Wiley – SID Series in Display Technology
Series Editor: Dr. Ian Sage
Advisory Board: Paul Drzaic, Ioannis (John) Kymissis, Ray Ma, Ian Underwood, Michael Wittek, Qun (Frank) Yan
Fundamentals and Applications of Colour Engineering
Phil Green
E-Paper Displays
Bo-Ru Yang
Liquid Crystal Displays - Addressing Schemes and ElectroOptical Effects, Third Edition
Ernst Lueder, Peter Knoll, and Seung Hee Lee
Flexible Flat Panel Displays, Second Edition
Darran R. Cairns, Dirk J. Broer, and Gregory P. Crawford
Amorphous Oxide Semiconductors: IGZO and Related Materials for Display and Memory
Hideo Hosono, Hideya Kumomi
Introduction to Flat Panel Displays, Second Edition
Jiun-Haw Lee, I-Chun Cheng, Hong Hua, and Shin-Tson Wu
Flat Panel Display Manufacturing
Jun Souk, Shinji Morozumi, Fang-Chen Luo, and Ion Bita
Physics and Technology of Crystalline Oxide Semiconductor
CAAC-IGZO: Application to Displays
Shunpei Yamazaki, Tetsuo Tsutsui
OLED Displays: Fundamentals and Applications, Second Edition
Takatoshi Tsujimura
Physics and Technology of Crystalline Oxide Semiconductor
CAAC-IGZO: Fundamentals
Noboru Kimizuka, Shunpei Yamazaki
Physics and Technology of Crystalline Oxide Semiconductor
CAAC-IGZO: Application to LSI
Shunpei Yamazaki, Masahiro Fujita
Interactive Displays: Natural Human-Interface Techniques
Achintya K. Bhowmik
Addressing Techniques of Liquid Crystal Displays
Temkar N. Ruckmongathan
Modeling and Optimization of LCD Optical Performance
Dmitry A. Yakovlev, Vladimir G. Chigrinov, and Hoi-Sing Kwok
Fundamentals of Liquid Crystal Devices, Second Edition
Deng-Ke Yang and Shin-Tson Wu
3D Displays
Ernst Lueder
Illumination, Color and Imaging: Evaluation and Optimization of Visual Displays
P. Bodrogi, T. Q. Khan
Liquid Crystal Displays: Fundamental Physics and Technology
Robert H. Chen
Transflective Liquid Crystal Displays
Zhibing Ge and Shin-Tson Wu
LCD Backlights
Shunsuke Kobayashi, Shigeo Mikoshiba, and Sungkyoo Lim (Eds.)
Mobile Displays: Technology and Applications
Achintya K. Bhowmik, Zili Li, and Philip Bos (Eds.)
Photoalignment of Liquid Crystalline Materials: Physics and Applications
Vladimir G. Chigrinov, Vladimir M. Kozenkov, and HoiSing Kwok
Projection Displays, Second Edition
Mathew S. Brennesholtz and Edward H. Stupp
Introduction to Microdisplays
David Armitage, Ian Underwood, and Shin-Tson Wu
Polarization Engineering for LCD Projection
Michael G. Robinson, Jianmin Chen, and Gary D. Sharp
Digital Image Display: Algorithms and Implementation
Gheorghe Berbecel
Color Engineering: Achieving Device Independent Color
Phil Green and Lindsay MacDonald (Eds.)
Display Interfaces: Fundamentals and Standards
Robert L. Myers
Reflective Liquid Crystal Displays
Shin-Tson Wu and Deng-Ke Yang
Display Systems: Design and Applications
Lindsay W. MacDonald and Anthony C. Lowe (Eds.)
Fundamentals and Applications of Colour Engineering
Edited By Phil Green
This edition first published 2024 © 2024 John Wiley & Sons Ltd
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by law. Advice on how to obtain permission to reuse material from this title is available at http://www.wiley.com/go/permissions.
The right of Phil Green to be identified as the author of the editorial material in this work has been asserted in accordance with law.
Registered Offices
John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA
John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK
For details of our global editorial offices, customer services, and more information about Wiley products visit us at www.wiley.com.
Wiley also publishes its books in a variety of electronic formats and by print-on-demand. Some content that appears in standard print versions of this book may not be available in other formats.
Trademarks: Wiley and the Wiley logo are trademarks or registered trademarks of John Wiley & Sons, Inc. and/or its affiliates in the United States and other countries and may not be used without written permission. All other trademarks are the property of their respective owners. John Wiley & Sons, Inc. is not associated with any product or vendor mentioned in this book.
Limit of Liability/Disclaimer of Warranty
While the publisher and authors have used their best efforts in preparing this work, they make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives, written sales materials or promotional statements for this work. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice and strategies contained herein may not be suitable for your situation. You should consult with a specialist where appropriate. The fact that an organization, website, or product is referred to in this work as a citation and/or potential source of further information does not mean that the publisher and authors endorse the information or services the organization, website, or product may provide or recommendations it may make. Further, readers should be aware that websites listed in this work may have changed or disappeared between when this work was written and when it is read. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.
A catalogue record for this book is available from the Library of Congress
Hardback ISBN: 9781119827184; ePub ISBN: 9781119827207; ePDF ISBN: 9781119827191; oBook ISBN: 9781119827214
Cover image: © iSam iSmile/Shutterstock
Cover design by Wiley
Set in 10/12pt WarnockPro by Integra Software Services Pvt. Ltd, Pondicherry, India
Contents
Series Editor’s Foreword xvii
Preface xix
Introductory Notes xxi
1 Instruments and Methods for the Colour Measurements Required in Colour Engineering 1 Danny Rich
1.1 Introduction 1
1.1.1 The Need for Colorimetry 1
1.1.2 The Principles of Colorimetry 1
1.1.3 Making the Transition from What We “See” to Quantifying How We “Match” a Colour 2
1.2 Visual Colorimetry 3
1.2.1 A Method to Uniquely Map the Colour of Lights and Objects 3
1.2.2 Development of the CIE Method of Visual Colorimetry 4
1.2.3 Applications of Visual Colorimetry 6
1.2.4 Disadvantages of Visual Colorimetry 7
1.3 Analogue Simulation of Visual Colorimetry 7
1.3.1 Replacing the Human Eye with an Optoelectronic Sensor 7
1.3.2 Substituting Coloured Filters to Approximate the CIE Colour-Matching Functions 8
1.3.3 Assessing the “Goodness of Fit” of a Set of Colorimeter Filters 10
1.3.4 Schematic Description of Analogue Filter Colorimeters 11
1.3.5 Disadvantages of Analogue Filter Colorimeters 11
1.4 Digital Simulation of Visual Colorimetry 12
1.4.1 Replacing the Analogue Filters with an Abridged Spectrometer 12
1.4.2 Assessing the “Goodness of Fit” of Abridged Spectrometers 13
1.4.3 Schematic Description of Digital Spectrocolorimeters 13
1.4.4 Advantages and Disadvantages of Digital Spectrocolorimeters 14
1.5 Selecting and Using Colorimeters and Spectrocolorimeters 15
1.5.1 Reading and Understanding Specifications and Technical Literature 15
1.5.2 Verifying Performance Specifications 17
1.5.3 Standards of Colour and Colour-difference 17
1.5.4 Sources of Error and Uncertainty in the Measurement of Reflectance, Transmittance and Radiance 18
1.6 Geometric Requirements for Colour Measurements 18
1.6.1 Colour Measurements from Self-Luminous Objects 18
1.6.2 Colour Measurements from Reflecting or Transmitting Objects 19
1.7 Conclusions and Expectations 22
1.7.1
Current CIE and ISO Activities in Colour and Colour-difference Measurements 22
1.7.2 Quality Management Systems and Colour Measurements 22 References 23
2 Colorimetry and Colour Difference 27 Phil Green
2.1 Introduction 27
2.2 Colorimetry 27
2.3 Normalization 28
2.4 Colour Matching Functions 29
2.5 Illuminants 29
2.6 Data for Observers and Illuminants 30
2.7 Range and Interval 30
2.8 Calculation of Chromaticity 31
2.9 Calculation of CIE 1976 Uniform Colour Spaces 31
2.10 Inversion of CIELAB Equations 34
2.11 Colour Difference 34
2.12 Problems with Using UCS Colour Difference 35
2.13 Uniformity of the Components of Colour Difference 35
2.13.1 Chroma 35
2.13.2 Hue 36
2.13.3 Lightness 36
2.14 Viewing Conditions 36
2.15 Surface Characteristics 37
2.16 Acceptability of Colour Differences 37
2.17 Overcoming the Limitations of UCS Colour Difference with Advanced Colour Difference Metrics 37
2.18 CIE94 37
2.19 CIEDE2000 39
2.20 Progress on Colour Difference Metrics since CIEDE2000 41
2.21 3D Colour Difference 41
2.22 Colour Difference in High Luminance Conditions 41
2.23 Colour Difference Formulas Based on Colour Appearance Models 41
2.24 Limitations in the Use of Advanced Colour Difference Metrics in Colour Imaging 42
2.25 Basis Conditions 42
2.25.1 Illuminant 42
2.25.2 Illuminance 42
2.25.3 Sample Separation 42
2.25.4 Sample Size and Image Structure 43
2.26 Colour Difference in Complex Images 43
2.27 Acceptability and Perceptibility 44
2.28 Large vs Small Differences 44
2.29 Deriving Colour Difference Tolerances 44
2.30 Sample Preparation 45
2.31 Psychophysical Experiments 45
2.31.1 Observer Variability and Experience 45
2.32 Colour Difference Judgements by Observers with a Colour Vision Deficiency 46
2.33 Calculating Colour Tolerances from Experimental Data 46
2.34 Calculation of Discrimination Ellipsoids and Tolerance Distributions 46
2.34.1 Calculation of Parametric Constants in Weightings Functions 47
2.35 Calculation of Acceptability Thresholds 48
2.36 Evaluating Colour Difference Metrics 48
2.37 Conclusion 48 References 49
3 Fundamentals of Device Characterization 53
Phil Green
3.1 Introduction 53
3.1.1 Objectives 54
3.2 Characterization Methods 54
3.2.1 Test Charts 55
3.2.2 Calibration 55
3.2.2.1 Matching Aim Values 56
3.2.2.2 Optimizing Performance 56
3.2.2.3 Perceptual Uniformity of Device Values 56
3.2.2.4 Optimization for Machine Vision 56
3.2.3 Linearization 56
3.3 Numerical Models 57
3.3.1 Regression Methods Used in Characterization 58
3.3.1.1 First Order Model 58
3.3.1.2 Higher Order Models 59
3.3.1.3 Choosing the Polynomial Order 60
3.3.1.4 Spline Methods 60
3.3.1.5 Weighted Regression 60
3.3.2 Domain 61
3.3.3 Optimization 62
3.3.4 Noisy and Discontinuous Data 62
3.3.5 Machine Learning 62
3.4 Look-Up Tables with Interpolation 63
3.4.1 Packing 63
3.4.2 Extraction 64
3.4.3 Interpolation 64
3.4.4 LUT Implementation 66
3.4.4.1 LUT implementation in ICC profiles 67
3.5 Evaluating Accuracy – Training and Test Data 67 References 68
4 Characterization of Input Devices 71
Phil Green
4.1 Input Channels 71
4.2 Characterization Goals 72
4.3 Transform Encoding 73
4.4 Dynamic Range 73
4.5 Input Characterization Methods 74
4.5.1 Scanners 74
4.6 Targets 74
4.7 Modelling 74
4.7.1 Digital Cameras 75
4.8 Target-Based Characterization 75
4.9 Targets 75
4.10 Modelling 76
4.10.1 Spectral Sensitivity-based Methods 77
4.10.2 Machine Learning Methods 78
4.10.3 Spectral Characterization of Input Devices 78 References 78
5 Color Processing for Digital Cameras 81
Michael S. Brown
5.1 Introduction 81
5.2 Basics of a Camera Sensor 82
5.3 The Camera Pipeline 83
5.3.1 Defective Pixel Correction 83
5.3.2 Black-Level Correction and Normalization 84
5.3.3 Lens Shading Correction 84
5.3.4 Autofocus, Autoexposure, Auto White Balance 85
5.3.4.1 Autoexposure 85
5.3.4.2 Autofocus 86
5.3.5 White Balance and Auto White Balance 87
5.3.5.1 White Balance 87
5.3.5.2 Manual and Auto White Balance 87
5.3.6 Demosaicing 88
5.3.7 Noise Reduction 89
5.3.8 Color Space Transform to Device-Independent Color Space 89
5.3.9 Photo-Finishing/Rendering 90
5.3.9.1 General and Selective Color Manipulation 90
5.3.9.2 Global and Local Tone-Mapping 90
5.3.9.3 Sharpening/Noise and Grain 91
5.3.9.4 Image Resizing/Super-Resolution 91
5.3.10 Color Mapping to Final Image Encoding Color Space 92
5.3.11 Compression and Save to Storage 92
5.3.12 RAW Image Capture 92
5.4 Multi-Frame Processing 93
5.4.1 HDR Imaging 93
5.4.2 Low-Light/Night-Mode Imaging 94
5.5 Towards the Neural ISP 94
5.6 Concluding Remarks 95 Acknowledgment 95 References 95
6 Display Calibration 99 Catherine Meininger, Tom Lianza, and Grace Annese
6.1 Introduction 99
6.2 From CRT to Contemporary Display Technologies 99
6.3 The Display Never Sleeps… Merging Television and Computer Display Standards 102
6.4 The Evolution of Display Calibration Capabilities 103
6.4.1 Gamut Mapping 104
6.4.2 Manual Calibration 105
6.4.3 One Dimensional Lookup 108
6.3.4 The Matrix Shaper Architecture 108
6.4.5 Single 3-Dimensional LUT 109
6.4.5.1 3DLUT Considerations 111
6.4.6 Hybrid Matrix Shaper Utilizing 3DLUT Followed by a 1DLUT 111
6.5 Measurement Set Requirements 111
6.5.1 Pattern Generation 112
6.5.2 How Many Measurements are Needed? 112
6.5.3 Methods to Mitigate Drift in Display Measurements 112
6.6 Calibration Validation Methodologies 113
6.6.1 Numerical Scales 113
6.6.2 Visual Evaluation Targets and Methods 114
6.7 Low Blue Light Developments 114
6.8 Conclusions 117 References 117
7 Characterizing Hard Copy Printers 119 Phil Green
7.1 Introduction 119
7.2 Properties of Hard Copy Printers 120
7.3 Substrates and Inks 120
7.3.1 Fluorescent Whitening Agents 120
7.3.2 Inks 120
7.4 Colour Gamut 120
7.5 Halftoning 121
7.6 Mechanical Printing Systems 122
7.7 Printing Conditions 122
7.8 Digital Systems 122
7.9 RGB Printers 122
7.10 Test Charts 123
7.11 Printer Models 124
7.12 Block Dye Model 125
7.13 Physical Models 126
7.13.1 Density 126
7.13.2 Dot Area Models 126
7.13.2.1 Murray-Davies 127
7.13.2.2 Yule-Nielsen 127
7.13.2.3 Clapper-Yule 128
7.13.2.4 Additivity Failure 128
7.13.3 Neugebauer 128
7.13.3.1 Modified and Extended Neugebauer Equations 130
7.13.3.2 N-Modified Neugebauer Equations 130
7.13.4 Vector-Corrected Neugebauer Equations 130
7.13.4.1 Cellular Extensions 130
7.13.4.2 Spectral Extensions 131
7.13.4.3 Evaluation of Different Forms of the Neugebauer Equations 131
7.13.5 Colorant Models 131
7.13.5.1 Masking Equations 131
7.13.6 Beer-Bouguer 132
7.13.7 Kubelka-Munk 132
7.13.8 Extensions 134
7.14 Numerical Models and Look-up Tables 134
7.14.1 Black Printer 134
7.14.1.1 Spectral Grey-Component Replacement 136
7.14.1.2 Black Generation Algorithm 136
7.15 Inverting the Model 137
7.16 Multi-Colour and Spot Colour Characterization 137
7.17 Spectral Characterization 137
7.18 White Ink 138
7.19 Reducing the Frequency of Characterization 138
7.20 Conclusions 138 References 138
8 Colour Encodings 143 Phil Green
8.1 Introduction 143
8.2 Colour Encoding Components 143
8.3 Colour Spaces 144
8.4 Device and Colour Space Encodings 144
8.5 Colorimetric Interpretation 144
8.6 Image State 145
8.7 Standard 3-Component Colour Space Encodings 146
8.8 Colour Gamut 146
8.8.1 Extended Colour Gamut 147
8.9 Precision and Range 147
8.9.1 High Dynamic Range 148
8.9.2 Negative Values 149
8.10 Luminance/Chrominance Encodings 149
8.11 Conversion to Colorimetry 150
8.12 Implementation Issues 150
8.13 File Formats 152 References 152
9 Colour Gamut Communication 155 Kiran Deshpande
9.1 Introduction 155
9.1.1 Device Colour Gamut and the Usable Colour Gamut 155
9.1.2 Colour Space 156
9.1.3 Factors Affecting Colour Gamut 157
9.1.4 Gamut of an Image 157
9.2 How to Describe Colour Gamuts 157
9.2.1 Convex Hull 158
9.2.2 Alpha-Shapes 158
9.2.3 The Segment Maxima Method 159
9.2.4 Gamut Based on a Printer Model 159
9.2.5 Gamulyt Method 160
9.2.6 The Mountain Range Method 161
9.2.7 Defining Gamut Boundaries with a Test Target 161
9.3 How to Obtain a Colour Gamut of a Printing System 162
9.4 How to Obtain a Colour Gamut of a Display 163
9.5 How to Calculate Gamut Volume 163
9.6 How to Analyse Colour Gamuts 164
9.6.1 Metrics for Comparing Colour Gamuts 165
9.6.2 Gamut Analysis of an N-Colour Printing Process 166
9.7 How to Visualize Colour Gamuts 167
9.7.1 Venn Diagram 168
9.7.2 Gamut Rings 169
9.8 How to Communicate Colour Gamuts 171
9.8.1 How to Encode a Colour Gamut Description 172
9.8.1.1 Encoding Based on CxF 173
9.8.1.2 Encoding Based on an ICC Profile 173
9.8.1.3 Encoding Based on Tab-Delimited Text Files 173
9.9 Summary 173
References 174
10 The ICC Colour Management Architecture 177
Phil Green
10.1 Origins of the ICC 177
10.2 Fundamentals of the ICC Architecture: The PCS, the ICC Profile, Transforms and the CMM 178
10.2.1 Range and Precision 179
10.2.2 Tags and Types 180
10.2.3 Media-Relative Colorimetry 180
10.2.4 Image State 181
10.2.5 Rendering Intents 181
10.2.6 Profile Classes 183
10.2.7 Features of ICC v4 183
10.2.8 Making Profiles 184
10.2.9 Embedding Profiles 184
10.2.10 CMMs 184
10.3 Other CMM Operations 185
10.3.1 Function Inversion 185
10.3.2 Black Point Compensation 185
10.3.3 Channel Preservation 186
10.3.4 Gamut Mapping 186
10.3.5 Copyright and Security 187
10.4 Workflows 187
10.5 Current Status of ICC.1 188
10.5.1 Limitations of ICC.1 188
10.6 ICC.2 189
10.6.1 PCS 189
10.6.2 Data Types 189
10.6.3 Multiprocess Elements 189
10.6.4 Calculator 189
10.6.5 Workflows 190
References 190
11 iccMAX Color Management – Philosophy, Overview, and Basics 193
Max Derhak
11.1 Background and Philosophy Leading to iccMAX 193
11.2 Overview 194
11.2.1 Making Connections 194
11.2.2 Transform Connection and Application 199
11.2.3 Encoding Transforms 200
11.2.4 MultiProcessing Element Transforms 202
11.2.4.1 Matrix Elements 202
11.2.4.2 Curve Elements 203
11.2.4.3 Tint Array Elements 203
11.2.4.4 CLUT Elements 203
11.2.4.5 Tone Mapping Element 203
11.2.4.6 Calculator Element 204
11.2.4.7 Emissive Elements 204
11.2.4.8 Emission Observer Element 205
11.2.4.9 Reflectance CLUT Element 205
11.2.4.10 Reflectance Observer Element 205
11.2.4.11 CAM Elements 205
11.2.5 Profile Structure Generalization 206
11.3 Creating Transforms 207
11.4 Specification Subsets via ICSs 209
11.5 Domain Specific Examples 210
11.5.1 Photography 210
11.5.2 Packaging 210
11.5.3 Medical Imaging 211
11.5.4 Fine Art 211
11.5.5 Critical Color on Wide Gamut Displays 211
11.6 Getting Started with iccMAX (Where Color Engineering Comes to Play) 212
11.7 Conclusion 213 References 213
12 Sensor Adjustment 215 Phil Green
12.1 Introduction 215
12.2 Aims of Sensor Adjustment 215
12.3 Luminance Adjustment 216
12.4 Chromatic Adaptation 218
12.4.1 Chromatic Adaptation in Colour Management 219
12.4.2 Chromatic Adaptation in ICC.2 220
12.5 Material-Equivalent Adjustment 220
12.6 Local Adaptation 221
12.7 Incomplete Adaptation 222 References 223
13 Evaluating Colour Transforms 227 Phil Green
13.1 Introduction 227
13.2 Accuracy 227
13.2.1 Metamerism 228
13.2.2 Smoothness 229
13.2.3 Spatial Artefacts 229
13.2.4 Spectral Accuracy 229
13.2.5 Acceptability 230
13.2.6 Sources of Error in Colour Transforms 230
13.2.7 Procedures for Colorimetric Transform Evaluation 231
13.2.8 Media-Relative Colour Transforms 232
13.3 Cost 232
13.4 Subjective Preference 233
13.4.1 Test Data 234
13.4.2 Reporting Evaluation Results 234
13.4.2.1 Visualisation of Results 235
References 236
14 Appearance Beyond Colour: Gloss and Translucency Perception 239
Davit Gigilashvili and Jean-Baptiste Thomas
14.1 Introduction 239
14.2 Gloss Perception 240
14.2.1 Perceptual Dimensions of Gloss 241
14.2.2 Image Cues and Partial Models 241
14.2.3 Factors Impacting Perceived Gloss 242
14.2.3.1 Shape 242
14.2.3.2 Illumination 243
14.2.3.3 Motion 243
14.2.3.4 Observer 243
14.2.4 Summary and Open Questions 244
14.3 Translucency Perception 244
14.3.1 Transparency and Translucency 245
14.3.2 Image Cues and Partial Models 246
14.3.3 Factors Impacting Perceived Translucency 247
14.3.4 Summary and Open Questions 248
14.4 Interaction among Appearance Attributes 248
14.4.1 Impact of Colour on Gloss and Translucency 248
14.4.2 Interaction between Gloss and Translucency 249
14.5 Impact on Colour Technologies 250
14.6 Conclusion 252 References 253
15 Colour Management of Material Appearance 259
Tanzima Habib
15.1 Introduction 259
15.2 Material Appearance Modelling 260
15.2.1 Blinn‒Phong Model 262
15.2.2 Ward Model 262
15.2.3 Cook‒Torrance Model 262
15.3 Appearance Support in Colour Management 263
15.4 A Colour Management Workflow for Material Appearance 264
15.5 Conclusion 269 References 270
16 Color on the Web 271
Chris Lilley
16.1 Early History 271
16.2 Color on the Legacy Web 272
16.2.1 RGB Representations 272
16.2.2 Color Names 273
16.2.3 Color with Alpha 274
16.2.4 Hue-Wheel Systems 274
16.2.5 Gradients 276
16.3 Wide Color Gamut (WCG) Comes to the Web 277
16.3.1 The Importance of Display P3 277
16.3.2 WCG Raster Images, with ICC Profiles 279
16.3.3 Development of WCG Upgrades to Web Specifications 279
16.3.4 Limitations of CIELAB: Introducing OK Lab 280
16.4 Color on the Wide Gamut Web 281
16.4.1 Predefined RGB Color Spaces 281
16.4.2 Device-independent Color Spaces 283
16.4.3 WCG Gradients 284
16.4.4 Manipulating and Mixing Colors 285
16.5 HDR Comes to the Web 286
16.5.1 Introducing HDR 286
16.5.2 HDR in Canvas 287
16.5.3 HDR in WebGL and WebGPU 287
16.5.4 HDR in CSS 287
References 287
17 High Dynamic Range Imaging 293 Mekides Assefa Abebe
17.1 Introduction and Background 293
17.1.1 The Human Visual System 293
17.1.2 Color Imaging 294
17.2 High Dynamic Range Imaging 296
17.2.1 HDR Acquisition 296
17.2.1.1 Single Exposure HDR Acquisition 296
17.2.1.2 Multi-Exposure HDR Acquisition
17.2.1.3 HDR Image Synthesis 299
17.2.2 HDR Image Storage 299
17.2.2.1 HDR Image Formats and Encoding 300
17.2.3 HDR Rendering 302
17.2.3.1 Tone Mapping 303
17.2.3.2 Reverse Tone Mapping 303
17.3 Conclusion 308
References 308
18 HDR and Wide Color Gamut Display Technologies and Considerations 311 Timo Kunkel and Ajit Ninan
18.1 Introduction 311
18.2 Early HDR Display Systems 312
18.3 Transmissive Displays 313
18.3.1 Liquid Crystal Display Technology 313
18.3.2 Global Modulation 314
18.3.3 Dual Modulation 314
18.3.4 Dual LCD Displays 316
18.4 Emissive Displays 317
18.4.1 Organic Light Emitting Diodes 317
18.4.2 Direct LED Displays 318
18.5 Projection Systems 319
18.5.1 Projection-LCD Dual Modulation 319
18.5.2 Screen Projection 320
18.6 Reflective Displays 320
18.7 Achieving Wide Color Gamuts 321
18.7.1 Designing Narrow Primaries 323
18.7.2 Multi Spectral or Multi-Primary Displays 325
18.7.3 Metameric Error 325
18.8 Spatial Display Properties 326
18.9 Temporal Display Properties 327
18.10 Signaling 328
18.10.1 Signal, Display and Content Properties 328
18.10.2 Signal Reference vs. Display Preference Modes 329
18.10.3 Professional vs. Consumer Displays 329
18.11 Characterization and Calibration 330
18.12 Ambient Effects 330
18.13 Conclusion 332 References 332
19 Colour in AR and VR 335
Michael J. Murdoch
19.1 Introduction 335
19.2 Colour Synthesis in AR and VR Displays 337
19.2.1 GOG Display Model 337
19.2.2 Idealized Display Models 338
19.2.3 Spatial and Temporal Independence 338
19.2.4 HMD Optics 339
19.2.5 Measuring AR and VR Displays 339
19.2.6 Example: Measuring and Characterizing an AR Display 340
19.3 Colour Appearance in AR and VR 342
19.3.1 Limitations of CAMs for AR and VR 342
19.3.2 Chromatic Adaptation 343
19.3.2.1 Chromatic Adaptation in VR 343
19.3.2.2 Chromatic Adaptation in AR 344
19.3.3 Scission and Transparency in AR 344
19.3.3.1 Experimental Evidence for Scission in OST-AR 344
19.3.3.2 Interpretation of Transparency and Related Visual Effects 346
19.3.4 Example: Modelling an OST-AR Display and Colour Matching Results 347
19.4 Colour Imaging and Graphics in AR and VR 350
19.4.1 Colour Reproduction 350
19.4.2 Virtual Colour Reproduction 350
19.5 Conclusion 351
19.5.1 Open Questions in AR and VR 351 Acknowledgements 352 References 352
20 Colour Engineering Toolbox and Other Open Source Tools 355
Phil Green
20.1 Colour Engineering Toolbox 2.0 355
20.1.1 Colorimetry 357
20.2 Polar Calculations 357
20.3 Media-Relative and PCS Scaling 357
20.3.1 Adaptation 357
20.3.2 Difference 358
20.3.3 Characterization 358
20.3.4 Gamut 359
20.3.5 Utility Functions 359
20.3.6 Psychophysics 359
20.3.7 Documentation 360
20.3.8 Licensing and Use 360
20.4 DemoIccMax 360
20.5 Color.js 360
20.6 Little CMS 360
20.7 Argyll 361
20.8 Colour 361 References 361 Index 363
Series Editor’s Foreword
The central questions of colour engineering: “Are these two objects the same colour?” and “If not, are they close enough to be acceptable?” have an apparent beguiling simplicity based on the familiarity with colour that most of us share in everyday life. Display scientists and engineers know different. Comparison of reference objects with those made using different colourants, or with their representations rendered through different imaging devices, data pipelines, display technologies and hard copy devices, under different conditions of illumination and view becomes challenging, and compensating for the different behaviour of each device can be exquisitely complex. Fortunately, we do not have to walk this difficult path alone. International bodies such as CIE, ISO, ICC and SID have produced a multitude of standards and recommendations to guide best practice.
Against this background, Professor Green has provided an authoritative guide since the publication of his first book on the subject, Colour Engineering: Achieving Device-independent Colour almost 20 years ago. In this new book, Professor Green, aided by expert authors on specialist topics, brings us a thoroughly updated account of his subject, which covers the latest developments in the field. Here the reader will find guidance, formulae and best practice relating to all aspects of the colour recording, manipulation and reproduction pipeline with specialist chapters on such diverse topics as HDR rendering, AR and VR applications, web colour management and the impact of surface texture on colour perception and rendering. The text is logically and progressively presented, with sections covering the fundamentals of colorimetry, characterization and calibration of input and output devices, colour transformation and management protocols, followed by specialist topics. Explicit formulae and guidance are provided throughout the text, with copious references to the underlying adopted standards and recommendation documents in addition to research papers.
Colour reproduction is a topic of supreme importance, not only in display technology but also in manufacturing, graphic arts, publishing, broadcast and software development. This book will provide an invaluable reference to practitioners in all these disciplines and also serve as a guide to advanced students and those beginning their journey in colour engineering and deserves a place on the bookshelf of all whose concern is a faithful – or enhanced – rendering of colour.
Ian Sage Malvern, Worcestershire
Colour engineering, as presented in this book, represents the totality of disciplines involved in the acquisition, processing, synthesis and reproduction of colour images, using a wide range of devices. These colour imaging systems have become ubiquitous both in everyday life and in specialist, highly technical and highvolume applications.
Since the early days of digital colour imaging there has been a close collaboration between academics and industry-based scientists and engineers, who meet regularly in international scientific conferences and technical committees. This text aims to support the colour engineers of tomorrow, who are likely to be working in colour in web-based applications, in phones and in HDR displays, perhaps more so than in the more established industries of cameras, printers and SDR displays. The understanding of the relationship between device signals and the human vision system, and the colour gamut of an imaging device, are fundamental building blocks to all these application areas. Expectations of colour fidelity are no longer limited to 2D, planar diffuse colorimetry but are extending to spectral reproduction and total material appearance in 3D. International standards, developed by technical committees in ISO, IEC and CIE, play an important role in the interoperability of these technologies and their applications. The science and engineering of matching colour across different devices and platforms is defined in the well-established ICC.1 colour management architecture, while a key development in support of the colour engineering of tomorrow is the more flexible and more advanced ICC.2 (iccMAX) architecture, which has a chapter to itself.
I have been extremely fortunate that leading figures in all these cutting-edge areas of research and development agreed to contribute chapters to this volume. I have also added a few chapters myself to round out the range of subjects covered. No claim is made to be comprehensive, as it would take many volumes to do full justice to the state of the art in this field. A small number of the chapters include material that was previously published in Colour Engineering: Achieving Device-Independent Colour (2002) but have been comprehensively updated.
Series Editor Ian Sage provided valuable insights which helped to structure the content in some of the early chapters, and Wiley staff Sandra Grayson, Becky Cowan, Katherine Wong, Martin Tribe, Dilip Varma and Durgadevi Shanmugasundram all made important contributions at different stages in the development of the book.
I should very much like to thank all those who contributed to my own journey in the field of colour, although I fear that to do so would not leave much room in this book for the content. Instead, I will mention just a few and hope not to give too much offence to everyone else. Ronnier Luo and Tony Johnson supervised my PhD, and, with others at University of Leeds and the London College of Communication, provided much colourful inspiration; my friends and colleagues at the Norwegian University of Science and Technology provide a wonderful collegiate environment dedicated to colour and imaging; the members of the International Color
Preface
Consortium have shared their wide knowledge and experience; Pei-li Sun and his colleagues at National Taiwan University of Science and Technology allowed me to work with their excellent students; and my wonderful postgraduate students at LCC and NTNU and elsewhere over the years, have all added their own pieces to the puzzle. I should also particularly like to thank Eric Walowit, Peter Nussbaum and Aditya Sole, who kindly reviewed chapters in the book; and of course, to Ruth and Rosalie, my partner and daughter, who have always given their support.
The highly regarded colour scientist Danny Rich, who made an enormous contribution to colour science and standardisation activities over many decades, and wrote Chapter 1 in this book, tragically passed away in July 2022. I am grateful to Phyllis and Amanda Rich for their gracious support for the continuation of Danny’s work. The royalties from the book will be dedicated to a scholarship fund in his name (details are on the ICC web site www.color.org).
Phil Green NTNU
Introductory Notes
In colour engineering we are generally interested in recording colours in the world, communicating them, where necessary adjusting or enhancing them, and then producing them on some medium. The chapters in this book provide the necessary detail for each of these steps, but first we need to have some initial understanding of what colour is and what it means to record or generate a colour sensation by means of an imaging device.
Colour is best understood as a sensation produced in the brain, in response to an external stimulus. The human visual system is trichromatic, i.e., it has photoreceptor cells which respond differentially to three regions within the visible part of the electromagnetic spectrum, and as a result the visual response can be defined using a coordinate system with three dimensions and subsequently reconstructed using three primaries. The range of colours that can be recorded or reproduced is sometimes enhanced by adding intermediate hues to these three basic primaries.
THE COLOUR STIMULUS FUNCTION
The visual stimulus consists of light which strikes the photoreceptor cells after having been reflected off the surface of an object, transmitted through a transparent or translucent surface, or emitted by a source, which might be a lamp, a self-luminous display or some other emitter of light. (Note that an emitter of light is referred to as a “source” and not an “illuminant”, since the latter term is reserved for a numerical tabulation of the relative spectral power distribution of energy.)
A colour stimulus is one that is spectrally selective, and so the relative amounts of energy at different wavelengths within the visible spectrum are the starting point for any colour measurements or colorimetric data. To more precisely define the stimulus emitted in a particular direction and received by the eye, or by the aperture of a measurement instrument, a number of other physical quantities need to be defined.
The energy emitted in a flow of photons from a source is referred to as the flux, Ф. For the colour stimulus we are interested in the flux as a function of wavelength, which is denoted by the use of the Greek symbol λ, as in Ф(λ).
The flux within the visible spectrum can be weighted by a function that expresses the visual response to light at different wavelengths. This function is known as the photometric function, V(λ), and the flux thus weighted is known as the luminous flux. The SI unit of luminous flux is the lumen (lm), and when emitted uniformly over a solid angle of one steradian (sr), has a luminous intensity of one candela (cd). The brightness of an area emitting light (such as a surface viewed by reflected light, or a self-luminous display) can be
Introductory Notes
measured in candelas per square metre (cd m−2); while the intensity of luminous flux falling on a surface is known as the illuminance, whose unit is the lux. One lux (lx) falling uniformly on a surface corresponds to 1 lm m−2. For a perfectly diffuse reflecting surface, uniformly illuminated by x lux and emitting uniformly into an overlying hemisphere, the brightness of the surface in cd m−2 corresponds to x/π.
The luminous flux, luminous intensity, illuminance and luminance are all known as photometric quantities, since they describe the flux as seen by the human visual system. The corresponding unweighted radiometric units are the radiant flux, radiant intensity, irradiance and radiance, with units of Watts (W), W m−2, W sr−1 and W m−2 sr−1 respectively.
A more detailed treatment of the above quantities can be found in references 1–5.
RADIANCE AND REFLECTANCE
Radiance (L) is the flux from a surface for a given solid angle and projected area and has units of watts per steradian per square metre (W sr−1 m−2). Self-luminous systems such as displays, and natural scenes with varying illumination, can be considered in terms of the radiance reaching the observer (or the detector).
A perfectly diffuse surface which reflects flux Φ at the normal will reflect cos(θ)Φ at angle θ, although a surface with area A viewed at angle θ will have the same perceived brightness as when viewed at the normal, since its apparent area is cos(θ)A rather than A.
A radiance measurement is made directly from the surface of interest over a cone of approximately 5°. If the light from the sample is diffused before reaching the detector, the appropriate measured quantity is irradiance (the density of the incident radiant flux with respect to the area at a point on a surface), given the symbol E, in watts per square metre (W m−2).
While a spectral radiance measurement incorporates the spectral power distribution of the illumination, a spectral reflectance measurement is a measure of the surface independent of any illumination. The reflectance of light from a surface can be simplistically expressed as the ratio of reflected to incident flux, as shown in Equation 1, where Фr is the reflected flux and Фi is the incident flux.
(1)
Фi can be determined by direct measurement, although it is more practical to determine it by measuring a specimen with known reflectance (such as a calibration tile), and rearranging Equation 1 as Фi = Фr/R s , where Фr is the measured flux from the calibration tile and R s is its known reflectance.
The reflected quantity from a given surface will depend on several factors, including the geometry of both incident and detected flux, and is thus somewhat imprecise in the form shown in Equation 1. Instead, we can define reflectance factor more precisely as ratio of the flux reflected in the direction delimited by a given cone, Φr, to the flux reflected in the same direction by a perfect reflecting diffuser identically irradiated or illuminated, Φd. This is consistent with the definitions of reflectance factor adopted by CIE [6], ISO [7] and IEC [8] and applied in ICC colour management. It implies that a perfect diffuse reflector always has a reflectance factor of one at all wavelengths, and the reflectance factor of a given specimen is relative to this value.
Measured spectral reflectance factors of real surfaces do not normally have values above one, with certain exceptions such as some yellow inks, and white substrates with fluorescent whitening agents. (Note that reflectance factors may be given as either factors in the range [0,1] or percentages in the range [0,100].)
As described in Chapter 1, the CIE recommends two basic geometries for reflectance: 45:0 and d:0. Measurements made with a diffuser in the optical path between source and sample (d:0), or between sample and detector (0:d) can be made across the full range of geometric angles, or with the specular angle excluded. Where these specular-included (SPIN) and specular-excluded (SPEX) measurements for a sample are significantly different from each other it can be inferred that the sample has some level of gloss. The choice of 45:0 or d:0 depends mainly on the particular application and industry recommendation.
Spectral reflectance and spectral radiance (i.e., per unit wavelength) are given the notations R(λ) and L(λ).
To find the reflectance factor from a measurement of the flux (such as a measurement of spectral radiance), Equation 2 is used:
where Φr(λ) is the flux reflected by the specimen, and Φd(λ). is the flux reflected by a perfect diffuser with the same geometry of illumination and detection as for Φr(λ).
Measurements of reflectance or emission should where possible be traceable to a national standardising laboratory through a chain of calibration standards (such as calibrated tiles or lamps) with known uncertainties. A measurement R(λ) of a sample can by corrected by applying a calibration vector as shown in Equation 3:
where R(λ)calibration is found by measuring a calibration standard and dividing this measurement by the traceable reference data R(λ)reference for the same standard.
Materials that are viewed by transmitted light are measured in terms of their transmitted radiance or their transmittance. Transmittance factor, T, is usually measured relative to a perfect transmitting medium, and the equations for reflectance can be applied with T(λ) substituted for R(λ).
Radiance, reflectance and transmittance values are used in conjunction with a set of colour matching functions and an illuminant to calculate tristimulus values, as discussed in Chapters 1 and 2.
OBSERVER RESPONSE FUNCTIONS
The light sensitive cells in the human retina are known as rods (active mainly at low levels of illumination) and cones (active only at levels of illumination from approximately 50 lux). The cone cells have three different photopigments, with peak absorbences at 557, 532 and 426nm respectively (Merbs, 1992). These are known as the long (L), middle (M) and short (S) wavelength cones, and a trichromatic visual response can be defined in terms of the LMS ratios.
Since LMS cone activation cannot be measured directly, the observer response is quantified in terms of the relative response to three primary lights at different wavelengths across the spectrum in colour matching experiments. The relative response to any set of real red, blue and green primaries can be linearly transformed to any other set of primaries, and it is convenient to use three “imaginary” primaries X, Y and Z. This topic is discussed extensively in Chapters 1 and 2.
ACQUISITION AND RECONSTRUCTION
The visible spectrum can be divided into three regions, which are predominantly red, green and blue, with red at longer wavelengths and blue at shorter wavelengths. Imaging devices make use of these channels, either in wavelength-selective sensors in the case of an image acquisition device such as a camera; as selective emission regions, as in a display; or as selective absorption regions, as in a printer. These wavelength-selective regions in a colour device are optimized to ensure efficient coverage of the entire spectrum, but do not correspond to the selective absorption regions of the cone photopigments of the retina. So, while it is possible to think of the three channels of visual trichromacy as being “reddish”, “greenish” and “bluish”, it is important to avoid overly simplifying the relationship between trichromatic imaging device and trichromatic visual response. For this reason, the abbreviations L, M and S are used to indicate the long, middle and short-wavelength cones of the
Introductory Notes
visual system, while R, G and B indicate the three sensitivity bands of a camera, or the primary emission bands of a display.
Since the appearance of a colour is a matter of the observer response rather than a particular set of device signals, the appearance is said to be device-independent. In order to match a colour appearance or otherwise produce the desired colour on an imaging device, it is essential to know the relationship between the device signals and the observer response. This topic is covered in Chapters 3 to 7.
COMMUNICATION AND EXCHANGE
Rarely is colour communicated entirely within a closed system, and so accurate and reliable methods of communication of the different attributes of colour stimuli is essential. A range of internationally-agreed recommendations and standards exist to support such communication, and are referred to in Chapters 8 to 13.
APPLICATIONS
While the range of possible applications of colour engineering is extremely large, this book focuses on a number of emerging topics that are considered to be of interest to the colour engineer of the future. These include the appearance of material objects, high dynamic range imaging, augmented and virtual reality, and colour on the web. These topics are covered in Chapters 14 to 19. All these applications have the potential to be connected through the new ICC colour management architecture iccMAX, which is summarized in Chapter 11.
References
1 Grum, F. and Becherer, R. (1981). Optical Radiation Measurement: 1 Radiometry, Academic Press.
2 McDonald, R. (1997). Colour Physics for Industry Society of Dyers and Colourists.
3 Hunt , R.W.G. and Pointer, M. (2011). Measuring Colour, 4e. Wiley.
4 Wyszecki, G. and Stiles , W.S. (1982). Colour Science: Concepts and Methods, Quantitative Data and Formulae, 2e. Wiley.
5 Wright , W.D. (1973). The Measurement of Colour, 4e. Adam Hilger.
6 CIE 17:2020 ILV: International Lighting Vocabulary, 2e.
7 ISO 13655:2017 Graphic technology - Spectral measurement and colorimetric computation for graphic arts images.
8 IEC 60050-845:2020 International Electrotechnical Vocabulary (IEV) - Part 845: lighting.
Instruments and Methods for the Colour Measurements Required in Colour Engineering
Danny Rich
1.1 INTRODUCTION
1.1.1
The Need for Colorimetry
In the reproduction of colour and coloured images, the assessment of the colour appearance or colour match has historically been the job of experienced, trained artisans. They were schooled and apprenticed by a master colourist or artist who taught them how to hold the specimen or image, how to describe the appearance of the colour or colour difference and how to select the correct set of dyes, pigments or inks to create a workable recipe for the desired colour match. They also taught them their bias for certain colours and their methods for making any recipe uniquely their secret.
The master colourist could match any colour within the gamut of their known primaries and could find the most visually acceptable or pleasing near-match to any colour that was outside of their gamut. This approach worked well when the coloured product, be it an illuminated manuscript, a fine textile or the paint to decorate and protect the woodwork of a stately house, was conceived, reproduced and sold locally. In today’s fast moving, global marketplace, more objective and rigorous methods are required for creating, matching and reproducing coloured materials and images. This is especially true in the application of electronic imaging where the coloration process occurs in milliseconds rather than hours. Colorimetry is the technology that attempts to capture the essence of the visual sensation of the light reflected from or transmitted through coloured images or emitted by self-luminous images. Colorimetry then converts that essence into an objective nomenclature to communicate the colour or colour-differences to someone in a different place and time and still obtain the same level of fidelity and aesthetics. The remaining chapters in this book describe the engineering approaches to the creation and manipulation of coloured images and to many of the issues raised here, especially electronically communicating and reproducing coloured images, but all applications of colour engineering must begin with a basis in colorimetry.
1.1.2
The Principles of Colorimetry
Colorimetry, at its purest and most basic level, is quite simple. The technology was developed to answer one question, “Does this test colour match this reference colour?” Basic colorimetry can provide nothing more than
Sun Chemical
the answer to that question. To obtain further information is to extend the technology from the measurement of colour into the measurement of colour appearance. Topics like colour difference, colour constancy or lack of constancy, corresponding colour and answers to the question, “How does this image or image element appear?” require additional assumptions and technologies beyond the scope of colorimetry. If there is one principle that needs to be kept in mind while becoming proficient in colour engineering, it is that “Colorimetry does not describe what a person sees!” Colorimetry is fully enveloped by the technology of colour matching. The details in colorimetry are found in how the colour match is created and reported. There are many excellent introductory textbooks on colorimetry that provide the history and background of colour measurements. The one by Berns [1] is particularly useful as it introduces both object mode colorimetry and imaging colorimetry.
1.1.3 Making the Transition from What We “See” to Quantifying How We “Match” a Colour
It can sometimes be difficult for the novice colour engineer of the future to be completely comfortable with this limited definition of colorimetry. As pointed out in the earlier parts of this book, colour is what you see, and the human visual system is highly adapted to the visual perception of colour. Indeed, this is true, but words are often pregnant with deeper meanings and implications. So it is with the concept of perception. The perception of colour involves many factors, most of which require the interaction of neural processes and physical phenomena that are not yet fully understood. Sensation involves fewer processes and occurs at much earlier stages in the visual system. The assessment of a colour match involves the sensation of identity or difference in the two sources of colour. Colorimetry is still a visual measurement even though we may utilise a colorimeter to make the measurements more precise. In the following sections the various ways to construct a colorimeter will be described, as well as how to validate that the colorimeter is consistent with its design intent and how that colorimeter can be used to answer the question of what reference colour is matched by the test colour. From an understanding of visual colorimetry, the development of an international standard of colorimetry can be understood and finally its application to automatic, electronic colorimeters. In the years since the publication of the first edition of this book, many new international standards have been published providing guidance on the best practices for colour measurements.
Colorimetry was first commercialised for the characterisation of materials, mainly textiles and later for decorative coatings, commonly known as paints. Plastics and inks used for image reproduction by printing were added much later. Because the application was most analytical in terms of the content of colouring media in the materials the objects took on a central role. The traditional description of colorimetry was in terms of a triad of contributors, usually abbreviated as: “Source + Object + Observer”. This paradigm is not useful when describing the measurements of the colours in image reproduction. The image may be in the form of a reflective object, or a transmitting object or it may be completely self-luminous. Many times the desire is to have two or more of these viewing modes produce matching colours. For image reproduction, the more fundamental paradigm of a colour stimulus function and an observer function is more appropriate. This paradigm will be used in the following sections.
The colour stimulus function, Φ(λ), is the true complement to the observer functions or spectral tristimulus values, X(λ), Y(λ), Z(λ). The tristimulus values are then computed from the spectral product, wavelength by wavelength, of the colour stimulus function and the observer function. In most cases, the actual functional dependence is not known so the functions are approximated by tabulations at discrete points. The exact number of points depends on the spectral range and the required accuracy. The CIE has determined that the visible spectral range extends from 360nm to 830nm and recommends for highest accuracy, a table with 471 values or intervals of 1nm.
When performing colorimetry on self-luminous objects, displays, projections, the colour stimulus function is measured directly. When performing colorimetry on objects, the colour stimulus function is derived from the measurement of the incident light flux and the transmittance or reflectance of the image or image element, as a decimal not as a percentage.
1.2 VISUAL COLORIMETRY
1.2.1 A Method to Uniquely Map the Colour of Lights and Objects
Visual colorimetry is the most direct and accurate method to objectively quantify colour. It is also the most difficult. Visual colorimetry requires the colourist to mix lights of different colours until a match between the test colour and the mixture is obtained. There are several common components found in all colorimeters that include a source of light, a source of primary or mixing colours that together produce the colour stimulus function and the viewing optics that transfer the colour stimulus function to the observer. Figure 1.1 shows a block diagram of a visual colorimeter. The principle of visual colorimetry is very simple and very familiar to most people today. It is the same principle that is used to set the mood on the stage in a theatre, and to
Figure 1.1 Typical visual colorimeter with three primary filters.
Specimen
generate colour signals on television screens, computer displays and slide or motion picture film projectors. This approach to colorimetry is also appropriate to teach primary school children to mix tempera paints, dye our clothing or hair and is familiar to artists and technologists who mix colorants to match the colours of natural objects in a paint medium. The physics of how the primaries interact to form the final, coloured stimulus is different but the process and result is completely analogous.
The earliest commercial colorimeters were thus visual colorimeters. Donaldson [2] produced several different models of visual colorimeter. The light source was a stable, ribbon filament incandescent lamp which illuminated both halves of a bipartite visual field. A transparent, coloured specimen could be placed in one half of the visual field and located between the coloured primaries opened full and the viewing optics. In the other half, the mixture of the coloured primaries is adjusted by opening or closing shutters over the primaries allowing more or less of the light to pass through the primary filters. Opening the shutters together made the image brighter while closing them together made the image dimmer. The shutters on the test field could also be adjusted so that the light seen through the viewing optics, usually a telescope imaging the two light mixing chambers side by side into the observer’s eye, could be matched for both colour and brightness. When the observer was satisfied with the quality of the match the positions of each of the shutters would be recorded, fixing exactly the state of that colour mixture on that colorimeter. Essentially, this quantified the equality of the two colour stimulus functions. Those numbers could be communicated to anyone else in the world with the same model Donaldson colorimeter and the match could be visualised by setting the shutters to the same positions. But if the second laboratory did not have the same model of colorimeter, the shutter settings would be of little value. Visual colorimetry as practiced in this way was thus a rather limited tool, good for evaluation of repeated specimens of the same colour but not for communication of colours to engineers in outside laboratories. There remains a great deal of similarity between these early colorimeters and our modern high dynamic range visual display units that incorporate solid state light sources and electro-chemically controlled shutters.
1.2.2 Development of the CIE Method of Visual Colorimetry
In the late 1920s two researchers in the London area began studies to better quantify the methods of visual colorimetry. One was John Guild at the National Physical Laboratory, and the other was a graduate student at the Imperial College, W. David Wright. Being interested in both the basic science of colorimetry and the needs of commerce, Guild [3] built a colorimeter that was similar to those of Donaldson with red, green and blue filters as primaries but with finer resolution and the best optical characterisation available to a national standardising laboratory. Wright [4, 5], being a physics graduate student, was more interested in making the most thorough determination of these colour mixture functions and built his visual colorimeter using prism monochromators for the primaries and seems to have first used the term “trichromator” for such an instrument. Guild was able to convince seven people to go through the difficult task of making colour matches to 30 or so narrow bands of wavelengths. The result was a spectral curve of colour mixtures representing the amounts of each of Guild’s red, green or blue primaries that would be mixed to match a given wavelength of light. Figure 1.2 shows a typical set of colour mixture curves. Wright was a bit more successful at recruiting observers and had 10 people make matches on his trichromator. Wright’s colorimeter then produced curves that described the amounts of his primary wavelengths (460nm, 530nm, 650nm) required to match the spectral colours. Even before Guild had published Wright took the two sets of data and compared them. They found that the colour mixtures curves in terms of primaries, RGB, of both experiments were very similar and so, using linear algebra, they transformed the two sets of results to a common set of monochromatic primaries and then averaged the results. Finally, they normalised the results using a white point determined by NPL and made the middle wavelength or green mixture function identical to the CIE 1929 standard of photometry. Details of how they were able to achieve this can be found in the review by Fairman [6]. The resulting set of colour matching functions were transformed from real mixing primaries to an imaginary set of primaries with theoretical colours outside of the gamut of spectral colours and were adopted by the CIE in 1931. These colour matching functions, now known as spectral tristimulus values, have become an international Standard Observer for small visual fields (less than 4°). The
transformations resulted in colour mixing curves that were all positive since the primaries were taken from points outside of the gamut of real colours as can be seen in Figure 1.3. The CIE named the primaries X, Y and Z and the colour matching functions, since they were based on the average of 17 observers were given the symbols, x, y, z , which in modern notation would now be termed X(λ), Y(λ),and Z(λ), and hence the name spectral tristimulus values. Since the whole system was linear, by definition this would allow any visual colorimeter readings to be converted to CIE equivalent readings and thus standardise the whole process. Having the middle primary be equal to the photometric function allows one to estimate both chromaticness and luminance with one reading.
1.2 Colour mixture curves for matches to spectral lights using red, green, and blue primaries.
The whole process was repeated in the late 1950s by Stiles and Burch [7] in the UK and by Speranskya [8] in the USSR but using field sizes of 10°. Unfortunately, no large field luminance or brightness function has ever been standardised so they could not provide a reference to large field photometry. Even so, many people, most recently Trezona [9], have observed that the large field Y10(λ) function correlates well with luminance factor in surface colours. The data that were collected for the development of the large field colour matching functions are now considered to be the most accurate and reliable ever obtained. It has recently been confirmed by Stockman [10, 11] along with the most reliable estimates of cone fundamentals or spectral response curves. This new information was assembled by CIE TC 1–36 [12] into a new chromaticity diagram based on these physiological significant coordinates as well as a set of physiologically significant colour mixture functions [13, 14]. CIE Division 1 technical committee, 1–59 then developed a supplementary photopic photometric observer for large fields and the CIE published this information [15].
Just as the observer had to be characterised and standardised so too the sources of light used in object mode colorimetry had to be standardised. The CIE is primarily an organization of illumination engineers and not
1.3 CIE 1931 standard observer’s colour matching functions.
Figure
Figure
