify a lighting level for the general lighting between

17.2

Application-Specific Aspects

421

50 and 100 lux with a supplement of 150–300 lux local illuminance at the bed head for reading. This local lighting, covering the full width of the bed, must be easily controllable by the patient. The luminance of the luminaires for the local lighting should be less than 350 cd/m2. In the design of this local lighting, care should be taken that it cannot disturb other patients in the same room. Section 9.3.7 of Chap. 9 did show that the lighting in wardrooms plays an essential role in providing a robust and regular circadian rhythm for the patients. Studies have also been discussed that show that, in particular, morning light has positive effects on the sleep and mood of patients. It may also lead to less pain medication and even a shorter stay in the hospital. These effects of light in wardrooms have not yet taken into account in lighting standards. Until that is the case, it is recommended to use as a basis for the general lighting in wardrooms a customised version of the dynamic lighting scenario given in Fig. 6.9 for office lighting. The early morning part of the customised scenario provides 750 lux horizontal illuminance of high colour temperature (approx. 6500 K) as in office lighting. After some 2.5–3 h, the lighting level decreases not to 500 lux, as for offices, but to some 100 lux with a colour temperature of 3000–4000 K without any after-lunch rise of lighting level or colour temperature. Light (in the morning) with a large blue component (i.e. coloured light) radiated from the luminaires on the ceiling and upper parts of the walls in the morning can strengthen the desired effect (Fig. 17.5). Of course, the electric lighting should be daylight linked to reduce its levels when sufficient natural daylight reaches the beds of the patients. Night-time lighting: Chapter 5 explained the importance of both sufficient daytime lighting and night-time darkness for keeping a healthy circadian rhythm. Nighttime lighting in wardrooms should, therefore, be as minimal as possible. This, of course, starts with good curtains or screens. The lighting should be sufficient to provide the minimum amount of light necessary for nurses and patients to find their way about. For this, an illuminance of 1 lux is sufficient, provided that the adjacent corridor lighting is properly dimmed to some 5–10 lux so as to avoid adaptation problems for the medical staff. Dimming the corridor lighting also prevents disturbing light from entering the wardroom at each visit of a staff member. Fig. 17.5 Low-luminance direct wardroom lighting combined with coloured dynamic indirect lighting on the upper part of the wall (photograph: Philips Lighting)

422

17

Design Aspects

Examination lighting: Supplementary lighting may be needed for patient examination purposes. The amount and quality of that lighting depend on the actual department of the hospital, and the corresponding types of examinations carried out in the wardroom.

17.2.4.2

Intensive Care Units (ICUs)

Section 9.3.7.2 of Chap. 9 explained about the high occurrence of ICU delirium in intensive care patients and the possible connection with a disturbed circadian system. Studies investigating a possible effect of dedicated ICU lighting to reduce the occurrence of ICU delirium indicate sometimes positive, but not yet statistically relevant, results. Investigations are ongoing. Until more research results are available, ICU units should have, as far as the required treatments permit, at least the lighting described in the previous section for wardrooms with special attention for keeping as much as possible one and only one light-dark rhythm: the 24-h rhythm.

17.2.4.3

Nursing Homes

Nursing homes exist for different categories of people. As far as lighting is concerned nursing homes for the elderly set the highest requirements and are, therefore, discussed here. Age effects have been discussed in Chap. 8. From the discussion there, it is clear that considerably less light reaches the retinas of the elderly and consequently evokes less effectively visual and non-visual effects. The older eye is far more sensitive for glare than the younger eye, while adaptation to different lighting levels is slower. Chapter 8 did also show that higher lighting levels, therefore, are required, although they can compensate only partly for these adverse effects. In designing the interior design, reflectances and colours of objects, doors and stairs should be chosen to provide high contrasts. Glossy surface materials should not be used to avoid disturbing reflections. Large light transitions when the residents move around the building should also be avoided. Low-luminance luminaires should be applied, and luminaires carefully positioned to limit glare problems further. Windows must be large for maximum daylight penetration but also for the visual connection with the outside world. Again for glare protection, the windows should be equipped with sunscreens, preferably automatic controlled. Living rooms: The lighting described in the previous section for wardrooms in hospitals can also be used as a guideline for the lighting of the living rooms in nursing rooms. In the case of homes for the elderly, the lighting levels should be increased with a factor of 2, i.e. for the general lighting 100–200 lux with a supplement of 300–500 lux at positions suitable for reading and other hobby activities. The dynamic lighting scenario described in the section on wardrooms is recommended because also the healthy elderly benefit from lighting that stimulates the circadian rhythm (White et al. 2013). Fully indirect lighting is not recommended because of the resulting poor modelling of faces and three-dimensional objects. The

17.2

Application-Specific Aspects

423

latter would not only negatively influence pleasantness, but it also makes practising hobbies difficult or impossible. Smart, fully automated lighting control, incorporating daylight control, not only saves energy. It also avoids residents sitting in too low lighting levels without them realising the negative consequences that has for their circadian rhythm and, in turn, for their night-time sleep quality and daytime activity level. Night-time lighting should be around 1 lux if the resident has to be checked during the night. If checking is not needed, complete darkness is preferable unless the resident was during most of his or her life accustomed to sleeping in a not-sowell-darkened bedroom. In all cases, night-time orientation lighting to move around the room, for example for going to the bathroom, should be provided for with an easily accessible and simple-to-operate switch. Smart lighting control that switches the night-time lighting automatically on and off when the resident leaves and returns to the bed offers a more comfortable solution. Communal living rooms: If dynamic lighting in the living rooms is not possible, the communal living room where many of the residents stay during large parts of the day is the ideal location for a dynamic lighting installation that evokes robust circadian rhythms in the residents. The morning lighting level of the dynamic system should at least provide 1000 lux horizontally, which decreases after some 2.5–3 h to a general lighting level of 200 lux. Section 9.3.3 described that dynamic lighting with a (cool white) lighting level in the morning of some 1200 lux at the plane of the eye has a therapeutic effect for many Alzheimer’s patients with regard to their sleep-wake rhythm (Fig. 9.1). In homes for demented persons, therefore, such a dynamic lighting regime is recommended for the communal living room. For the same reason, homes for demented people should have an indoor garden. Demented persons who for safety reasons are not permitted to leave home unaccompanied can then go outside and benefit from natural outside daylight.

17.2.5 Emergency Lighting The most important objective of emergency lighting is to ensure the safety of users and visitors of a building when in the case of a calamity the normal lighting fails. Emergency lighting, therefore, uses a power source that is entirely independent of the source supplying the normal lighting. In most countries, the emergency lighting of buildings is regulated in laws. These laws may be aimed at those responsible for the building and building installations or those responsible for the people in the building. In many countries, these laws refer to national or international standards for emergency lighting. Such standards are usually split in, on the one hand, standards for specifying the emergency lighting quality and, on the other hand, for specifying the quality of the products used in the emergency lighting installation. The emphasis here is on the emergency lighting quality specifications.

424

17

Design Aspects

Much research has been carried out on what emergency lighting quality is required for different situations. Boyce (2014) gives an extensive overview of emergency lighting research, Lighting standards are based on this type of research. Watts (2012) gives detailed background information on emergency lighting standards. The international lighting commission, CIE, produced in 2007 an ISO/CIE Standard that specifies lighting requirements for emergency lighting (ISO/CIE 2007). Many national and regional standards are based on this standard. The 2013 emergency lighting standard of the European Standardisation Commission, CEN, is an example of such regional standard (CEN 2013). This standard has incorporated some clarifications and improvements where inconsistencies existed in the 2007 ISO/CIE standard. Therefore, this book uses the more recent European Standard as a basis for recommended lighting values. Emergency lighting levels specified in the next sections are maintained lighting levels and do not take into account indirect light contributions so that the minimum effect of the emergency lighting is not affected by changes in the reflectances of room surfaces, for example by repainting or redecoration.

17.2.5.1

Categories of Emergency Lighting

Emergency lighting comprises lighting that facilitates the evacuation of a building in an emergency, as well as standby lighting enabling the continuation of activities of vital importance during a power failure (Fig. 17.6). To facilitate evacuation, persons should be able to orientate themselves in their environment, to locate an escape route which is indicated with discernible signs, and safely follow that escape route to the exit. As shown in Fig. 17.6, escape lighting can be subdivided into escape-route

Emergency Lighting

Escape Lighting

Standby Lighting

enabling safe exit

enabling continuation of essential activities

Escape Route Lighting

Open Area Lighting

Points of Emphasis Lighting

High-Risk Task Area Lighting

guidance to exits

anti-panic

safety equipment and hazard points

enabling shutdown of dangerous processes

Fig. 17.6 Different categories of emergency lighting. In italic characters: the objective of lighting

17.2

Application-Specific Aspects

425

lighting, enabling a safe exit, anti-panic open-area lighting, lighting of points of emphasis such as places with safety equipment, and high-risk task area lighting enabling the shutdown of dangerous processes. Safety signs used for guidance in an emergency should have internal or external lighting that makes them conspicuous and legible from appropriate distances. Escape-route lighting: The emergency plan-designated escape routes, marked with safety signs using pictograms and colours, should have lighting that identifies the escape route and enables the safe use of it. For this purpose, the centre line of the escape route should have illuminance values at floor level, along each point of the line, of at least 1 lux. For escape routes up to 2 m in width, a band centred around that central line, consisting of at least half the width of the escape route, must have at all points at least an illuminance of 50% of the illuminance value provided on the central line (Fig. 17.7, top). Broader escape routes can be treated as a number of 2 m wide strips or as anti-panic open areas. If the diversity of the lighting, i.e. the difference between the minimum and maximum illuminance values, is too high, the higher illuminance values may affect visibility in the areas with lower values. To restrict this adverse effect, the diversity of the lighting (Emin/Emax) must be better than 1/40 (Fig. 17.7, middle). Disability glare should be limited. The ISO/CIE and CEN standards specify maximum luminous intensities for the zones shown in grey in Fig. 17.7, bottom. For equal-level floors the luminous intensities must be limited within the zone 60 to 90 from the downward vertical, whereas for non-level escape routes, such as stairs, the limiting values apply to all angles (see Fig. 17.7, bottom, again). The luminous intensity limit of 500 cd as shown in Fig. 17.7 is valid for luminaire mounting heights smaller than 2.5 m. The standards specify larger values for a series of larger mounting heights. The largest intensity tolerated in the standards is 5000 cd for mounting heights larger than 4.5 m. Fig. 17.7 Escape-route lighting requirements for routes up to 2 m in width. Top: illuminance levels; middle: illuminance diversity; bottom: zones in which luminous intensities have to be limited to restrict disability glare. Values based on CEN (2013)

d

0.5 d

Ecentreline ≥ 1 lux

≥ 0.5 Ecentreline ≥ 0.5 Ecentreline

Emax

Emin/Emax ≥ 1/40 Emin

60° 60°

I ≤ 500 cd (h < 2.5m); I ≤ 5000 cd (h > 4.5m)

426

17

Design Aspects

The emergency escape-route lighting must reach 50% of the required illuminance within 5 s, and 100% within 60 s. The escape-route lighting should continue to provide lighting for at least 1 h. Escape lighting should not only make the route itself visible but also the safety signs should be discernible. For this reason, the colour-rendering index Ra of the light sources used for the emergency lighting should be at least 40. Open-area lighting: Open-area lighting is also referred to as anti-panic lighting. The objective, apart from helping to avoid panic, is to facilitate reaching a place from where an escape route can be easily found. The horizontal illuminance at floor level must at least be 0.5 lux, and the diversity (Emin/Emax) should be better than 1/40. Disability glare should be limited just as specified for the escape-route lighting. Also the start-up time, duration and colour-rendering index of the light sources should be the same as for the escape-route lighting. Points-of-emphasis lighting: Points of emphasis are points where an escapelighting luminaire has to be placed (less than 2 m from the point). If relevant, such a luminaire can be part of the escape route or anti-panic open-area lighting. The most critical points of emphasis are: • • • •

Exit doors intended for use in an emergency Each flight of stairs or any other change in level Each change of direction and intersection of corridors First-aid posts, call points and points with escape equipment for the disabled; first-aid boxes and fire-fighting equipment must have emergency lighting providing a vertical illuminance of at least 5 lux • Disabled refugee points • Safety signs that are part of the emergency plan • Outside the final exit at a place of safety

High-risk task area lighting: The objective of this category of emergency lighting is enabling the shutdown of dangerous processes; these may range from short-lasting hazards like operating an electric saw or long-lasting hazards like a metal or glass oven with the danger of hot liquid pouring out. General emergency lighting standards may not cover all potential risks. In the analysis phase of the emergency lighting design process, it should be carefully checked with the responsible person for the processes whether the standard is relevant enough. Typically, high-risk task areas should at least have an illuminance on the task area of 10% of that required for the task under normal conditions, but not less than 15 lux. The illuminance uniformity, Emin/Eav, should be better than 0.1. The lighting must reach 100% of the required illuminance within 5 s unless the process is identified as one requiring a shorter start-up or permitting a longer one. The lighting should continue to provide lighting as long as the risk lasts. Disability glare should be limited just as specified for the escape-route lighting, and the colour-rendering index should also be the same, i.e. at least 40. Standby lighting: The objective of standby lighting, also called substitute lighting, is enabling an uninterrupted continuation of activities of vital importance during a power failure. The vital importance can have a safety but also a technical or

17.2

Application-Specific Aspects

427

economical reason. An example where standby lighting is required is the hospital operating theatre. The lighting requirements of the standby lighting have to be discussed with the responsible persons for the activity. The lighting requirements are often the same as the requirements under normal operating conditions.

17.2.5.2

Emergency Lighting Installation

The emergency lighting installation has its own power supply system, completely separate from the mains power supply of the normal lighting. The system can be centralised or decentralised. In a centralised system, one central power supply unit powers the emergency lighting luminaires. In a decentralised system, the individual emergency-lighting luminaires are equipped with a chargeable battery making it self-contained luminaires. Centralised system: A centralised system uses a set of accumulators or batteries of sufficient power to start and maintain the emergency lighting through an own cabling network. The units are located at a central location appropriately protected against disturbing external influences. A non-break generator can be used to take over from the batteries to extend the period that the emergency lighting stays on. The cabling network has to fulfil stringent requirements, in particular as far as fire protection is concerned. Centralised systems are so called non-permanent systems in the sense that they are normally switched on only in an emergency. The emergency luminaires can be given a double function by powering them from the mains under normal conditions and integrating them in the normal lighting. In an emergency, they continue to receive power, now from the emergency power system. Of course, such integrated luminaires must be of the same type (visually and light-technically) as the other luminaires used under normal conditions. For escaperoute lighting this may be not realistic, but for the open-area lighting it may be a possibility. Regular tests to check the proper operation of the emergency system can easily be carried out with central systems. Decentralised system: In this system, the individual emergency-lighting luminaires have an individual power supply in the form of a chargeable battery. The luminaires thus function self-contained or autonomous. The luminaire itself incorporates the functions of power supply, battery charging, mains failure detection and illumination. The battery is charged under normal conditions by the standard mains power. Self-contained luminaires can be used individually or in groups. Luminaires can be executed as permanent or non-permanent luminaires. Permanent luminaires provide lighting also under normal conditions. This can, for example, be applicable for the escape-route signalisation. In this way, the users of the space can get acquainted, under normal conditions, with the emergency escape routes. At night, they can serve, partly or entirely, as security lighting. An important advantage of decentralised systems using self-contained luminaires is that the system cannot be affected by faults in the central power supply or cabling network. In the case of a calamity, especially the latter can become a reality. If there

428

17

Design Aspects

is a power defect in only one or a limited number of groups, the self-sustained luminaires in that group or groups respond. Extensions of the emergency system are easily made with a decentralised system. A disadvantage of decentralised systems is that the regular tests of the proper functioning of the system are time consuming for the maintenance staff and may interrupt normal activities at the location of the luminaire. Self-test emergency luminaires can replace the tests carried out by the maintenance staff. When a defect is detected in the self-test carried out by the luminaire itself, it displays a visual warning.

References ANSI/IES (2012) Design guide: recommended practice for office lighting. Illuminating Engineering Society of North America, New York Barkmann C, Wessolowski N, Schulte-Markwort M (2012) Applicability and efficacy of variable light in schools. Physiol Behav 105(3):621–627 Boyce PR, Veitch JA, Newsham GL, Jones CC, Heerwagen J, Myer M, Hunter CM (2006) Lighting quality and office work: two field simulation experiments. Lighting Res Technol 38(3):191–223 Boyce PR (2014) Escape lighting, Chapter 9. In: Human factors in lighting, 3rd edn. CRC Press, Boca Raton De Boer JB, Fischer D (1981) Interior lighting, 2nd edn. Kluwer Technische Boeken, Deventer CEN (2011) EN Standard 12464-1. Light and lighting—lighting of work places—part 1: indoor work places CEN (2013) EN Standard 1838:2013. Lighting applications: emergency lighting. European Committee for Standardization, CEN, Brussels DiLaura DL, Houser KW, Mistrick RG, Steffy GR (2011) The lighting handbook: reference and application, 10th edn. Illuminating Engineering Society of North America, New York Figueiro MG, Rea MS (2010) Lack of short-wavelength light during school day delays dim light melatonin onset (Dimo) in middle school students. Neuro Endocrinol Lett 31:92–96 Gentile N, Govén T, Laike T, Sjöberg K (2018) A field study of fluorescent and LED classroom lighting. Lighting Res Technol 50:631–651 Govén T, Laike T, Raynham P, Sansal E (2010) Influence of ambient light on the performance, mood, endocrine systems and other factors of school children. Proceedings of the 27th Session of the CIE, Sun City Grangaard EM (1993) Effects of color and light on selected elementary students. Dissertation University of Nevada, Las Vegas ISO/CIE (2007) Publication ISO 30061:2007(E)/CIE S 020/E:2007 Joint ISO/CIE Standard: emergency lighting Keis O, Helbig H, Streb J, Hille K (2014) Influence of blue-enriched classroom lighting on students’ cognitive performance. Trends Neurosci Educ 3:86–92 Morrow BL, Kanakri SM (2018) The impact of fluorescent and led lighting on students attitudes and behavior in the classroom. Adv Pediatr Res 5:15 Mott MS, Robinson DH, Walden A, Burnette J, Rutherford AS (2012) Illuminating the effects od dynamic lighting on student learning. SAGE Open 2:1–9 Novotny P, Plischke H (2014) Lighting for health and well-being in education. In: Lighting for health and well-being in education, work places, nursing homes, domestic applications, and smart cities. SSL-erate consortium, Brussels, Belgium Rautkylä E, Puolakka M, Tetri E, Halonen L (2010) Effects of correlated colour temperature and timing of light exposure on daytime alertness in lecture environments. J Light Vis Env 34:59–68

References

429

Sleegers PJC, Moolenaar NM, Galeetzka M, Pruyn A, Sarroukh BE, Van Der Zande B (2013) Lighting affects students’ concentration positively: findings from three Dutch studies. Lighting Res Technol 45:159–175 SSL (2018) LH2018 SLL lighting handbook. CIBSE, London Veitch JA, Miller N, McKay H, Jones C (1996) Lighting system effects on judged lighting quality and facial appearance. Proceedings of 1996 IESNA Annual Conference, Cleveland. pp 519–541 Watts C (2012) A guide to emergency lighting, 2nd edn. The British Standards Institution, London White MD, Leed LC, Ancoli-Israel S, Wilson RD (2013) Senior living environments: evidencebased lighting design strategies. Meta-Analysis 7(10):60–78 Winterbottom M, Wilkins A (2009) Lighting and discomfort in the classroom. J Environ Psychol 29:63–75 Wohlfarth K, Sam C (1981) The effects of color/light changes on severely handicapped children. Planning and Research Branch, Alberta Education

Chapter 18

Calculations and Measurements

Abstract The lighting designer has to perform lighting calculations in order to arrive at solutions that satisfy the relevant lighting requirements. Universally applicable computer programs are available for this purpose. The lumen method of calculating the lighting level on the working plane is a simplified “calculation-byhand” method. It provides unexperienced lighting designers with a tool to learn, for different types of light distribution, the effect the room dimensions and reflectances have on the resulting average horizontal illuminance of the room. The measurements carried out in connection with interior lighting fall into three categories: those to determine the lamp properties, the luminaire properties and the installation properties. The measurements of the first two categories are mostly carried out in laboratories. They concern the measurement of the luminous flux of lamps and luminaires, the light distribution and light-emitting area of luminaires and the spectral data of lamps. Field measurements are carried out on new installations to check whether they fulfil the quality specifications, and on installations already longer in use to reveal whether there is a need for maintenance, modification or perhaps replacement. They concern illuminance, luminance and glare measurements. Light-logging devices are used to gather information about the light dose persons are receiving under different circumstances. The principles of both laboratory and field measurements are discussed in this chapter.

18.1

Calculations

Lighting calculations for interior lighting installations made for lighting design purposes are typically made under the following assumptions: • The space is empty. • The room surfaces reflect perfectly diffuse (Lambertian reflection for which the relationship between the luminance of the surface, L, in all directions, the illuminance, E, on the surface, and the reflectance of the surface, ρ, is given by L ¼ E.ρ/π). • The room surfaces are spectrally neutral (i.e. light of all spectrums reflects the same). © Springer Nature Switzerland AG 2019 W. van Bommel, Interior Lighting, https://doi.org/10.1007/978-3-030-17195-7_18

431

432

18 Calculations and Measurements

Each calculation should clearly state whether the values calculated are new installation values or maintained values. In the latter case, the maintenance factor for which the calculations are made must be explicitly specified.

18.1.1 The Lumen Method The lumen method of calculating the lighting level on the working plane is a simplified “calculation-by-hand” method. It does not take account of the actual distribution of the luminaires in the space. It comes in handy for estimating quickly how many luminaires of a particular type are roughly required for illuminating a specific space. Probably in the computer era more important, it provides unexperienced lighting designers with a tool to learn, for different types of light distribution, the effect the room dimensions and reflectances have on the resulting average horizontal illuminance of a room. The lumen method can be applied if the utilisation factors, UF, are available for the luminaire for which the calculation is intended. Most luminaire manufacturers provide utilisation factors with the photometric data of their luminaires. Section 13.1.2.2 “Utilisation Factors” described that the utilisation factor specifies how efficiently the lamp light of a specific luminaire reaches the working plane, taking into account both the direct and indirect contributions. Table 13.3 of that section gives an example of a utilisation factor table for a specific luminaire. That table gives utilisation factor values in dependence of a set of different reflectances of the ceiling, walls and working plane surface for different room index values. As explained in Chap. 13, the room index value characterises the dimensions of the room and is calculated from the room length and width and the distance between the luminaires and the working plane. The average working plane illuminance is calculated from E av ¼

Φlamps  UF room area

with φlamps ¼ total luminous flux of all lamps (lm) The utilisation factors are calculated for uniformly distributed luminaires in an arrangement defined by CIE and given in Table 13.2. This means that care must be taken with the results obtained in the case of non-uniform-distributed luminaire arrangements. Of course, the lumen method calculation does not provide any information about the uniformity of the illuminance pattern. Note that IES of North America uses, instead of the term utilisation factor, the term coefficient of utilisation (same definition) and, instead of room index k, room cavity ratio, which is equal to 5/k (IES 2011).

18.2

Measurements

433

18.1.2 Computerised Calculations Many advanced commercial lighting calculation software programs are available that permit lighting calculations to be executed by computer. In most cases, lighting manufacturers pay for having the required photometric data of their luminaires included in their software programs. In this way, the lighting designer has, usually free of charge, access to a professional calculation tool and the photometric data of many luminaires of many different manufacturers. Most of these packages include high-level visualisations of the calculated projects. Many of the lighting calculation software programs can set the calculation method according to specifications of specific international and regional lighting standards.

18.2

Measurements

18.2.1 Light Detectors 18.2.1.1

Types

For all light measurements, photoelectric cells are used that convert the light incident upon them into electrical energy, the amount of which is measured. The principle of these cells is the same as that of solar cells of the photovoltaic type: they convert incident light directly into electric power without the need for an external voltage source. However, photovoltaic cells used for accurate and low-lighting-level measurements usually employ external batteries to amplify the output current of the cell. Photovoltaic cells consist of two layers of semiconductor material: n- and p-types (as with LED light sources). Again as with LED light sources, they are in fact semiconductor diodes and are therefore often referred to as photodiodes. Formerly, selenium, with a barrier layer in between the n- and p-layers, was used as semiconductor material. They were referred to as barrier-layer cells. Today, silicon and silicon dioxide are commonly used for light-measuring photodiode cells. They are available in sizes ranging from 0.5 mm to something like 30 mm in diameter (Fig. 18.1).

Fig. 18.1 Detector cells. From left to right: selenium cell, two silicium cells, silicium array and 2 megapixel CCD cells

434

18 Calculations and Measurements

Since photoelectric cells measure the light falling on the surface area of the cell, it is the illuminance that is measured. In photometry, therefore, all photometric quantities that have to be measured have to be converted to illuminances. For luminance measurements, this means that the image of the surface whose luminance has to be measured, as the bright area of a luminaire or a wall or ceiling, has to be projected on the face of the photocell. A luminance meter, therefore, contains a photocell with an optical system in front of it that projects an image of the scene to be measured onto the face of the photocell. Usually, the luminance meter has a second optical system that enables it to be aimed accurately at the location to be measured. By employing an adjustable diaphragm, the size of the area projected onto the cell can be changed from spot measurements to larger area measurements. A possibility for measuring the complete luminance pattern of a scene in just one measurement is obtained with the introduction of charge-coupled device (CCD) and complementary metal-oxide semiconductor (CMOS) chips, which were developed for digital photo and video cameras. CCDs consist of one piece of n-p silicon semiconductor material, which is, in fact, a silicon photodiode cell. The top layer is divided by lines of electrodes, often of aluminium, and lines of insulator material, into a raster of many tiny dots called pixels. Today, chips with a size smaller than 1 cm2 containing more than 10 megapixels are no exception: a matrix with more than 3000  3000 dots. The light incident on each pixel area is, as in ordinary photodiodes, transformed into an electrical charge. CCDs have no traditional wiring to all individual pixels. They transport the charge at a pixel across the chip to the edge of the chip where it is converted into a voltage and read while memorising the precise location from where it originates in the cell’s matrix. In CMOS chips, which are made of metal-oxide-silicon, each pixel has its own charge-to-voltage conversion and reading circuit. This reduces the area for light capture and thus the resolution and sensitivity of the chip. Whereas CCDs require a complicated manufacturing process, CMOS chips are manufactured in much the same way as most microchips and can, therefore, be produced considerably cheaper in a mass production process. Today, CMOS technology is rapidly developing to include high-end applications, thus decreasing the quality gap with CCDs. A digital photo camera with a CCD or CMOS chip records and remembers the amount of light arriving from each scene element, which in turn is proportional to the luminance of that element. With specific software, it is now possible to analyse the complete luminance pattern of that picture. For certain digital photo cameras, such specific software is commercially available. For absolute values, appropriate calibration is required. Specific luminance meters that need no separate calibration, also using CCD or CMOS chips, are being produced as well.

18.2.1.2

V(λ) Correction

The spectral response of a photocell differs considerably from that of the eye as defined by the spectral sensitivity V(λ) of the CIE Standard Observer, on which all light units are based. For this reason, a so-called correction filter or filters must be

18.2

Measurements

435

employed over the cell window in the cell housing. The various makes of cell, and even cells from the same manufacturer, differ in the degree of correction needed to match cell to spectral eye sensitivity. Each cell should, therefore, be supplied with its own appropriate correction filter. It should be realised that exact correction using filters is difficult to achieve and makes meters expensive. Cheaper meters, those with insufficient V(λ) correction, need “colour-correction factors” for each different type of light source. These correction factors have to be supplied by the meter manufacturer. Such meters are not suitable for the measurement of mixtures of different light sources especially those including daylight. Given the large number of different spectra that LED light sources may have, it is impossible for a meter manufacturer to supply correction factors for all the possible different types of LEDs. Poor V(λ) corrected light cells should therefore not be used for the measurement of LEDs or LED installations.

18.2.1.3

Cosine Correction

Light falling on the surface of the cell should, according to the definition of illuminance, be weighted according to the cosine of the angle of incidence. A measurement error can occur because at large angles of light incidence part of the light is reflected by the transparent, somewhat glossy, protective layer covering the cell. This part does not reach the light-sensitive surface of the cell, and therefore the cell does not precisely follow the cosine law. In high-quality photometers, the cell and its housing, in conjunction with the covering material of the cell, are so shaped as to obtain correct cosine correction (Fig. 18.2). For cylindrical, semi-cylindrical and spherical illuminance measurements there are special adapters available that have to be placed over the normal photocell (Fig. 18.2, right).

18.2.1.4

Pulsed Light Measurement

To dim LEDs, pulse width modulation (PWM) is one of the dimming methods that can be employed. When measuring the dimmed lighting level, it should be noted that the measuring result of the photometer may deviate from the arithmetic average of Fig. 18.2 Example of photocell housings with a protective layer that together provide cosine correction, and (on the right) an adapter for the measurement of semi-cylindrical illuminance

436

18 Calculations and Measurements

the modulated light. The deviation especially depends on the modulation frequency and on the internal response time of the photocell and the amplifier.

18.2.1.5

Ambient Influences

Some photocells are affected more than others by changes in ambient temperature. Where this affects measurement accuracy, it is necessary to have the instrument calibrated for different ambient temperatures. Appropriate correction factors can then be applied to the readings obtained. This may be relevant for industrial areas with extreme temperatures. Moisture on the photocell, or on any filter or corrector in use, can detract considerably from the accuracy of the measurements. The cell should, therefore, be kept dry, and it is better to avoid making measurements altogether during periods of extremely high humidity. Although photodiode cells have only very small fatigue effects, the amplification and reading circuitry may have a larger fatigue effect exhibiting hysteresis. This leads to a too high or too low reading, depending on the previous light condition in which the photometer was used. In situations where there are large differences in lighting levels being measured, adaptation times of up to as much as several minutes may be required until a stable reading is obtained.

18.2.1.6

Ageing Effects

The absolute and spectral sensitivity of a photoelectric cell and its sensitivity to ambient temperature change with time. Also, the electrical amplification circuitry and reading circuitry may change with time. The complete instrument has, therefore, to be calibrated at least once a year. Many high-quality manufacturers offer a calibration service for this purpose.

18.2.1.7

Accuracy

The previous sections have shown that the accuracy of light measurements depends on the operational conditions, but perhaps even more so on the characteristics of the photometer itself. The CIE Standard, “Characterization of the performance of illuminance meters and luminance meters” (CIE 2013), defines quality indices for different quality aspects of photometers, the more relevant ones having been dealt with in the previous sections. Table 18.1 lists the quality aspects that are defined in this standard, together with the quality index. The smaller the value of the index, the smaller the inaccuracy and thus the better the quality of the relevant aspect. If photometer manufacturers were to provide the values of the relevant quality indices for their meters, it would allow not only for mutual comparison of the quality of different meters but also for determining which meter is more suitable for use in specific applications and under specific conditions.

18.2

Measurements

Table 18.1 CIE quality indices for photometers (CIE 2013)

437 Quality aspect V(λ) mismatch UV response IR response Cosine response Directional response Surround field influence Linearity Display unit Fatigue Temperature dependence Humidity resistance Modulated light

Quality index f10 fUV fIR f2 f2,g f2,u f3 f4 f5 f6,T

Polarisation Spatial non-uniformity response Range change Focusing distance

f8 f9

f6,H f7

f11 f12

Remark

Illuminance only Luminance only Luminance only

For relevant frequencies

Luminance only

18.2.2 Measuring Lamps 18.2.2.1

Luminous Flux

The luminous flux of a lamp or of a luminaire can be obtained by measuring the luminous intensities in all directions with a goniophotometer. This is described in a next Sect. 18.2.3.1. Through the calculation of the solid-angle-weighted average of all intensities, the total luminous flux is obtained. This is an accurate but timeconsuming method. Thus, especially for quality control in the production process, an easier and more rapid method is required. With a sphere-shaped integrating photometer, called an Ulbricht’s sphere after the scientist who described the principle in 1900, a much more direct measurement of luminous flux is possible (CIE 1989, 2007). The hollow interior of the sphere is painted matt white so that it reflects the light of a lamp suspended in the interior of the sphere perfectly diffusely (Fig. 18.3). Every point of the interior surface reflects both the light arriving directly from the lamp and that arriving by way of inter-reflection from every other point of the sphere. Consequently, the illuminance on any part of the sphere’s interior wall due to reflected light is proportional to the luminous flux of the lamp in the sphere, independent of its light distribution. A small window in the wall, which is screened from direct lamp light by a small baffle, allows this illuminance to be measured and the luminous flux of the lamp to be calculated. In practice, the baffle and the lamp itself disturb part of the inter-reflections. Moreover, in addition to these disturbances, the matt-white paint will in practice not be completely uniform and not reflect

438

18 Calculations and Measurements

Fig. 18.3 Integrating photometer or sphere of Ulbricht with a diameter of 4 m

perfectly diffusely. Because of this, the integrating sphere has to be calibrated against a standard lamp of known luminous flux and, preferably, of comparable shape. To further limit measurement inaccuracies, the sphere dimensions have to be large relative to the lamp to be measured and relative to the window size required (integrating photometers with a diameter of 2–4 m are not unusual). Nevertheless, this is still too small for accurate measurements of absolute values of the luminous flux where large luminaires are concerned. However, for relative luminous flux measurements of luminaires for quality control reasons, they may be adequate.

18.2.2.2

Spectral Data

The spectrum of a light source is measured with a spectrophotometer. In a spectrophotometer the light from the light source entering the meter is dispersed into its spectral colours, that is to say into different wavelengths, by a diffraction grating that works as a prism (Fig. 18.4). In traditional spectrometers, a moving, unfiltered detector scans the dispersed radiation and so measures the energy radiated at the different wavelengths. From the energy spectrum obtained all the spectral parameter values can be calculated. Normally, the instruments have built-in software that makes the required calculations and directly yields the resulting values of the parameters directly. Many newer spectrometers employ an (unfiltered) CCD linear array detector instead of a single scanning detector. The light that is dispersed by the diffraction grating is projected onto the detector array so that in one measurement all information over the whole wavelength range is collected (see again Fig. 18.4). Such a spectrophotometer without moving parts is also available in portable versions. These allow the instrument to be used in the field to determine the spectral data with a reasonable degree of accuracy. Most such meters have embedded software that calculates from the measured spectrum all kind of colour characteristics such as chromaticity coordinates x and y, position in any chromaticity diagram, and CCT, Ra, Rf and Rg values.

18.2

Measurements

439

Fig. 18.4 Principle of a spectrophotometer with a detector array

detector array

entrance slit diffraction grating

These values can be displayed together with the spectrum itself on the meter immediately after the measurement. The light colour parameters CCT, Ra, Rf and Rg can be measured without a prism or diffraction grating, with a so-called colorimeter that has at least three filtered detectors that simulate the x, y and z CIE colour-matching functions. Often a fourth detector with a V(λ) filter is used in conjunction with the other three detectors. From the resulting chromaticity coordinates x and y, the values for correlated colour temperature and colour rendering are calculated (Schanda 2008).

18.2.3 Measuring Luminaires 18.2.3.1

Light Distribution

The luminous intensity distribution of a luminaire is the basis for all lighting application calculations. In the calculations, the luminaire is considered as a point source. Luminous intensity measurements, therefore, have to be done at a large distance from the luminaire. Inaccuracies will be negligible provided that the optical path length of the measuring set-up is ten times the length of the largest size of the light-emitting area of the luminaire (CIE 1996). Measurements satisfying this requirement are called far-field measurements. A set-up that enables the measurement of the luminous intensity together with the measurement of the angles of azimuth and elevation of each light direction of a luminaire (C- and gamma-angles) is called a goniophotometer. The simplest form is a set-up in which the luminaire intensities are measured by rotating the luminaire about its longitudinal and vertical axes, while the photocell remains in a fixed position (Fig. 18.5, left). However, with a luminaire rotated around its longitudinal axis, the burning position of the lamp changes which, with some gas-discharge lamps, affects the lamp’s light output and luminance distribution. This, in turn, may lead to unacceptable errors in the measured light distribution. A technique in which this problem is avoided is shown in Fig. 18.5, right, where the luminaire only rotates around its vertical axis while the photocell is moved over a spherical path (left part of the figure on the right). An advanced modification of this technique is that in which photodetectors are mounted at many locations on the spherical path. For the standard requirement of

440 Fig. 18.5 Schematic representation of goniophotometers. Left: luminaire is rotated about two axes while the photometer detector is stationary; right: luminaire remains stationary while the detector is moved over a spherical path

18 Calculations and Measurements

luminaire

detector array

detector

moving detector

Fig. 18.6 Example of a goniophotometer with the luminaire rotating about its horizontal position and with a rotating mirror that directs the light towards a stationary photocell

detector

luminaire

mirror

measurements to be made each 2.5 in the vertical plane, 72 detectors are required. They can be used in a circuitry that permits for simultaneous measurement of all detectors, so considerably reducing the total measuring time. However, the laboratory space then has to be very large. If only compact LED luminaires have to be measured, this technique can become feasible. To decrease the laboratory space needed, a rotating mirror or mirrors are generally employed. The principle is sketched in Fig. 18.6. Here the luminaire rotates, always in its horizontal position, to radiate its light towards a fixed photocell via a mirror mounted on a boom that rotates vertically. By the combination of the rotating luminaire and the rotating mirror, all relevant directions can be successively measured. The angular settings at which measurements are made must be precisely reproducible. Many different alternative solutions using a rotating mirror in conjunction with a moving, horizontal, luminaire are in practical use. Figure 18.7 shows a photograph of a typical commercial example.

18.2

Measurements

441

Fig. 18.7 Commercial photometer with a rotating mirror and fixed luminaire position

If the luminous flux of the lamp in the luminaire is known, only relative values are needed for the light distribution measurement. This means that unfiltered photocells can be used, thus avoiding inaccuracies due to incorrect V(λ) correction. Results from near-field photometry can be used to calculate the far-field light distribution (Ngai 1987). With the introduction of CCD matrix photodetectors, nearfield photometry requiring only a limited laboratory space has become feasible. In near-field photometry, the luminaire is not considered as a point source but as a combination of bright surfaces, each with its own photometric properties that can be characterised by their luminances in many different directions. Near-field goniophotometers map the luminance distribution of these light-emitting surface areas using a luminance meter with CCD detectors.

18.2.3.2

Light Output

As has been mentioned in the section about the measurement of luminous flux of lamps, the total luminous flux of a lamp (or luminaire) can be obtained from the calculation of the solid angle-weighted average of all luminous intensities. In fact, using high-quality goniophotometers for obtaining these luminous intensities offers the most accurate method for determining the luminous flux or light output of a luminaire. Since here absolute values are required, calibration against a light source with known luminous flux is needed.

442

18 Calculations and Measurements

18.2.4 Measuring Lighting Installations 18.2.4.1

Illuminance and Luminance Measurements

Generally speaking, where the aim is to determine the overall quality of an installation this has to be done by way of measurements of the quality parameters, namely the illuminances or luminances that are relevant for the type of application. Section 18.2.1 has described the detectors and correction devices that can be used for illuminance measurements. In the same section, it was explained that a luminance meter consists of a photocell with an optical system in front of it that projects an image of the scene to be measured onto the face of the photocell. For luminance measurements of large areas like walls or part of walls, the optical system can simply consist of a frame made by baffles (screens) that restricts the field of view of the photocell to only that large area. An illuminance meter with such frame mounted on top of its cell can be calibrated for making luminance measurements. A frame produced so that it has a field of view corresponding with a band with a width of 40 can be used for measuring in a simple way the average luminance in the B40 band. For spot luminance measurements, more sophisticated meters are required. The amount of light reflected from a small spot may be very small, so the meter has to be highly sensitive. To aim the meter accurately at the spot, a second optical system for aiming the meter is required. Large-megabyte CCD cameras with high spatial resolution allow for high-quality luminance measurements (Reinhard et al. 2005; Wienold and Christoffersen 2006; Rami and Lorge 2007; Bouroussis and Topalis 2009; Kruisselbrink et al. 2018). With this CCD camera-based luminance mapping technology, spatial measuring resolutions of well below 20 are possible. The perspective image of the scene to be measured is projected on the CCD’s pixel matrix. The signal of each pixel is proportional to the luminance of the corresponding scene element. Depending on the meter’s position and orientation relative to the scene, the perspective image can be converted through complex mathematical calculation software into a plane, non-perspective, image in which each pixel represents a small, same-sized scene area. From a single measurement, the software subsequently calculates the average scene luminance and the luminance uniformities. Cameras with incorporated luminance mapping software are also referred to as imaging luminance-measuring devices (ILMDs). In the next section, glare measurements making use of ILMD meters are discussed. Whatever the type of measurement to be made, a decision must be made concerning the number and positions of the measuring points: in other words, the measuring grid must be defined. Where good accuracy is required, for example for checking compliance with the specifications, it is recommended to use as many measuring points as specified in the specifications and used as basis for the calculations on which the design is based. Usually, a lighting design is based on an empty space. If measurements are done in a fully furnished space, the light levels are affected negatively because of obstruction and absorption of light by the furniture.

18.2

Measurements

18.2.4.2

443

Glare Measurements

UGR measurements can be done by measuring in a photometric laboratory the luminous intensity distribution and light-emitting area of a representative sample of the luminaire type used in the actual installation. The resulting photometric data is combined with the geometrical information of the installation to calculate either UGR for distinct observation positions and viewing directions or UGRL as being representative for the glare situation of the installation as a whole. The geometrical information concerns the position of the luminaires relative to the position of the observer. Sometimes it is desired to measure glare on location. For this purpose, imaging luminance-measuring devices (ILMDs) discussed in the previous section can be used (Wolf 2004; Hsu et al. 2012; Yamada and Kohko (2013); Porsch et al. 2015; Sawicki et al. 2019). From the position of the observer a luminance map of the actual room is made. Subsequently, processing algorithms identify and isolate the luminaires from the remaining part of the scene so that the luminaires can be treated as glare sources while the remaining part of the scene is treated as the background luminance. The processing algorithms can get both the photometrical data of the luminaires (luminances and light-emitting areas) and the geometrical positions of the luminaires. From these data, application algorithms can calculate UGR. All this can be done in real time providing near-instantaneous results. Figure 18.8 shows an example for an office with one single luminaire (left); middle: the luminance map with the glare source eliminated and right: the luminance map of the isolated glare source. By making an image map from different positions and for different viewing directions, UGR can be measured for the critical positions in a space. Using the luminance map to determine the UGR value of an installation, UGRL, is more complicated because it involves changing, in the calculation, the actual geometry of the luminaires to a reference geometry as explained in Sect. 4.4.2.2, “UGR Value

Fig. 18.8 Luminance mapping for measuring UGR for an office with one desk and one luminaire. Left: map of the whole office; middle: map with the glare source eliminated to determine the background luminance; right: map with the isolated glare source to determine the glare source luminance and light-emitting area. Red circles in the centre of the map: viewing direction. The geometrical position of the luminaire, needed for the determination of the position index, is also obtained from the map. Photograph: Christoph Schierz

444

18 Calculations and Measurements

for a Lighting Installation”. But with appropriate application algorithms this can also be done. As with all field measurements, it must be realised that there is a difference between the empty space (used for design purposes) and the furnished space. Where the glare source luminances are not or hardly affected by the furniture, the background luminance is likely to be lower than in the empty space because of screening of light by the furniture. It increases the measured value of UGR. Although laboratory ILMDs for glare measurements have been developed, glare meters are not yet commercially available. The International Lighting Committee CIE has Technical Committee TC 2-86 “glare measurement by imaging luminance measurement device” preparing a Technical Report to summarise the characteristics of ILMDs and best practices for measuring glare.

18.2.4.3

Spectral Data

Affordable portable spectrophotometers are commercially available which permit for the measurement of the spectrum of the light of a lighting installation radiated from the different light sources in a room. The price level is comparable with that of a high-quality lux meter. The principle of such spectrophotometers is discussed in Sect. 18.2.2.2.

18.2.5 Light-Logging Devices For the study of non-visual biological effects of light, it is often required to determine the light dose test persons receive under different circumstances and at different periods of the day or night. The light dose is the accumulated amount of light received over a certain period of time. For measuring light dose, light-logging devices are used that are worn on the body. The photocell of the device should preferably be able to measure all five α-opic irradiances or at least the photopic (lux) and melanopic irradiance. Logging devices are of the wrist-worn, chest-worn or spectacle type. The wrist-worn type is often combined with a movement sensor to also measure activity (action watch). The chest-worn type can be worn as a hanger or attached to the clothes. The spectacle type has the sensor mounted at the side of the glasses. All these devices do not exactly measure at the position and plane of the eye and therefore lead inherently to deviations. A study in which the measured radiation with different logging types was compared with the radiation measured at the position and plane of the eye showed, as expected, that the deviation is largest for the wrist-worn type and smallest for the spectacle type (Aarts et al. 2017). The mean deviation from the measured radiation at the eye was in this study for the wrist-worn type 26%, for the chest-worn type 17% and for the spectacle type 7%. The deviations in radiation dose, measured over a longer time, are smaller because of the levelling

References

445

influence of the varying positions and attitudes of the persons wearing the device (Figueiro et al. 2013; Aarts et al. 2017). The different types of logging device also differ in how they are experienced by the test persons (Van Duijnhoven et al. 2017). If a device is experienced as annoying or irritating, the risk exists that a test person does not wear it properly, or takes it off, without the research leader realising it. The type of device planned to be used should be tested on the actual test persons also in this respect.

References Aarts M, Van Duijnhoven J, Aries MBC, Rosemann A (2017) Performance of personally worn dosimeters to study non-image forming effects of light: assessment methods. Build Environ 117:60–72 Bouroussis CA, Topalis FV (2009) Automated luminance measurements of tunnel and road lighting installations. Proceedings of the 11th European Lighting Conference Lux Europe, Istanbul CIE (1989) International Commission on Illumination Publication 83:1989, The measurement of luminous flux. CIE, Vienna CIE (1996) International Commission on Illumination Publication 121:1996, The photometry and goniophotometry of luminaires. CIE, Vienna CIE (2007) International Commission on Illumination Publication 127:2007, Measurement of LEDs, 2nd edn, CIE, Vienna CIE (2013) CIE International Standard S 023/E:2013, Characterization of the performance of illuminance meters and luminance meters, Vienna Dilaura DL, Houser KW, Mistrick RG, Steffy GR (eds) (2011) The lighting handbook, 10th edn. Illuminating Engineering Society of North America, New York Figueiro MG, Hamner R, Bierman A, Rea MS (2013) Comparisons of three practical field devices used to measure personal light exposures and activity levels. Lighting Res Technol 45:421–434 Hsu S-W, Chen C-H, Jiaan Y-D (2012) Measurements of UGR of LED light by a DSLR colorimeter. Proceedings SPIE, 12th International conference on solid state lighting and 4th International conference on white LEDs and solid-state lighting 848415 Kruisselbrink T, van Duijnhoven J, Dangol R, Rosemann ALP (2018) Improving lighting quality by practical measurements of the luminance distribution. Proceedings of the 20th congress of the International Ergonomics Association (IEA 2018), vol X, pp 190–198 Ngai PY (1987) On near-field photometry. J Illum Eng Soc 2:129–136 Porsch T, Funke C, Schmidt F, Schierz Ch (2015) Measurement of the unified glare rating (UGR) based on using ILMD. Proceedings of the 28th CIE Session, Manchester, pp 536–542 Rami JP, Lorge G (2007) Road luminances dealt with a digital camera. CIE 26th Session, Beijing, pp 417–424 Reinhard E, Ward G, Pattanaik S, Debevec P (2005) High dynamic imaging, acquisition, display and image-based lighting. Morgan Kaufmann Publishers, San Francisco Sawicki D, Wolska A, Porsch T (2019) Glare assessment for research and development of measuring methods. Przeglad Elektrotechniczny R95(1):169–176 Schanda J (ed) (2008) Colorimetry: understanding the CIE system. John Wiley & Sons, Hoboken, NJ Van Duijnhoven J, Aarts MPJ, Aries MBC, Bömer MN, Rosemann ALP (2017) Recommendations for measuring non-image-forming effects of light: a practical method to apply on cognitive impaired and unaffected participants. Technol Health Care 25:171–186

446

18 Calculations and Measurements

Wienold J, Christoffersen J (2006) Evaluation methods and development of a new glare predictionmodel for daylight environments with the use of CCD cameras. Energy Build 38(7):742–756 Wolf S (2004) Entwicklung und Aubau eines leuchtdichte-Analysators zur Messung von Blendungskennzahlen, PHD thesis, Technische Universität Ilmenau, Germany Yamada T, Kohko S (2013) Glare evaluation system using imaging photometry. In: CIE centennial congress, vol x038. CIE, Paris, pp 627–633

Appendix A: Standardised Relative Spectral Sensitivity Values V(λ)

λ 400 405 410 415 420 425 430 435 440 445 450 455 460 465 470 475 480 485 490 495 CIE (1926)

V(λ) 0.000396 0.000640 0.001210 0.002180 0.004000 0.007300 0.011600 0.016840 0.023000 0.029800 0.038000 0.048000 0.060000 0.073900 0.090980 0.112600 0.139020 0.169300 0.208020 0.258600

λ 500 505 510 515 520 525 530 535 540 545 550 555 560 565 570 575 580 585 590 595

V(λ) 0.323000 0.407300 0.503000 0.608200 0.710000 0.793200 0.862000 0.914850 0.954000 0.980300 0.994950 1.000000 0.995000 0.978600 0.952000 0.915400 0.870000 0.816300 0.757000 0.694900

λ 600 605 610 615 620 625 630 635 640 645 650 655 660 665 670 675 680 685 690 695 700

© Springer Nature Switzerland AG 2019 W. van Bommel, Interior Lighting, https://doi.org/10.1007/978-3-030-17195-7

V(λ) 0.631000 0.566800 0.503000 0.441200 0.381000 0.321000 0.265000 0.217000 0.175000 0.138200 0.107000 0.081600 0.061000 0.044580 0.032000 0.023200 0.017000 0.011920 0.008210 0.005723 0.004102

447

Appendix B: Calculation of x–y Chromaticity Coordinates

The CIE 1931 standard colorimetric observer is defined by the set of three colourmatching functions x(λ), y(λ) and z(λ) given in Fig. 2.4. For the calculation, the values of these functions are given in tabular form in 5 nm wavelength intervals below (ISO/CIE 2007 and 2012). With these functions the tristimulus values X, Y and Z are calculated for a given spectral distribution Φ(λ), which should also be available in tabular form in the same wavelength intervals, from X¼

n X i¼1

Y¼

n X i¼1

Z¼

n X i¼1

ΦðλÞi  xðλÞi ΦðλÞi  yðλÞi ΦðλÞi  zðλÞi

The chromaticity coordinates x and y are then calculated with x¼

X XþY þZ

y¼

Y XþY þZ

© Springer Nature Switzerland AG 2019 W. van Bommel, Interior Lighting, https://doi.org/10.1007/978-3-030-17195-7

449

450

Appendix B: Calculation of x–y Chromaticity Coordinates

CIE Colour-Matching Functions λ (nm) 380 385 390 395 400 405 410 415 420 425 430 435 440 445 450 455 460 465 470 475 480 485 490 495 500 505 510 515 520 525 530 535 540 545 550 555 560 565 570 575 580

xðλÞ 0.0014 0.0022 0.0042 0.0076 0.0143 0.0232 0.0435 0.0776 0.1344 0.2148 0.2839 0.3285 0.3483 0.3481 0.3362 0.3187 0.2908 0.2511 0.1954 0.1421 0.0956 0.0580 0.0320 0.0147 0.0049 0.0024 0.0093 0.0291 0.0633 0.1096 0.1655 0.2257 0.2904 0.3597 0.4334 0.5121 0.5945 0.6784 0.7621 0.8425 0.9163

yðλÞ 0.0000 0.0001 0.0001 0.0002 0.0004 0.0006 0.0012 0.0022 0.0040 0.0073 0.0116 0.0168 0.0230 0.0298 0.0380 0.0480 0.0600 0.0739 0.0910 0.1126 0.1390 0.1693 0.2080 0.2586 0.3230 0.4073 0.5030 0.6082 0.7100 0.7932 0.8620 0.9149 0.9540 0.9803 0.9950 10.000 0.9950 0.9786 0.9520 0.9154 0.8700

zðλÞ 0.0065 0.0105 0.0201 0.0362 0.0679 0.1102 0.2074 0.3713 0.6456 10.391 13.856 16.230 17.471 17.826 17.721 17.441 16.692 15.281 12.876 10.419 0.8130 0.6162 0.4652 0.3533 0.2720 0.2123 0.1582 0.1117 0.0782 0.0573 0.0422 0.0298 0.0203 0.0134 0.0087 0.0057 0.0039 0.0027 0.0021 0.0018 0.0017

x10 ðλÞ 0.0002 0.0007 0.0024 0.0072 0.0191 0.0434 0.0847 0.1406 0.2045 0.2647 0.3147 0.3577 0.3837 0.3867 0.3707 0.3430 0.3023 0.2541 0.1956 0.1323 0.0805 0.0411 0.0162 0.0051 0.0038 0.0154 0.0375 0.0714 0.1177 0.1730 0.2365 0.3042 0.3768 0.4516 0.5298 0.6161 0.7052 0.7938 0.8787 0.9512 10.142

y10 ðλÞ 0.0000 0.0001 0.0003 0.0008 0.0020 0.0045 0.0088 0.0145 0.0214 0.0295 0.0387 0.0496 0.0621 0.0747 0.0895 0.1063 0.1282 0.1528 0.1852 0.2199 0.2536 0.2977 0.3391 0.3954 0.4608 0.5314 0.6067 0.6857 0.7618 0.8233 0.8752 0.9238 0.9620 0.9822 0.9918 0.9991 0.9973 0.9824 0.9556 0.9152 0.8689

z10 ðλÞ 0.0007 0.0029 0.0105 0.0323 0.0860 0.1971 0.3894 0.6568 0.9725 12.825 15.535 17.985 19.673 20.273 19.948 19.007 17.454 15.549 13.176 10.302 0.7721 0.5701 0.4153 0.3024 0.2185 0.1592 0.1120 0.0822 0.0607 0.0431 0.0305 0.0206 0.0137 0.0079 0.0040 0.0011 0.0000 0.0000 0.0000 0.0000 0.0000 (continued)

Appendix B: Calculation of x–y Chromaticity Coordinates λ (nm) 585 590 595 600 605 610 615 620 625 630 635 640 645 650 655 660 665 670 675 680 685 690 695 700 705 710 715 720 725 730 735 740 745 750 755 760 765 770 775 780

xðλÞ 0.9786 10.263 10.567 10.622 10.456 10.026 0.9384 0.8544 0.7514 0.6424 0.5419 0.4479 0.3608 0.2835 0.2187 0.1649 0.1212 0.0874 0.0636 0.0468 0.0329 0.0227 0.0158 0.0114 0.0081 0.0058 0.0041 0.0029 0.0020 0.0014 0.0010 0.0007 0.0005 0.0003 0.0002 0.0002 0.0001 0.0001 0.0001 0.0000

yðλÞ 0.8163 0.7570 0.6949 0.6310 0.5668 0.5030 0.4412 0.3810 0.3210 0.2650 0.2170 0.1750 0.1382 0.1070 0.0816 0.0610 0.0446 0.0320 0.0232 0.0170 0.0119 0.0082 0.0057 0.0041 0.0029 0.0021 0.0015 0.0010 0.0007 0.0005 0.0004 0.0002 0.0002 0.0001 0.0001 0.0001 0.0000 0.0000 0.0000 0.0000

zðλÞ 0.0014 0.0011 0.0010 0.0008 0.0006 0.0003 0.0002 0.0002 0.0001 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

451 x10 ðλÞ 10.743 11.185 11.343 11.240 10.891 10.305 0.9507 0.8563 0.7549 0.6475 0.5351 0.4316 0.3437 0.2683 0.2043 0.1526 0.1122 0.0813 0.0579 0.0409 0.0286 0.0199 0.0138 0.0096 0.0066 0.0046 0.0031 0.0022 0.0015 0.0010 0.0007 0.0005 0.0004 0.0003 0.0002 0.0001 0.0001 0.0001 0.0000 0.0000

y10 ðλÞ 0.8256 0.7774 0.7204 0.6583 0.5939 0.5280 0.4618 0.3981 0.3396 0.2835 0.2283 0.1798 0.1402 0.1076 0.0812 0.0603 0.0441 0.0318 0.0226 0.0159 0.0111 0.0077 0.0054 0.0037 0.0026 0.0018 0.0012 0.0008 0.0006 0.0004 0.0003 0.0002 0.0001 0.0001 0.0001 0.0000 0.0000 0.0000 0.0000 0.0000

z10 ðλÞ 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

Appendix C: RVP Model of Weston

Formulas to calculate, based on Weston (1953) and CIE (2002), Weston’s relative visual performance RVP values for different contrast values (C), background luminances (Lb), visual angle α and age (a). The actual meaning of some of the intermediate parameters to be calculated is given in reference CIE (2002). Here we treat these parameters just as intermediate calculation functions. For C  0.35 and α  1.5: X ¼ 0:119  ðlogLb þ 1:923Þ0:0840  ðC þ 1:516Þ0:655 Y ¼ 0:814  ðα  1:182Þ0:783  ðC þ 1:054Þ3:062 Z ¼ 0:575  ðlogLb þ 0:267Þ0:390  ðα  0:830Þ0:764 E ¼ ðC þ 0:199Þ0:148 þ 1:024   AF ¼ 1  1:317  104  ða  20ÞE RVP ¼ 0:930  ðα  1:499ÞX  ðlogLb þ 0:0920ÞY  ðC  0:253ÞZ  AF For C < 0.35 and α  1.5: X ¼ 0:082  ðlogLb þ 0:113Þ0:638  ðC þ 0:0224Þ0:23 Y ¼ 0:145  ðα  0:0041Þ0:185  ðC  0:099Þ0:117 Z ¼ 1:291  ðlogLb þ 0:264Þ0:387  ðα  0:218Þ0:523 E ¼ ðC þ 0:199Þ0:148 þ 1:024   AF ¼ 1  1:317  104  ða  20ÞE RVP ¼ 1:137  ðα  1:499ÞX  ðlogLb þ 0:035ÞY  ðC  0:0852ÞZ  AF

© Springer Nature Switzerland AG 2019 W. van Bommel, Interior Lighting, https://doi.org/10.1007/978-3-030-17195-7

453

Appendix D: RVP Model of Rea

Go through steps 1–7 to calculate relative visual performance RVP for different contrast values (Cv), adaptation luminances (LA) and age (a). The actual meaning of the intermediate parameters to be calculated in the different steps is given in reference (Rea and Ouellette 1991). Here we treat these parameters just as intermediate calculation functions. • Step 1 I R0 ¼ P  LA  π  r 2 where LA ¼ adaptation luminance in cd/m2 r ¼ 2:39  f1:22  tanhð0:3  logLA Þg P ¼ 1  0:017  ða  20Þ where a ¼ age of observer for 20–65 years • Step 2 C 0t, d ¼ ε  10ð1:3640:179A0:813Lþ0:226A

2

0:0772L2 þ0:169ALÞ

where A ¼ logtanhð20; 000  ωÞ

© Springer Nature Switzerland AG 2019 W. van Bommel, Interior Lighting, https://doi.org/10.1007/978-3-030-17195-7

455

456

Appendix D: RVP Model of Rea

  I R0 L ¼ loglog 10  π ε ¼ 1 þ ð0:00251Þ  ða  20Þ where ω ¼ area of the object in steradians • Step 3 2 2 K ¼ 10ð1:7630:175AA0:0310LLþ0:112AA Þþ0:171LL þ0:0622AALL

AA ¼ logtanhð5000  ωÞ   I 0 LL ¼ logtanh 0:04  R π • Step 4 Rmax ¼ 0:000196  logI R0 þ 0:0027 • Step 5 RT ¼

ΔC0:97 þ K 0:97 d ΔC 0:97  Rmax d

ΔC d ¼ C v  C 0t, d ; ΔC d > 0 Cv ¼

jLB  LT j jE  ρ=π  LT j ¼ LB LB

where: LB ¼ background luminance in cd/m2 (with uniform luminance LB ¼ LA) LT ¼ target (object) luminance in cd/m2 Lveil ¼ equivalent veiling luminance in cd/m2 E ¼ illuminance on object in lux ρ ¼ reflectance factor of object • Step 6 ΔT vis ¼ 305:4  RT • Step 7  RVP ¼ 0:998  RVP < 0 ! RVP ¼ 0

ΔT vis þ 800 777



RVP > 1 ! RVP ¼ 1

Appendix E: Evector/Escalar Ratio

Calculationbased on cubic illumination (Cuttle 1997) (Fig. E.1). E(x) and E(x) are two opposing illuminances on the x-axis E( y) and E(y) are two opposing illuminances on the y-axis E(z) and E(z) are two opposing illuminances on the z-axis 0 E(x) ¼ E(x)  E(x) is the magnitude of the illumination vector in the x-axis 0 E( y) ¼ E( y)  E(y) is the magnitude of the illumination vector in the y-axis 0 E(z) ¼ E(z)  E(z) is the magnitude of the illumination vector in the z-axis

z

Fig. E.1 Cubic illumination at a point

E(z) E(-x) y

E(y)

E(-y) E(x) E(-z)

© Springer Nature Switzerland AG 2019 W. van Bommel, Interior Lighting, https://doi.org/10.1007/978-3-030-17195-7

x

457

Appendix E: Evector/Escalar Ratio

458

jE j ¼

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 0 E ðxÞ2 þ 0 E ðyÞ2 þ 0 E ðzÞ2

with |E| is the magnitude of the illumination vector (as a result of all light incidences) eðxÞ ¼ 0 EðxÞ= j E j

eðyÞ ¼ 0 EðyÞ=jEj

eðzÞ ¼ 0 EðzÞ= j E j

with e(x, y, z) the direction of the illumination vector in the x, y, z coordinating system ~E(x) is the lesser of E(x) and E(x) ~E( y) is the lesser of E( y) and E(y) ~E(z) is the lesser of E(z) and E(z)

Esr ¼

jEj E ðxÞþ EðyÞþ EðzÞ þ 4 3

with Esr is the scalar illumination E vector =E scalar ratio ¼

jE j Esr

Appendix F: Position Indices, p

Position indices, p, are needed for the calculation of UGR. The parameters x, y and h0 refer to the rectangular coordinating system shown in Fig. 4.22 (CIE 1995)

© Springer Nature Switzerland AG 2019 W. van Bommel, Interior Lighting, https://doi.org/10.1007/978-3-030-17195-7

459

h0 /y x/y 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 1.1 1.20 1.30 1.40 1.50 1.60 1.70 1.80

0.00 1.00 1.05 1.12 1.22 1.32 1.43 1.55 1.70 1.82 1.95 2.11 2.30 2.40 2.55 2.70 2.85 2.95 3.10 3.25

0.10 1.26 1.22 1.30 1.38 1.47 1.60 1.72 1.88 2.00 2.20 2.40 2.55 2.75 2.90 3.10 3.15 3.40 3.55 3.70

0.20 1.53 1.46 1.50 1.60 1.70 1.82 1.98 2.12 2.32 2.54 2.75 2.92 3.12 3.30 3.50 3.65 3.80 4.00 4.20

0.30 1.90 1.80 1.80 1.87 1.96 2.10 2.30 2.48 2.70 2.90 3.10 3.30 3.50 3.70 3.90 4.10 4.25 4.50 4.65

0.40 2.35 2.20 2.20 2.25 2.35 2.48 2.65 2.87 3.08 3.30 3.50 3.72 3.90 4.20 4.35 4.55 4.75 4.90 5.10

0.50 2.86 2.75 2.66 2.70 2.80 2.91 3.10 3.30 3.50 3.70 3.91 4.20 4.35 4.65 4.85 5.00 5.20 5.40 5.60

0.60 3.50 3.40 3.18 3.25 3.30 3.40 3.60 3.78 3.92 4.20 4.40 4.70 4.85 5.20 5.35 5.50 5.75 5.95 6.10

0.70 4.20 4.10 3.88 3.90 3.90 3.98 4.10 4.30 4.50 4.75 5.00 5.25 5.50 5.70 585 6.20 6.30 6.50 6.75

0.80 5.00 4.80 4.60 4.60 4.60 4.70 4.80 4.88 5.10 5.30 5.60 5.80 6.05 6.30 6.50 6.80 7.00 7.20 7.40

0.90 6.00 5.80 5.50 5.45 5.40 5.50 5.50 5.60 5.75 6.00 6.20 6.55 6.70 7.00 7.25 7.50 7.65 7.80 8.00

1.00 7.00 6.80 6.50 6.45 6.40 6.40 6.40 6.60 6.60 6.75 7.00 7.20 7.50 7.70 8.00 8.20 8.40 8.50 8.65

1.10 8.10 8.00 7.60 7.40 7.30 7.30 7.35 7.40 7.50 7.70 7.90 7.16 8.30 8.55 8.70 8.85 9.00 9.20 9.35

1.20 9.25 9.10 8.75 8.40 8.30 8.30 8.40 8.50 8.60 8.70 8.80 9.00 9.20 9.35 9.50 9.70 9.80 10.00 10.10

1.30 10.35 10.30 9.85 9.50 9.40 9.40 9.40 9.50 9.50 9.65 9.75 9.90 10.00 10.20 10.40 10.55 10.80 10.85 11.00

1.40 11.70 11.60 11.20 10.85 19.60 10.50 10.50 10.50 10.60 10.75 10.80 10.95 11.02 11.20 11.4 11.50 11.75 11.85 11.90

1.50 13.15 13,00 12.70 12.10 11.90 11.75 11.70 11.70 11.75 11.80 11.90 12.00 12.10 12.25 12.40 12.50 12.60 12.75 12.80

1.60 14.70 14.60 14.00 13.70 13.20 13.00 13.00 12.85 12.80 12.90 12.95 13.00 13.10 13.20 13.25 13.30 13.40 13.45 13.50

1.70 16.20 16.10 15.70 15.00 14.60 14.40 14.10 14.00 14.00 14.00 14.00 14.00 14.00 14.00 14.05 14.05 14.20 14.20 14.20

1.90

16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00

1.80

16.00 15.70 15.40 15.20 16.10 15.00 15.00 15.00 15.00 15.00 15.00 15.02 15.10 15.10 15.10

460 Appendix F: Position Indices, p

1.90 2.00 2.10 2.20 2.30 2.40 2.50 2.60 2.70 2.80 2.90 3.00

3.43 3.50 3.60 3.75 3.85 3.95 4.00 4.07 4.10 4.15 4.20 4 22

3.86 4.00 4.17 4.25 4.35 4.40 4.50 4.55 4.60 4.62 4.65 4.67

4.30 4.50 4.65 4.72 4.80 4.90 4.95 5.05 5.10 5.15 5.17 5.20

4.75 4.90 5.05 5.20 5.25 5.35 5.40 5.47 5.53 5.56 5.60 5.65

5.20 5.35 5.50 5.60 5.70 5.80 5.85 5.95 6.00 6.05 6.07 6.12

5.70 5.80 6.00 6.10 6.22 6.30 6.40 6.45 6.50 6.55 6.57 6.60

6.30 6.40 6.60 6.70 6.80 6.90 6.95 7.00 7.05 7.08 7.12 7.15

6.90 7.10 7.20 7.35 7.40 7.50 7.55 7.65 7.70 7.73 7.75 7.80

7.50 7.70 7.82 8.00 8.10 8.20 8.25 8.35 8.40 8.45 8.50 8.55

8.17 8.30 8.45 8.55 8.65 8.80 8.85 8.95 9.00 9.05 9.10 9.12

8.80 8.90 9.00 9.15 9.30 9.40 9.50 9.65 9.60 9.65 9.70 9.70

9.50 9.60 9.75 9.85 9.90 10.00 10.05 10.10 10.16 10.20 10.23 10.23

10.20 10.40 10.50 10.60 10.70 10.80 10.85 10.90 10.92 10.95 10.95 10.95

11.00 11.10 11.20 11.30 11.40 11.50 11.55 11.60 11.63 11.65 11.65 11.65

12.00 12.00 12.10 12.10 12.20 12.25 12.30 12.32 12.35 12.35 12.35 12.35

12.82 12.85 12.90 12.90 12.95 13.00 13.00 13.00 13.00 13.00 13.00 13.00

13.55 13.60 13.70 13.70 13.70 13.75 13.80 13.80 13.80 13.80 13.80 13.80

14.20 14.30 14.35 14.40 14.40 14.45 14.50 14.50 14.50 14.50 14.50 14.50

15.10 15.10 15.10 15.15 15.20 15.20 15.25 15.25 15.25 15.25 15.25 15.25

16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00

Appendix F: Position Indices, p 461

Appendix G: Groningen Sleep Quality Scale

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

I had a deep sleep last night I feel like I slept poorly last night It took me more than half an hour to fall asleep last night I felt tired after waking up this morning I woke up several times last night I feel like I didn’t get enough sleep last night I got up in the middle of the night I felt rested after waking up this morning I feel like I only had a couple of hours of sleep last night I feel I slept well last night I didn’t sleep a wink last night I didn’t have any trouble falling asleep last night After I woke up last night, I have trouble falling asleep again I tossed and turned all night last night I didn’t get more than 5 h of sleep last night

True True True True True True True True True True True True True True True

False False False False False False False False False False False False False False False

Scoring: The first question doesn’t count towards the total score One point if answer is “True” for questions 2, 3, 4, 5, 6, 7, 9, 11, 13, 14, 15 One point if answer is “false” for questions 8, 10, 12 Maximum score of 14 points indicates poor sleep the night before Note: The Groningen Sleep Quality Scale (Meijman et al. 1988) is a tool that can be used to understand your patterns in overall sleep quality. Answer these 15 questions for at least 14 days in a row to help understand your individual sleep patterns

© Springer Nature Switzerland AG 2019 W. van Bommel, Interior Lighting, https://doi.org/10.1007/978-3-030-17195-7

463

Appendix H: Normalised Formula for SVM according to CIE (2016)

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi X Ci 3:7 SVM ¼ Ti 3:7

SVM ¼ stroboscopic visibility measure Ci ¼ the amplitude of the ith Fourier component Ti ¼ the visibility threshold for a sine wave at the frequency of the ith Fourier component T ¼ 1/Stroboscopic sensitivity of Fig. 10.8 T as formula: T i ðf Þ ¼

1 þ 20ef =10 1 þ eaðf bÞ

f ¼ frequency of the ith Fourier component in Hz a ¼ 0.00518 b ¼ 306.6

© Springer Nature Switzerland AG 2019 W. van Bommel, Interior Lighting, https://doi.org/10.1007/978-3-030-17195-7

465

References

CIE (1926) Proceedings CIE 6th Session 1924, Geneva. Recueil des Traveaux et Compte Rendu de Séances. University Press, Cambridge, pp 67–69 CIE (1995) International Commission on Illumination CIE Publication 117:1995, Technical report, 1221 Discomfort glare in interior lighting, Vienna CIE (2002) Publication 145:2002 The correlation of models for vision and visual performance. CIE, Vienna CIE (2016) CIE technical note CIE TN 006:2016 Visual aspects of time-modulated lighting systems—definitions and measurement models, Vienna Cuttle C (1997) Cubic illumination. Lighting Res Technol 29(1):1–14 ISO/CIE (2007) Publication ISO 11664-1:2007/CIE S 014-1:2006 Joint ISO/CIE Standard: colorimetry—Part 1: CIE standard colorimetric observers ISO/CIE (2012) Publication ISO 11664-3:2012/CIE S 014-3:2011 Joint ISO/CIE Standard: colorimetry—Part 3: CIE tristimulus values Meijman TF, De Vries-Griever AH, De Vries G, Kampman R (1988) The evaluation of the Groningen sleep quality scale. Heymans Bulletin (HB 88-13-EX), Groningen Rea MS, Ouellette MJ (1991) relative visual performance: a basis for application. Lighting Res Technol 23(3):135–144 Weston HC (1953) The relation between illumination and visual performance, Reprint IHRB Rep. No. 87 (1945) and Joint Rep. (1935). Medical Research Council, HMSO, London

© Springer Nature Switzerland AG 2019 W. van Bommel, Interior Lighting, https://doi.org/10.1007/978-3-030-17195-7

467

Author Index

A Aarts, M., 444, 445 Achermann, P., 171 Adelson, E.H., 20 Agogino, A.M., 351 Akashi, Y., 92, 125, 354, 355 Åkerstedt, T., 171, 176, 189 Albrecht, U., 222 Allison, S., 216 Alpern, M., 82 Alterman, T., 187 Ancoli-Israel, S., 215, 216 Anikeeva, P.O., 297 Ao-Thongthip, S., 129 Archer, S.N., 171, 188, 253 Aries, M., 172, 173, 181 Arnon, S., 381 Aschoff, J., 139, 141, 143, 144, 149 Atamian, H.S., 139

B Bae, G., 42 Bailes, H.J., 158 Bailey, I., 65 Bakker, I., 176 Balsalobre, A., 143 Bargary, G., 115 Bargiello, T.A., 143 Barkmann, C., 419 Basner, M., 190 Baumgartner, H., 272 Beauchamp, M.T., 227 Begemann, S., 292 Behar-Cohen, F., 250

Bell, A.G., 376 Benedetti, F., 228 Benloucif, S., 216 Bennett, C.A., 213 Berman, S.M., 65, 82–85, 90, 94, 115 Bernhofer, E.I., 228 Berrutto, V., 100 Berson, D.M., 19, 144, 147 Bevacqua, S.F., 292 Bhattacharya, I., 381 Birch, D.G., 79 Bizjak, G., 161 Blackwell, H.R., 63–65, 71, 80 Blask, D.E., 254, 255 Bloch, K.E., 215 Bodington, J.D., 238, 239 Bodmann, H.W., 68, 71–73, 90, 128 Bodrogi, P., 125 Boivin, D.B., 141, 142, 189, 195 Borbély, A.A., 170, 171 Born, J., 170 Boubekri, M., 172 Boudreau, P., 188, 190, 191 Bouma, H., 82 Bouma, P.J., 31 Bouman, M.A., 61, 62 Bouroussis, C.A., 442 Boyce, P.R., 21, 46, 51, 65, 74, 80, 92, 108, 126, 129, 177, 190, 191, 354, 356, 414, 424 Brainard, G.C., 153–155, 157 Brandenberger, G., 188 Brandston, H.M., 90 Broszio, K., 149 Brown, S.A., 143, 222 Bruce, V.G., 142

© Springer Nature Switzerland AG 2019 W. van Bommel, Interior Lighting, https://doi.org/10.1007/978-3-030-17195-7

469

470 Buijs, R.M., 150 Bullough, J.D., 125, 237, 239, 240, 250 Burgess, H.J., 146, 192, 195 Buysse, D.J., 171

C Cai, H., 121 Caicedo, D., 351, 352, 375 Cajochen, C., 151, 176, 191 Campbell, F.W., 82 Campbell, J.H., 236 Canazei, M., 172, 178, 198, 199 Canellas, F., 228 Cappuccio, F.P., 169 Cardinali, D.P., 215 Carrillo-Vico, A., 254 Cassone, V.M., 142 Cermakian, N., 189 Chang, A.M., 157, 191, 194, 195 Chapdelaine, S., 191 Cheal, C., 356 Chellappa, S.L., 152 Chen, S.K., 19, 151 Cheung, I.N., 169, 222 Chew, E.Y., 210 Chichilnisky, E.J., 18 Cho, J., 295 Choi, J.H., 228 Chovnick, A., 138 Chraibi, S., 96 Christie, A.W., 212 Christoffersen, J., 325, 442 Chung, T., 121 Clear, R.D., 65, 85, 121 Coaton, J.R., 21, 91, 278, 344 Cohen, M., 254 Colau, A., 181 Conway, B.R., 18 Coogan, A.N., 226 Cooke, J.R., 216 Cornélissen, G., 215 Costa, G., 188, 253 Costa, I.C., 169 Cox, D.D., 20 Crawford, B.H., 77 Creemers, P.T.J., 355 Crowley, S.J., 191, 192, 195 Cseh, T., 377 Curcio, C.A., 209 Cuttle, C., 21, 22, 102–106, 108–112, 114, 129, 457 Czeisler, C.A., 141–143, 192

Author Index D da Vinci, L., 4, 5 Daan, S., 170 Dacey, D.M., 158 Dai, X., 297 Daneault, V., 217 Danz, N., 309 Daurat, A., 191 David, A., 42, 52 Davis, R.G., 129 Davis, W., 42, 46, 356 Dawson, D., 190 De Beer, E., 42 De Boer, J.B., 79, 95, 111, 128, 417 De Candolle, A., 139 De Kort, Y.A.W., 181 De Lange, H., 237 De Mairan, J.J.O., 139 De Valois, R.L., 17 De Vries, H.J.A., 95, 100 Debney, L.M., 243 Dickmeis, T., 151 Dijk, D.-J., 169, 171, 189, 216 Dikel, E.E., 37, 129 DiLaura, D.L., 80, 228, 402, 407, 409 Dimitrov, S., 377, 383 Djordjevoic, I.B., 377 Do, M.T.H., 148 Donners, M.A.H., 115 Drake, C.L., 189 Driver, J., 5 Duff, J., 100, 103–106 Duffy, J.F., 143, 216 Dumont, M., 188, 194, 195 Dunlap, J.C., 147 Dyble, M., 311

E Eastman, A.A., 236 Eastman, C.I., 146, 188, 190, 192 Eble-Hankin, M., 121, 404 Eckel-Mahan, K., 253 Ehrman, M., 183 Eklund, N.H., 75 Ely, E., 229 Emery, P., 143 Enezi, J., 392 Enoch, J.M., 79 Escuyer, S., 356 Esposito, T., 51 Estrup, S., 229

Author Index F Fairchild, M.D., 35 Fan, J.J., 272 Fan, X.J., 293 Fargason, R.E., 227 Faria, M., 380 Farnsworth, D., 50 Farr, C.A., 127, 129 Fechner, G.T., 90 Fein, M., 37 Fekete, J., 125 Ferlazzo, F., 181 Fetveit, A., 216 Ficca, G., 183 Figueiro, M.G., 155–157, 172, 174, 175, 177, 182, 192, 215, 224, 225, 419, 445 Figueroa-Ramos, M.I., 229 Fischer, D., 79, 95, 109, 111, 128, 417 Fisher, A.J., 212 Fisher, R., 242 Flynn, J.E., 95, 100, 111 Focan, C., 222 Folkard, S., 171, 188, 190 Fonken, L.K., 253 Fontoynont, M., 356 Foster, R.G., 141, 143, 147, 158 Fotios, S.A., 93, 94, 129, 130, 181, 356 Freedman, M.S., 147 Freyssinier-Nova, J.P., 42 Fritschi, L., 257 Frost, P., 188, 253 Fry, G.A., 115 Funke, C., 123

G Galasiu, A.D., 356 Gamlin, P.D., 19, 82, 158 Geerdinck, L.M., 121 Gentile, N., 419 Gershun, A., 109, 112, 113 Geyer, U., 345 Ghassemlooy, Z., 377, 381 Gifford, R., 178 Gillberg, M., 176 Giménez, M.C., 162, 228 Ginsburg, L.M., 64 Ginthner, D.N., 129 Girgert, R., 254 Glickman, G., 149 Golden, R.N., 223 Gooley, J.J., 157 Gordijn, M.C.M., 223 Gorgels, G.M.F., 248

471 Gornicka, G.B., 181 Govén, T., 419 Grangaard, E.M., 418 Grimes, K.J., 210 Gronfier, C., 195 Gu, H.T., 45 Guild, J., 28 Güler, A.D., 19 Guth, S.K., 116, 117, 126

H Haas, H., 377, 382, 383 Ham, W.T., 247 Han, S., 129 Hanford, N., 224 Hanifin, J.P., 157 Hankins, H.W., 157 Hanselaer, P., 42 Hara, N., 122 Harada, T., 18 Hardeland, R., 215 Harding, G.F.A., 243 Härmä, M., 189 Hasegawa, S., 122 Hashimoto, K., 42 Hattar, S., 148, 149, 157 He, L., 297 Hébert, M., 157 Henderson, B.A., 210 Hering, E., 18 Hewitt, H., 108 Hill, S.M., 254 Hoekstra, J., 62 Holladay, L.L., 77 Holonyak, N. Jr., 292 Hood, B., 216 Hopkins, S., 217, 224 Hopkinson, R.G., 116 Horowitz, T.S., 192 Houser, K.W., 42, 45, 51, 85, 95, 96, 111 Howarth, P.A., 115 Hoyle, N.P., 139 Hsu, S.-W., 443 Hu, W., 356 Huang, A., 113 Huang, C., 378 Huang, W.J., 125 Hubalek, S., 172–174, 181 Hubel, D.H., 18 Huiberts, L.M., 177, 178, 180 Hunt, R.W.G., 34, 46 Hunter, J.J., 246, 247 Hysing, M., 226

472 I Ijaz, S., 256 Inoue, Y., 96 Ishii, H., 181 Iskra-Golec, I., 181 Islam, M.S., 129 Islim, M.S., 377

Author Index

J Jackson, G.R., 213 James, F.O., 188, 190, 195 James, S.M., 189 Jansen, J., 331 Jasser, S.A., 157 Jeavons, P., 243 Jhalani, V.A., 296 Jiménez, R.P., 380 Joarder, A.R., 228 Johnson, E.N., 18 Johnstone, B., 292 Jones, B.E., 152 Juslén, H.T., 108 Just, M.A., 85

Klein, D.C., 149 Klerman, E.B., 145, 216 Knab, B., 139, 141 Knez, I., 108, 181 Knight, J.A., 215 Knoop, M., 149 Knower, K.C., 254 Knudsen, E.I., 5 Kobav, M.B., 161 Kobayashi, R., 216 Kobayashi, S., 108 Kock, A., 292 Koga, Y., 121 Kolb, H., 8 Koonen, A.M.J., 383 Kosslyn, S.M., 20 Kozaki, T., 157 Kreitzman, L., 141, 143 Kruisselbrink, T., 442 Kruithof, A.A., 127–129 Kuehni, R.G., 93 Kuffler, S.W., 13 Küller, R., 244 Kumbalasiri, T., 148

K Kaida, K., 176, 178, 183 Kalsbeek, A., 150 Kamdar, B.B., 256 Kanakri, S.M., 419, 420 Kantermann, T., 253, 257 Karasek, M., 215 Karlicek, R.F., 375, 381 Kartashova, T., 113 Kato, M., 107 Kaur, G., 139 Kavehrad, M., 380 Kayumov, L., 197 Kazemi, R., 190 Keis, O., 419 Kelly, D.H., 237 Kepler, J., 4 Kers, C., 108, 181 Khanh, T.C., 293 Kim, M.H., 296 Kim, W., 121 Kim, Y., 169 King, S., 356 King, V.M., 115 Kini, S.G., 272 Kioupakis, E., 296 Kirsch, R.M., 98–100, 104, 107, 127 Kittler, R., 318, 325

L La Toison, M., 90 Laike, T., 244 Lall, P., 272 Lasko, T.A., 149 Lassane, C.J.M., 299, 347 Lee, C.M., 121 Lee, J.-G., 309 Lee, K.A., 176 Leger, D., 178 Lerner, A.B., 150 Li, C., 42 Li, S., 226 Li, Z.-Y., 296, 312 Lieverse, R., 223 Lin, Y., 38 Lipetz, L.E., 66 Lockley, S.W., 151 Loe, D.L., 68, 95–102, 104, 318 Logadóttir, A., 356 Lorge, G., 442 Lou, D.L., 123 Lowden, A., 151 Lucas, R.J., 19, 148, 151, 157, 158 Luckiesh, M., 116, 117, 126 Lundgren, J.D., 227 Luo, M.R., 35 Lynes, J.A., 21, 22, 80, 109, 113

Author Index M MacAdam, D.L., 32 Macaluso, E., 5 Mahler, H., 51 Mann, M.D., 9 Manov, B., 96 Mansfield, K.P., 95, 97, 100 Mardaljevic, J., 316, 325 Markwell, E.L., 82 Marrelec, G., 5 Marsden, A.M., 21, 90–92, 278, 344 Martin, S.K., 190 Mathews, E., 374 Mawad, K., 154 McColl, S.L., 244 McDougal, D.H., 19, 82 McIntyre, I.M., 151 McMurrich, J.P., 4 McNelis, J., 68 Mednick, S.C., 183 Meesters, Y., 223 Mehrabian, A., 176 Meijman, T.F., 171 Mély, D.A., 18 Meneghini, M., 272 Meyer, C., 278 Miller, N.J., 100 Mills, P.M., 181 Miñano, J.C., 345 Minerva, R., 372 Mishima, K., 215, 222 Mochizuki, E., 213 Mondriaan, P., 200 Moore, R.Y., 142, 149 Moore, T., 356 Morgenstern, Y., 111 Morin, L.P., 149 Morris, C.J., 253 Morrow, B.L., 419, 420 Mott, M.S., 418, 419 Muck, E., 68, 71 Muneer, T., 325 Munsell, A.H., 26 Mure, L.S., 154 Murray, R.F., 115 Mury, A.A., 112, 113

N Najjar, R., 215 Naka, K.I., 66 Nakamura, S., 292 Navvab, M., 83

473 Neches, J., 354, 355 Newman, L.A., 157 Newsham, G.R., 95, 100, 244, 353–356, 375 Ngai, P.Y., 112, 126, 441 Niedling, M., 125 Nienhuis, H., 278 Nightingale, F., 227 Noguchi, H., 181 Novotny, P., 418

O Odds, W., 169 Oh, S., 37 Ohno, Y., 37, 42, 46 Oi, N., 95, 100 Oren, D.A., 215 Organisciak, D.T., 246, 247 Oster, H., 188, 253 Ouellette, M.J., 74–76, 455 Ouweltjes, J.L., 39 Owsley, C., 207

P Padmasali, A.N., 272 Paech, G.M., 189 Palmer, C.R., 224 Pandajonas, S., 209 Pandharipande, A., 351, 353, 375 Park, N.K., 127, 129 Park, S.P., 209 Patel, J., 229 Pauers, M.J., 145 Paus, S., 226 Peirson, S., 158 Perera, S., 223 Peruffo, A., 351 Perz, M., 37, 238–241 Philip, P., 216 Phipps-Nelson, J., 151, 178 Piazena, H., 149 Pilat, L., 352 Pittendrigh, C.S., 142 Plischke, H., 418 Pointer, M.R., 34, 46 Polin, D., 311 Pont, S.C., 113 Poppe, A., 299, 347 Porsch, T., 443 Price, A.D.F., 228 Priest, I.G., 128 Protzman, J.B., 111

474 Provencio, I., 19, 148 Pruessner, J.C., 222 Pustjens, T., 229

R Radke, R.J., 381 Rahman, S.A., 197, 198 Rami, J.P., 442 Rasch, B., 170 Rautkylä, E., 152, 419 Raynham, P., 107 Rea, M.S., 42, 46, 65, 68, 72, 74–76, 151, 154–157, 181, 419, 455 Refinetti, R., 145, 221, 222 Regente, J., 192, 197, 198 Reinhard, E., 442 Reiter, R.J., 215, 254 Revell, V.L., 144, 216 Richter, C., 147 Riemersma-Van Der Lek, R.F., 216, 222, 224 Rodriguez, C., 254 Roenneberg, T., 143–145, 253, 257 Rosenthal, N.E., 144, 223, 224 Roth, T., 189 Rouch, I., 190 Round, H.J., 292 Rowlands, E., 100 Royer, M.P., 43, 51, 95, 269 Rubinstein, F., 312 Rüger, M., 149, 178, 192 Rushton, W., 66 Russell, J.A., 176 Rybak, Y.E., 227

S Sack, R.L., 192, 215 Safdar, M., 115 Sakaguchi, T., 181 Sam, C., 418 Sanderson, A., 113 Santhi, N., 181, 191–196 Saper, C.B., 152 Sasseville, A., 197 Sassone-Corsi, P., 253 Sawicki, D., 213, 443 Schanda, J., 26, 35, 36, 90, 439 Scheir, G.H., 115, 121, 123, 124 Scherder, E., 222 Schernhammer, E.S., 256 Schierz, C.H., 123 Schubert, E.F., 293

Author Index Schwab, R.N., 213 Schwartz, J.R.L., 189 Sekiguchi, K., 107 Sekuler, R., 207 Sephton, S.E., 222 Serre, T., 18 Shang, X., 352 Sharpe, L.Y., 31 Shepherd, A.J., 243 Shi, L., 181 Shochat, T., 215 Simons, K., 229 Simons, R.H., 51 Sinoo, M.M., 215 Skene, D.J., 215, 216 Sleegers, P.J.C., 419 Sletten, T.L., 216 Sliney, D.H., 246, 249 Sliwinski, T., 254 Smet, K., 42, 45 Smith, M.R., 189, 191, 195, 196 Smith, S., 68 Smolders, K.C.H.J., 178–182 Smolensky, M.H., 222 Smyth, V.O., 243 Söllner, G., 116 Song, H., 209 Souêtre, E., 222 St Hilaire, M.A., 195 Stampi, C., 183 Steckel, J., 297 Stephan, F.K., 142 Stevens, S.S., 90 Stevenson, R., 296 Stockman, A., 31 Stone, P.T., 115 Swaab, D.F., 215 Swanston, M., 90

T Taguchi, T., 229 Takahashi, H., 121, 127 Takahashi, M., 183 Takahashi, Y., 19, 151 Talapin, D.V., 297 Tashiro, T., 121–123 Teclemariam-Mesbah, R., 149 Terman, J.S., 223 Terman, M., 144, 222, 223 Teunissen, C., 38, 45 Thapan, K., 153–155 Theeuwes, J., 213

Author Index Thorn, L., 151 Thornton, W.A., 51 Tian, Z., 378 Tiller, D.K., 95 Tokura, H., 228 Topalis, F.V., 442 Tovée, M.J., 8 Travis, R.C., 256 Tregenza, P., 100, 316, 318, 323–325 Tsongos, N.F., 213 Turnage, J.J., 177 Turner, E.J.D., 225 Turner, P.L., 210, 215

475

V Valberg, A., 237 Van Bommel, W.J.M., 149, 161, 182, 214, 340 Van Cauter, E., 222 Van De Kraats, J., 210 Van De Werken, M., 197, 198 Van Der Burgt, P.J.M., 42, 48, 52 Van Diepen, H.C., 145, 157 Van Driel, W.D., 293 Van Duijnhoven, J., 445 Van Gelder, R.N., 154 Van Kemenade, J.T.C., 42, 48 Van Nes, F.L., 61, 62 Van Norren, D., 210, 248 Van Ooyen, M.H.F., 96, 100 Van Someren, E.J.W., 222, 224–226 Van Someren, K.I., 356 Vanderwalle, G., 151, 152, 178 Vaughan, D.K., 246, 247 Veitch, J.A., 89, 95–97, 100, 244, 356, 414 Vetter, C., 188 Vidovszky-Nemeth, A., 90 Viénot, F., 90, 129 Viola, A.U., 172, 181 Visser, E.K., 149 Von Kries, J.A., 41 Von Schantz, M., 189 Vos, J.J., 212

Walmsley, L., 145 Walraven, J., 8 Wams, E.J., 171, 172 Wandell, B.A., 18 Wang, K., 345 Wang, L., 239–241 Wang, X.S., 188, 253 Wang, Y., 37 Waters, C., 404 Waters, I.M., 68 Watson, A.B., 9, 211 Watts, C., 424 Weale, R.A., 75, 207 Wegrzyn, L.R., 256 Wei, J., 272 Wei, M., 37, 43, 129 Weibel, L., 188 Weiss, B., 229 Wen, Y.-J., 351 Werner, J.S., 207 Wertli, J., 352 Westerlund, A., 171 Westheimer, G., 62 Weston, H.C., 8, 66–68, 70–73, 75, 76, 78, 453 Wetterberg, L., 179 White, M.D., 422 Whitehead, L., 42 Wienold, J., 442 Wilkins, A.J., 115, 242–244, 419 Wilkinson, R.T., 216 Williams, A., 352 Willis, G.L., 225, 226 Wilson, M., 323–325 Winget, C.M., 147 Winter, A.L., 243 Winterbottom, M., 419 Wirz-Justice, A., 170, 223 Witten, I.B., 5 Woelders, T., 147 Wohlfarth, K., 418 Wolf, S., 443 Wolska, A., 213 Woodstock, T.-K., 381 Wright, K.P., 191 Wright, W.D., 28 Wuerger, S., 18 Wynchank, D.S., 226

W Wade, N.J., 90 Wakamura, T., 228 Walch, J.M., 228

X Xia, L., 111–114, 121, 126 Xiao, K., 18 Xu, W., 149

U Uttley, J., 356

476 Y Yamazaki, S., 144 Yang, Y., 121, 122, 124, 125 Yaodong, C., 127 Ye, M., 181 Yellott, J.I., 211 Yilmaz, F.S., 356 Yin, J., 169

Author Index Z Zaal, I.J., 229 Zeitzer, J.M., 157, 192, 215 Zhang, W., 380 Zhang, X., 378 Zhu, J., 297 Zhu, L., 345 Zucker, I., 142 Zulley, J., 139, 141

Subject Index

A AC-LED, 303 Action spectrum, 152–157 Action watch, 147, 171, 225, 444 Adaptation, 8, 19 Additive colour mixing, 17 Adrenal cortex, 149–151 Age alertness, 216 blue light loss effect, 209 circadian rhythm, 216 eye transparency, 209 increased glare, 212 melatonin concentration, 215 pupil size, 209 reduced pathway signal, 209, 215 reduction of pupil size, 210 sleep, 216 visual performance, 70 yellowing of eye lens, 209 Alertness, 141, 151, 176, 190, 193 Alpha-opic, 159, 163 Alzheimer’s disease, 224, 423 Amalgam, 282, 284 Ambient-temperature sensitivity, 306 Ambient temperature Ta, 347 American National Standards Institute (ANSI), xix ANSI/IES Design Guide RP-1-12, 402 Antioxidant properties of melatonin, 254 Arousal, 176, 190 Artificial intelligence (AI), 373 Attention-deficit hyperactivity disorder (ADHD), 226 Aversion reflex, 248

B Ballast for gas discharge lamps, 290 Ballast losses, 291 Banning of ballast, 291 Banning of lamp type, 264, 272, 275, 282 B40 band, 98, 101, 391 Binning, 33, 304 Biological clock, 188, see Suprachiasmatic nucleus (SCN) Blackbody locus, 28 Blackbody radiator, 35 Blue block goggles, 197 Blue block lens, 210 Blue light hazard, 247 Bluetooth protocol, 365 Body temperature, 141, 222 Breast cancer, 254 Bright light pulses, 195 Bus topology, 357

C CAD workstations, 81 Cancer, 188, 197, 253 Carcinogenic risk, 254 Cardiovascular disorder, 188 Cataract, 209 Centre-surround processing, 17 Ceramic gas discharge tube, 287 C-γ system of coordinates, 330 Charge-coupled device (CCD) detector, 434 Chip-on-board (COB), 302 Chip on glass (COG), 303 Chroma, 26 Chromatic adaptation, 21, 35, 93

© Springer Nature Switzerland AG 2019 W. van Bommel, Interior Lighting, https://doi.org/10.1007/978-3-030-17195-7

477

478 Chromaticity coordinates, 27, 449 Chronobiology, 139 Chronotherapy, 221 Chronotype, 98, 130, 143, 145, 224 CIECAM02-UCS colour space, 34, 45 CIE colour appearance model, 34 CIE colour matching functions, 30 CIE general colour rendering index, 38 CIELAB colour space, 34 CIELUV colour space, 34 CIE overcast sky, 321 CIE standard daylight illuminant, 319 CIE standard sky luminance distributions, 321 Circadian disruption, 222, 253 Circadian misalignment, 188 Circadian rhythm, 138, 142, 177, 188 Circadian sleep process, 170 Circadian stimulus (CS), 155, 174, 392 Classroom lighting, 417 COB LED, 302 Colorimeter, 439 Colorimetric observer, BNF–31 Colour appearance, 35–37 Colour constancy, 21, 26 Colour-designation system, 305 Colour discrimination, 50 Colour fidelity index, 38 Colour matching, 29 Colour-matching functions, 29, 30, 450–451 Colour metrics, 52 Colour rendering, 38–52, 395 Colour rendering index, 38 Colour spaces, 33–35 Colour temperature, 35 Colour triangle, 27 Colour vector graphics, 48–50 Communication protocols, 359 Compact fluorescent lamp (CFL), 264, 282 Complementary metal-oxide semiconductor (CMOS) detector, 434 Cones, 7 Continuous spectrum, 274, 276, 296 Contrast effective, 77 sensitivity, 64 threshold, 60, 63 Contrast rendering factor (CRF), 80 Control gear for discharge lamps, 289 Cornea, 5 Correlated colour temperature (CCT), 35–37, 83, 94, 127, 155, 161 Cortisol, 141, 157, 170, 222 Cosine correction, 435 Cylindrical illuminance, 393

Subject Index D Daisy chain topology, 357 Daylight colour, 319 factor, 323 flow, 326 levels, 318 spectrum, 319 Daylight harvesting, 354 Daylighting, 316 Daysimeter, 174 Department of Energy (DOE), xix Depressions, 223 Diffuseness, 113 Diffusers, 347 Digital addressable lighting interface (DALI) protocol, 361 Dim light melatonin onset (DLMO), 143 Dimming, 292, 311, 352 Directionality, 108 Direct lighting, 412 Direct photobiological effect, 151–152, 177, 191 Disability glare, 76, 114, 212 Discomfort glare, 114, 119, 121, 125, 126 DLMO, see Dim light melatonin onset (DLMO) DMX 512 protocol, 362 Dominant wavelength, 37 Drivers for LEDs, 309 Droop effect, 295, 311 Duty cycle, 235 Duv, 36 Dynamic lighting scenario, 181, 199, 229

E Eating disorder, 227 Edge detection, 16, 116 Efficiency droop, 295, 311 800 lamp colour series, 281 Electrodeless lamp, 285 Electroencephalography (EEG), 176, 178 Electromagnetic ballast, 290 Electronic ballast, 291 Emergency lighting, 423 Emitter, 280, 281 EN 12464, 396 Entrainment, 143, 181 Epileptic seizure, 242 Equivalent veiling luminance (Lveil), 77, 212 Equivalent visual efficiency (EVE), 84–85 Estrogen, 254 European Commission (EC), xix European Normalization Commission (CEN), xix

479

Subject Index EVE factor, 85 Excitation purity, 38 Extraction efficiency, 299

F Far-field photometry, 439 Farnsworth-Munsell 100 Hue (FM-100) test, 50 Filament LED lamp, 303 Fitting, 329 Fixture, 329 Flicker, 233 Flicker index, 236 Flickermeter, see IEC flickermeter Flicker sensitivity, 237 Flicker severity value Pst, 238 Flip-chip LED, 300 Flow of lighting, 109–112 Fluorescent lamp, 264, 277 Fluorescent powder, 279 Fourier transformation, 241 Fovea, 8 Fraunhofer lines, 316 Free-form optical surfaces, 345 Free-running rhythm, 141, 143 Functional lighting, 411–414

G Gamut area, 46–47, 94 Gamut index, 38, 46–47 Ganglion cells, 6, 13, 115 OFF-centre, 14 ON-centre, 14 Gas discharge lamps, 277 General colour fidelity index, 42–45 General colour rendering index, 38 Ghosting effect, 242 Glare disability (see Disability glare) discomfort (see Discomfort glare) indirect (see Indirect glare) overhead (see Overhead glare) Glare measurement, 443 Global Lighting Forum (GLF), xix Goggles, 192, 196 Goniophotometer, 439 Grating, 61, 63

H Halogen lamp, 264, 275 Hazardous effects, 188 Heating, ventilation and air conditioning (HVAC), 374

Heat sink, 299, 303, 347 Helmholtz-Kohlrausch effect, 93 HF electronic ballast, 291 High-bay lighting, 416 High-level vision, 20 Homeostatic sleep process, 170 Hospital lighting, 227, 420 Hue, 26, 31, 38, 39, 48 Human-centric lighting, 138, 182, 201 Hypophysis, see Pituitary gland

I IEC flickermeter, 238 IES TM-30-18, 42 Igniter, 290 Illuminance vector, 109 Illuminating Engineering Society of North America (IESNA), xix Imaging luminance-measuring devices (ILMDs), 442 Incandescent lamp, 266–267, 272, 275 Indirect glare, 79, 127, 401, 405 Indirect illuminance at the eye, 102–104 Indirect lighting, 414 Indoor navigation, 380 Indoor positioning system (IPS), 380 Induction lamp, 264, 285 Inductive ballast, 290 Industrial lighting, 415–417 Insomnia, 189 Integrative lighting, see Human-centric lighting Intensity tables, 331 Intensive care unit (ICU) delirium, 229 Intensive care unit (ICU) lighting, 422 International Electrotechnical Commission (IEC), xix International Lighting Commission (CIE), xix Internet of Things (IoT), 371 Intraocular lens, 210 Intrinsically photosensitive retinal ganglion cell (ipRGC), 148 Ionization, 278 IoT lighting system, 373 ipRGC, see Photosensitive retinal ganglion cell (pRGC) Iris, 5 International Standards Organization (ISO), xx I-table, 331, 332

J Jetlag, 146, 222 social, 145 Junction temperature, 304

480 K Kruithof’s law, 127

L Lamp failure catastrophic, 269 parametric, 269 Lamp flicker, 233 Lamp lifetime, see Lifetime Lamp lumen depreciation, see Lumen depreciation Lamp risk groups, 249 Light at night (LAN), 253 Landolt ring, 66 L-cones, 10 LED luminaires matrix, 16 Lenses, 344 Letter size, 60 LGN-ganglion cells, 17 Lifetime, 267, 348 Light as sensor, 381–382 Light beyond illumination, 371–383 Light colour preference, 127–130 Light distribution, 330 Light dose, 173 Light emitting area, 121, 123 Light-emitting diode (LED), 264, 292, 295 engine, 302 module, 301 package, 299 Light extraction efficiency, 299 Light Fidelity (Li-Fi), 366, 382 Light field, 109 Lighting control, 352 Lighting design process, 408 Lighting field, 113 Light logging device, 444 Light output ratio, 333 Light pulses, 195 Light therapy, 222 Light tubes, 112 LM-80 document, 270 Low-bay lighting, 415 Lumen depreciation, 267 Lumen maintenance testing, 270 Lumen method, 336, 432 Luminaire lifetime, 348 Luminaires, 329 Luminance contrast, 58 diagram, 337 map, 123

Subject Index mapping, 442, 443 meter, 434 Luminous efficacy, 266 theoretical maximum, 266 Luminous intensity distribution, 330

M MacAdam ellipses, 33, 304 Macular degeneration, 210 M-cones, 10 Mean indirect cubic illuminance (MICI), 107 Mean room surface exitance (MRSE), 102–107, 391–392 Measurement of light and lighting, 433 Melanopic daylight efficacy ratio, see Melanopic equivalent daylight (D65) factor Melanopic equivalent daylight (D65) factor, 161, 181, 392 Melanopic equivalent daylight (D65) illuminance, 162, 181, 201, 392 Melanopic irradiance, 161, 392 Melanopsin, 148, 158 Melatonin, 141, 156, 170, 222 Melatonin suppression, 152–154, 162, 191 Mesh topology, 358 Metabolic disorder, 188 Metal-halide (MH) lamp, 264, 287 Metamerism, 51–52 Micro lenses, 345 Migraine, 242 Mirror reflector, 342 Modelling, 108–114 Modulation depth, 236 Multilayer LED chip, 296 Munsell book of colour, 26

N Near-field photometry, 441 Network topologies, 357 Neural pathway, 149 Neurophysiological adverse effects, 242 Night shift, 188 Nighttime lighting strategy, 192 900 lamp colour series, 281 Non-image-forming (NIF) effects, 138 Nursing home lighting, 422

O Obesity, 188 Occupancy sensing, 354

Subject Index 1–10 V DC protocol, 359 ON-OFF signal processing, 17 Open Architectures for Intelligent Solid-state lighting systems (OpenAIS), 374 Opposing colour mechanism, 18 Opsin, 7, 157 Optical radiation hazards, 246 Orange glasses, 197 Organic light-emitting diode (OLED), 264, 295, 306 Overhead glare, 126

P Paraventricular nucleus (PVN), 150 Parkinson’s disease, 225 Perceived adequacy of the illumination (PAI), 104 Percent flicker, 236 Perceptual constancy, 20 Peripheral clocks, 143 Peripheral vision, 9 Personal control system, 356 Phantom array effect, 242 Phase-cutting, 274, 277, 292 Phase response curve, 145 Phase shift, 145, 147, 188, 192, 195 Phosphor coating, 301 Phosphors, 279, 297, 301 Phosphor white LED, 296 Photocell, 434, 442 Photochemical damage, 246 Photodiode cells, 433 Photoluminescence, 279 Photometer, 436 accuracy, 436 integrating, 437 Photometric data, 331 Photometry, 434 Photophone, 376 Photopic vision, 11 Photopigment, 8, 157 Photopsin, 9 Photo-retinitis, 247 Photosensitive retinal ganglion cell (pRGC), 19, 82, 148, 149, 154, 158 Photosensitivity, 242 Photosphere, 316 Phototransduction, 8 Phototropism, 79 Photovoltaic cells, 433 Pineal gland, 149–151 Pituitary gland, 151 Planckian locus, 28 p-n junction, 294

481 p-n sandwich, 294 Polar luminous intensity diagram, 332 Polysomnography (PSG), 171 Position index p, 117, 119, 459 Power nap, 183 Power over Ethernet (PoE), 367–368 pRGC, see Photosensitive retinal ganglion cell (pRGC) Primary colours, 28 Prisms, 344 Psychophysics, 90 Pulse-width modulation (PWM) dimming, 234, 311 Pupil, 5 Pupil diameter, 83 Pupillary reflex, 19, 82, 151, 248

Q Quantum dots, 297

R Ra, 39–42 Rare-earth metals, 279 Rayleigh scattering, 316 Receptive field, 13, 115 Recombination, 294 Reference light sources, 40–41, 44 Relative visual performance (RVP) model, 75, 78, 453, 455 Remote device management (RDM) protocol, 363 Remote phosphor, 301 Retina, 6 Retinal neural wiring, 148, 154 Retrofit LED lamp, 302 Rf, 42–45 RF communication, 375, 379 Rg, 46–47 RGB LED, 296 Rhodopsin, 9 Risk groups, see Lamp risk groups Rods, 7 Room appearance, 95–108 Room index, 335 Rotating shift, 187, 199, 200

S Saccadic eye movements, 248 SAD, see Seasonal affective disorder (SAD) Saturation, 26, 31, 38, 46, 48 SCN, see Suprachiasmatic nucleus (SCN) S-cones, 10

482 Scotopic-photopic (S/P) ratio, 85, 94 Scotopic vision, 11 Seasonal affective disorder (SAD), 223 Shielding angle, 341, 401 Shift work, 187, 199 Shift work sleep disorder (SWSD), 189 Short-wavelength depleted light, 197 Short-wavelength filtered light, 197 Silhouette effect, 326 Silicon encapsulation, 300 Skylight, 316 Sky luminance, 321 Sleep, 189 Sleep disorders, 189, 224 Sleepiness, 176 Sleep mechanism, 170–171 Sleep phase syndrome, 224 Sleep pressure, 170 Sleep quality, 173 Sleep window, 171 Smart lighting, 352 Smart networks, 357 Solid-state light source (SSL), 293 Spatial brightness, 90–95 Spatial frequency, 61 Spectral sensitivity, 11 Spectrophotometer, 438 Specular reflector, 342 S/P ratio, see Scotopic-photopic (S/P) ratio Standard colorimetric observer, 28, 30 Star topology, 358 Stroboscopic effect, 238 Stroboscopic sensitivity, 240 Stroboscopic visibility measure (SVM), 240 Subjective brightness, 91 Sunlight, 316 Suprachiasmatic nucleus (SCN), 142, 149, 152

T Ta, 347 Target/ambient illuminance ratio (TAIR), 106 Temporal light artefacts (TLA), 233 Thyristor dimmer, 274 Time modulated light artifacts, see Temporal light artefacts (TLA) Time of flight (ToF), 381 TIR optics, 345 TLA, see Temporal light artefacts (TLA) TM-21 document, 271 TM-28-14 document, 348 Total internal reflection (TIR), 299, 345–347 Transmission Control Protocol/Internet Protocol (TCP/IP), 366

Subject Index Tree topology, 358 Tristimulus values, 31

U UCS diagram, 32 UGR, see Unified glare rating (UGR) Ulbricht’s sphere, 437 Unified glare rating for a lighting installation (UGRL), 119 Unified glare rating (UGR), 115–125, 393 diagram, 340 table, 337 Uniform chromaticity scale diagram (UCS), see UCS diagram Utilisation factor, 334 UV-blocking quartz, 277, 287 u–v chromaticity diagram, 32–33

V V(λ) correction, 435 curve, 11, 447 VCP glare system, 404 Vector to scalar ratio, 109, 394 Veiling luminance, 77 Vigilance test, 177 Visibility supra-threshold, 65–75 threshold, 60–65 Visible flicker, 237 Visible light communication (VLC), 375–383 Visible light positioning (VLP), 380 Visual acuity, 60, 61, 83 Visual angle, 59 Visual attractiveness, 97 Visual comfort probability (VCP), 404 Visual cortex, 5, 8 Visual interest, 97, 102 Visual lightness, 97, 102 Visual performance model, 74–75 relative, 70, 76, 78 Visual satisfaction, 89 Vitality, 176

W Wafer, 298 Wardroom lighting, 420 Weber-Fechner law, 90 Wi-Fi protocol, 365 Winter depression, 223

Subject Index Wireless protocols, 363 WUV colour space, 34, 42

X x–y chromaticity diagram, 26–32 x–y coordinates, 27 XYZ colour space, 31

483 Z Zeitgeber, 144 0–10 V DC protocol, 359 Zhaga, xx, 302 ZigBee protocol, 364 Z-Wave protocol, 365