Tài liệu Introduction to Flat Panel Displays docx

Brennesholtz Liquid Crystal Displays: Addressing Schemes and Electro-Optical Effects Ernst Lueder Reflective Liquid Crystal Displays Shin-Tson Wu and Deng-Ke Yang Colour Engineering: Achi

Introduction to

Flat Panel Displays

Series Editor:

Anthony C Lowe

Consultant Editor:

Michael A Kriss

Display Systems: Design and Applications

Lindsay W MacDonald and Anthony C Lowe (Eds)

Electronic Display Measurement: Concepts, Techniques, and Instrumentation

Peter A Keller

Projection Displays

Edward H Stupp and Matthew S Brennesholtz

Liquid Crystal Displays: Addressing Schemes and Electro-Optical Effects

Ernst Lueder

Reflective Liquid Crystal Displays

Shin-Tson Wu and Deng-Ke Yang

Colour Engineering: Achieving Device Independent Colour

Phil Green and Lindsay MacDonald (Eds)

Display Interfaces: Fundamentals and Standards

Robert L Myers

Digital Image Display: Algorithms and Implementation

Gheorghe Berbecel

Flexible Flat Panel Displays

Gregory Crawford (Ed.)

Polarization Engineering for LCD Projection

Michael G Robinson, Jianmin Chen, and Gary D Sharp

Fundamentals of Liquid Crystal Devices

Deng-Ke Yang and Shin-Tson Wu

Introduction to Microdisplays

David Armitage, Ian Underwood, and Shin-Tson Wu

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 Hoi-Sing Kwok

Projection Displays, Second Edition

Matthew S Brennesholtz and Edward H Stupp

Introduction to Flat Panel Displays

Jiun-Haw Lee, David N Liu and Shin-Tson Wu

Registered office

John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex,

PO19 8SQ, United Kingdom

For details of our global editorial offices, for customer services and for information about how to apply for permission to reuse the copyright material in this book please see our website at www.wiley.com.

The right of the author to be identified as the author of this work has been asserted in accordance

with the Copyright, Designs and Patents Act 1988.

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 the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher.

Wiley also publishes its books in a variety of electronic formats Some content that appears in print

may not be available in electronic books.

Designations used by companies to distinguish their products are often claimed as trademarks All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners The publisher is not associated with any product or vendor mentioned in this book This publication is designed to provide accurate and authoritative information in regard to the subject matter covered It is sold on the understanding that the publisher is not engaged in rendering professional services.

If professional advice or other expert assistance is required, the services of a competent professional should

be sought.

Library of Congress Cataloging-in-Publication Data

Lee, Jiun-Haw.

Introduction to flat panel displays / by Jiun-Haw Lee,

David N Liu, and Shin-Tson Wu.

Set in 9/11pt Times by Integra Software Services Pvt Ltd, Pondicherry, India

Printed in Great Britain by Antony Rowe Ltd, Chippenham, Wiltshire

Contents

2.3.5.1 Sunlight and blackbody radiators 22

4.9 Vertical alignment 83

6 Light-emitting diodes 137

7.5.1.3 HIL, EIL and p–i–n structure 200

Series Editor’s Foreword

Article 2 of the bylaws of the Society for Information Display begins “1 The purpose of SID shall be: a) To encourage the scientific, literary and educational advancement of information display and its allied arts and sciences .’’ This book series was begun eleven years ago with the express object of extending

that encouragement, which in the printed form amounted to publishing conference proceedings and aJournal of peer refereed papers, to the provision of a series of books which would satisfy the needs ofscientists and engineers working in the wide and complex field of displays More recently in 2006, wepublished “Fundamentals of Liquid Crystal Devices’’ by Deng-Ke Yang and Shin-Tson Wu (who – notcoincidentally – is a co-author of this book) That book extended the readership because it was writtenprimarily as a post graduate textbook

This latest volume in the series extends that educational scope still further by describing the operatingprinciples and the methods of fabrication of technologies used or of potential use in flat panel displays,their methods of addressing, systems aspects and the underpinning science Although general books

on flat panel displays have been published in the past, this is the first comprehensive flat panel displaytextbook to have been written at this academic level Its readership and its use will extend far beyond postgraduate courses as it offers in a single volume material of great value to practising industrial engineersand scientists across the whole range of flat panel technologies

In my foreword, I usually provide a précis of the contents of a book, but the authors have done this socomprehensively that such an effort on my part would be superfluous It merely remains for me to thankthem for the great effort they have put into writing this book and wholeheartedly to commend it to ourpresent and expanding readership

Anthony C LoweSeries EditorBraishfield, UK

About the authors

Jiun-Haw Lee

Jiun-Haw Lee received BSEE, MSEE and PhD degrees in electrical engineering in 1994, 1995 and 2000,respectively, all from National Taiwan University, Taipei, Taiwan From 2000 to 2003 he was with theRiTdisplay Corporation as the director In 2003 he joined the faculty of National Taiwan University inthe Graduate Institute of Photonics and Optoelectronics and the Department of Electrical Engineering,where he is currently an associate professor His research interests include organic light-emitting devices,display technologies and solid-state lighting

Dr Lee is a member of the IEEE, OSA, MRS and SPIE He received the Exploration Research Award

of Pan Wen Yuan Foundation and Lam Research Award in both 2005 and 2006 He has published over

40 journal papers, 100 conference papers and 20 issued patents

David N Liu

David N Liu has been the director of the Strategic Planning Division in the Display Technology Center(DTC) of the Industrial Technology Research Institute (ITRI) since 2006 He worked on IC and fieldemission displays at ERSO (Electronics Research and Service Organization)/ITRI and Bellcore (BellCommunication Research) from 1983 to 1996 He started his research and development work on plasmadisplay panels at Acer Peripheral Inc and AUO from 1996 to 2002 After his service at AUO, he was incharge of the flat panel display technology division in ERSO/ITRI until 2006

Dr Liu received his PhD degree in electrical engineering from New Jersey Institute of Technology in

1992 He has over 45 issued patents, 18 published papers and a contributed chapter of the Semiconductor Manufacturing Handbook (McGraw-Hill, 2005) He also successfully developed field emission displays,

plasma display panels and flat panel displays followed by the receipt of many awards from ITRI, PhotonicsIndustry and Technology Development Association, Administration Bureau of Science Base Industry Parkand the Ministry of Economic Affairs (MOEA) He was also a recipient of the Outstanding Project LeaderAward from MOEA in 2006

Shin-Tson Wu

Shin-Tson Wu is a PREP professor at the College of Optics and Photonics, University of Central ida (UCF) Prior to joining UCF in 2001, Dr Wu worked at Hughes Research Laboratories (Malibu,California) for 18 years He received his PhD in physics from the University of Southern California (LosAngeles) and BS in physics from National Taiwan University (Taipei)

Flor-Prof Wu has co-authored four books: Fundamentals of Liquid Crystal Devices (Wiley, 2006), duction to Microdisplays (Wiley, 2006), Reflective Liquid Crystal Displays (Wiley, 2001) and Optics and Nonlinear Optics of Liquid Crystals (World Scientific, 1993), six book chapters, over 300 journal

Intro-publications and 75 issued and pending patents

Prof Wu is a fellow of the IEEE, OSA, SID and SPIE He is a recipient of the SPIE G.G Stokesaward, SID Jan Rajchman Prize, SID Special Recognition Award, SID Distinguished Paper Award,Hughes team achievement award, Hughes Research Laboratories outstanding paper award, UCF Distin-guished Researcher Award and UCF Research Incentive Award He was the founding editor-in-chief of

the IEEE/OSA Journal of Display Technology.

Flat panel displays (FPDs) are everywhere in our daily lives: mobile phones, notebooks, monitors, TVs,traffic signals and electronic signage are a few examples Several FPD technologies, such as liquid crystaldisplays (LCDs), plasma display panels (PDPs), light-emitting diodes (LEDs), organic light-emittingdevices (OLEDs) and field emission displays (FEDs), have been developed They coexist because eachtechnology has its own unique properties and applications

However, due to the diversity of display materials and operating mechanisms, there has not been atextbook covering the fundamental physics of such a wide spectrum of display technologies There arebooks dedicated to a specific display technology or book chapters covering different display technologies.This book is intended as a textbook for senior undergraduate and graduate students with a wide variety

of backgrounds, such as electrical engineering, electronics, material science, applied physics and opticalengineering It can also be used as a reference book for engineers and scientists working in displayindustries Parts of the material in this book and its organization follow the course ‘Introduction todisplay technologies’, which has been taught by Jiun-Haw Lee in the Graduate Institute of Photonics andOptoelectronics (GIPO) and Department of Electrical Engineering, National Taiwan University (NTU),Taipei, Taiwan, since 2003

This book introduces basic operation principles and underlying physics for thin-film transistors (TFTs)LCDs, PDPs, LEDs, OLEDs and FEDs in each chapter The LCD is a nonemissive display From the elec-trical viewpoint, each pixel is a light switch driven by a TFT To reduce leakage current of the capacitor,the liquid crystal material should have a high resistivity Moreover, to achieve a high contrast ratio, mostdirect-view TFT LCDs require two absorption-type sheet polarizers These polarizers not only reduce thelight efficiency but also limit the LCD’s viewing angle Therefore, phase compensation films are requiredfor wide-view LCDs In contrast, the PDP is an emissive display It can be considered as consisting of mil-lions of miniature fluorescent lamps on a single panel LEDs and OLEDs are electroluminescent deviceswith crystallized semiconductors and amorphous organic materials, respectively Compared with liquidcrystal materials which are also organic compounds, OLED materials should exhibit a low resistivity toreduce ohmic losses A FED is a type of flat cathode ray tube, which has all the advantages of this maturetechnology

In this book, both basic physics and practical issues (such as material requirements, device tions, fabrication methods and driving techniques) of different display technologies are addressed Eachdisplay technology is at a different development stage; some are more mature than others Generallyspeaking, they are still advancing so rapidly that it is difficult to keep up with the technological advance-ments Thus, in this introductory book we have decided to emphasize the fundamental science and onlyhighlight the key technological advancements of each technology

configura-Another objective of this book is to provide background knowledge for readers from ary fields to stimulate new ideas Since display technologies cover very broad scientific spectra, anybreakthrough from any aspect may result in substantial progress in this industry Sometimes there is notonly competition but also cooperation among different display technologies For example, LCDs andLEDs are distinct technologies for different display applications However, LEDs can be also used as

interdisciplin-backlights for LCDs As a result, the color gamut is widened, the dynamic contrast ratio is enhancedand power consumption is reduced After reading this book, one may expect to have a whole picture ofdisplay technologies from scientific, technical and engineering viewpoints There are different kinds oftechnologies suitable for different sizes (ranging from smaller than an inch to more than a hundred inches

in diagonal measurement) and applications (such as outdoor, indoor and mobile displays) Furthermore,this book may serve as a stepping stone to more advanced research and development

The organization of this book is as follows Chapter 1 introduces the classifications and tions of display technologies, which are guidelines for developing a display and judging performance.Applications suitable for different technologies (LCD, PDP, LED, OLED and FED) are also illustrated.Displays are used to produce or reproduce color images In Chapter 2 we introduce the stages of colorformation from a scientific viewpoint Then, the chromaticity diagram is used to quantitatively describecolors Finally, one can use the background of color science to engineer the color performance of a dis-play Chapter 3 describes the TFTs based on semiconductor material, which are used to drive LCDs andOLEDs Since this is an introductory textbook, some basic semiconductor physics are first introduced,which is also useful knowledge for Chapter 6 Material aspects of amorphous silicon and polycrystallinesilicon are discussed Then, device structures and their performances are introduced Finally, drivingtechniques and circuits for LCDs and OLEDs are demonstrated Emerging TFT technologies, such asorganic and oxide TFTs, are briefly discussed

specifica-In Chapter 4 we begin with basic liquid crystal compound structures, mixture formulations and theirphysical properties, and then extend the discussion to device structures and display characteristics Threemajor LCDs are introduced: transmissive, reflective and transflective Most modern LCDs are of thetransmissive type However, these displays might be washed out by direct sunlight In contrast, reflectivedisplays work well under sunlight but are not readable in dark ambient To retain the good images of atransmissive display while keeping good sunlight readability, transflective LCDs have been developed.Chapter 5 gives an overview of PDP fundamentals We begin with a discussion of the physics of a

gas discharge, covering the reactions of gas discharges and I–V characteristics DC PDP and AC PDP

panels as well as surface discharge and vertical discharge approaches are introduced The panel processtechnologies and useful process approaches are also described Finally, we discuss system techniqueswith cell operation and driving mechanism

Semiconductor LEDs are discussed in Chapter 6 We start from the material system because this ines the emission wavelength Electrical properties of LEDs, typically p–n junctions, and correspondingoptical characteristics are then discussed The fabrication process is introduced, which highlights thepractical electrical, optical and thermal issues Finally, applications of LEDs for displays are described.Chapter 7 describes OLEDs, with fabrication processes and operation principles similar to LCDs andLEDs, respectively The chapter starts from the material aspect Opto-physical processes in an organicmaterial are introduced Electrical injection and transport in organic materials are then described Devicestructures and fabrication are then discussed One serious disadvantage of an OLED is its short lifetime;this issue is also addressed In Chapter 8 an overview of FED fundamentals is provided We begin bydiscussing the physics of field emission, covering the field enhancement and vacuum mechanism FEDstructure, display mechanism and various emitters are introduced The advantages and disadvantages ofusing low- and high-voltage phosphor are compared The panel process technology and useful processapproaches are also described Finally, system techniques are discussed

determ-Jiun-Haw Lee, TaiwanDavid N Liu, TaiwanShin-Tson Wu, Florida, USA

David N Liu is grateful to his colleagues in ITRI and AUO for useful discussions, and Ted Knoy forhis professional proofreading In particular, he would like to express his gratitude to his wife Janice forher patience and support during the period of writing the book.

Shin-Tson Wu is deeply indebted to his present and former group members at the University ofCentral Florida for their numerous technical contributions, and to Chi-Mei Optoelectronics for the fundingsupport He is grateful to his wife Cho-Yan for spiritual support during the writing of the book

Introduction

1.1 Flat panel displays

A display is an interface containing information which stimulates human vision Information may bepictures, animation, movies and articles One can say that the functions of a display are to produce orreproduce colors and images Using ink to write, draw or print on paper is a traditional display, like

a painting or a book However, the content of such a traditional display is motionless and typicallyinerasable In addition, a light source, synthetic or natural, is needed for reading a book or seeing apicture There are lots of electronic displays that use an electronic signal to create images on a panel andstimulate the human eye Typically, they can be classified as emissive and nonemissive Emissive displaysemit light from each pixel which constitutes an image on the panel In contrast, nonemissive displaysmodulate light, by means of absorption, reflection, refraction and scattering, to display colors and images.For a nonemissive display, a light source is needed Hence, these can be classified into transmissive andreflective displays One of the most successful display technologies for home entertainment is the cathoderay tube (CRT), which is in widespread use in televisions (TVs) CRT is already a mature technologywhich has the advantages of self-emission, wide viewing angle, fast response, good color saturation, longlifetime and good image quality However, a major disadvantage is its bulky size The depth of a CRT isroughly equal to the length and width of the panel For example, a monitor’s depth is about 40 cm for a19-inch (38.6 cm× 30.0 cm) CRT with an aspect ratio of 4:3 Hence, it is not very portable The bulkysize and heavy weight limit its applications

In this book, we introduce various types of flat panel displays (FPDs) As the name implies, thesedisplays have a relatively thin profile, i.e several centimeters or less For instance, the liquid crys-tal display (LCD) is presently the dominant FPD technology with diagonal sizes ranging from lessthan 1 inch (microdisplay) to over 100 inches Such a display is usually driven by thin-film transistors(TFTs) A liquid crystal (LC) is a light modulator because it does not emit light Hence, a backlightmodule is required for a transmissive LCD In most LCDs, two crossed polarizers are employed in order

to obtain a high contrast ratio The use of two polarizers limits the maximum transmittance to about35–40 %, unless a polarization conversion scheme is implemented Moreover, the optical axes of twocrossed polarizers are no longer perpendicular to each other when viewed at oblique angles A LC is abirefringent medium which means its electro-optic effects are dependent on the incident light direction.Therefore, the viewing angle of a LCD is an important issue Most wide-view LCDs require multipleoptical phase compensation films; one for compensating the crossed polarizer and another for the birefrin-gent LC Film-compensated transmissive LCDs exhibit a high contrast ratio, high resolution, crisp image,good color saturation and wide viewing angle However, the displayed images can be washed out under

Introduction to Flat Panel Displays J.-H Lee, D.N Liu and S.-T Wu

c

 2008 John Wiley & Sons, Ltd

direct sunlight For example, if we use a notebook computer at outdoor ambient, the images may not bereadable This is because the reflected sunlight from the LCD surface is much brighter than that trans-mitted from the backlight so that the signal-to-noise ratio is low A broadband antireflection coating willdefinitely help to improve the sunlight readability.

Another way to improve sunlight readability is to use reflective LCDs.1A reflective LCD uses ambientlight to produce the displayed images It does not carry a backlight; thus, its weight is reduced A wrist-watch is such an example Most reflective LCDs have inferior performances compared to the transmissiveones in contrast ratio, color saturation and viewing angle Moreover, at dark ambient a reflective LCD isnot readable As a result, its application is rather limited

To overcome the sunlight readability issue while maintaining high image quality, a hybrid displaycalled a transflective liquid crystal display (TR-LCD) has been developed.2In a TR-LCD, each pixel isdivided into two subpixels: transmissive (T) and reflective (R) The area ratio between T and R can beadjusted depending on the application For example, if the display is mostly used outdoors, then one candesign to have 80 % reflective area and 20 % transmissive area In contrast, if the display is mostly usedindoors, then one can have 80 % transmissive area and 20 % reflective area Within this TR-LCD family,there are still some varieties: double cell gap versus single cell gap, and double TFTs versus single TFT.These approaches are trying to solve the optical path length disparity between the T and R subpixels

In the transmissive mode the light from the backlight unit passes through the LC layer once, but in thereflective mode the ambient light traverses the LC medium twice To balance the optical path length, wecould make the cell gap of the T subpixels twice as thick as that of the R subpixels This is the so-calleddual cell gap approach The single cell gap approach has a uniform cell gap throughout the T and Rregions To balance the different optical path lengths, several approaches have been developed, e.g dualTFTs, dual fields (stronger field for T region and weaker field for R region) and dual alignments Presently,the majority of TR-LCDs adopt the double cell gap approach for two reasons: (1) both T and R modes canachieve maximum light efficiency, and (2) the gamma curve matching between the voltage-dependenttransmittance (VT) and reflectance (VR) is almost perfect However, the double cell gap approach hastwo shortcomings: first, the T region has a slower response time than the R region because its cell gap

is about twice as thick as that of the R region; second, the viewing angle is relatively narrow, especiallywhen homogeneous cells are employed To widen the viewing angle, a special rod-like LC polymericcompensation film has to be used Chapter 4 gives detailed descriptions of various types of LCDs

A plasma display panel (PDP) is an emissive display which can be thought of as very many miniaturefluorescent lamps on a panel As an emissive display it typically has a better display performance, such

as good color saturation and wide viewing angle Due to the limitation of fabrication, the pixel size of aPDP cannot be too small For a finite pixel size, the video content is increased by enlarging the panel size.PDPs are suitable for large-screen applications In 2008, Panasonic demonstrated a 150-inch PDP TVwith 4096× 2160 pixels This resolution is four times higher than that of the present full high-definitiontelevision (HDTV)

Light-emitting diodes (LEDs) and organic light-emitting devices (OLEDs) are electroluminescentdevices with semiconductor and organic materials, respectively Electrons and holes recombine withinthe emissive materials, where the bandgap of the materials determines the emission wavelength

A field emission display (FED) uses sharp emitters to generate electrons These electrons bombard thephosphors that are present to emit red (R), green (G) and blue (B) light A FED is like a ‘flat’ CRT Due

to the mature technologies developed in CRTs, FEDs exhibit all the advantages of CRTs plus the smallerpanel thickness

Compared to conventional displays (such as books, magazines and newspapers), electronic displays(such as TVs, mobile phones and monitors) are rigid because they are typically fabricated on glasssubstrates Flexible FPDs are emerging Several approaches have been developed, such as electrophoreticdisplays and polymer-stabilized cholesteric displays Flexible displays are thin, robust and lightweight

In the remainder of this chapter, we first introduce FPD classifications in terms of emissive and missive displays, where nonemissive displays include transmissive and reflective displays Specifications

none-of FPDs are then outlined Finally, the FPD technologies described in the later chapters none-of this book arebriefly introduced.

1.2 Emissive and nonemissive displays

Both emissive and nonemissive FPDs have been developed For emissive displays, each pixel emits lightwith different intensity and color which stimulate the human eye directly CRTs, PDPs, LEDs, OLEDsand FEDs are emissive displays An emitter is called Lambertian when the luminances from differentviewing directions are the same Most emissive displays are Lambertian emitters which results in a wideviewing angle performance Also, due to the self-emissive characteristics, they can be used even undervery low ambient light When such displays are turned off, they are completely dark (ignoring the ambientreflection) Hence, display contrast ratios (see also Section 1.3.3) are high

Displays that do not emit light themselves are called nonemissive displays A LCD is a nonemissivedisplay in which the LC molecules in each pixel work as an independent light switch The externalvoltage reorients the LC directors which causes phase retardation As a result, the incident light fromthe backlight unit or ambient is modulated Most high-contrast LCDs use two crossed polarizers Theapplied voltage controls the transmittance of the light through the polarizers If the light source is behind

the display panel, the display is called a transmissive display It is also possible to use ambient light as

the light source This resembles the concept of a conventional display, such as reading a book, which is

called a reflective display Since no backlight is needed in a reflective display, its power consumption is

relatively low

In a very bright environment, images of emissive displays and transmissive LCDs can be washed out

In contrast, reflective displays exhibit an even higher luminance as the ambient light increases However,they cannot be used in a dim environment Hence, transflective LCDs have been developed, which aredescribed in Chapter 4

1.3.1 Physical parameters

The basic physical parameters of an FPD include display size, aspect ratio, resolution and pixel format.The size of a display is typically described by diagonal length, in units of inches For example, a 15-inchdisplay means the diagonal of the viewable area of this display is 38.1 cm There are three kinds ofdisplay format: landscape, equal and portrait, corresponding to the display width being larger than, equal

to and smaller than its length Most monitors and TVs use landscape format with a width-to-length ratio,which is called the ‘aspect ratio’, of 4:3, 16:9 or 16:10, typically

An FPD typically consists of a ‘dot matrix’ which can display images and characters To increaseresolution, one may use more dots in a display Table 1.1 lists some standard resolutions of FPDs For

Table 1.1 Resolution of FPDs.

SXGA Super extended graphics array 1280 × 1024

UXGA Ultra extended graphics array 1600 × 1200

WXGA Wide extended graphics array 1366 × 768

WSXGA Wide super extended graphics array 1680 × 1050

WUXGA Wide ultra extended graphics array 1920 × 1200

example, VGA means the display is 640 dots in width and 480 dots in length Higher resolution typically(but not necessarily) means better image quality There are some resolutions listed in Table 1.1 startingwith the letter ‘W’, which means wide screen with an aspect ratio larger than 4:3 Once the resolution,display size and aspect ratio are known, one may obtain the pitch of the pixels For example, a 19-inchdisplay with aspect ratio of 4:3 and resolution of UXGA has a pitch of 190.5 m Note that not all ofthe pixel area contributes to the display One can define the ‘fill factor’ or ‘aperture ratio’ as the ratio ofthe display area in a pixel over the whole pixel size, with its maximum value of 100 % Besides, for afull-color display, at least three primary colors are needed to compose a color pixel Hence, each colorpixel is divided into three subpixels (RGB) sharing the area For example, let us assume a color pixel hassize of 240 m× 240 m; then the dimension of each subpixel is 80 m × 240 m If the fill factor is

81 % which actually contributes to light emission or transmission, then the usable pixel area is reduced

to 72× 216 m2

There are different layouts for RGB subpixels, as shown in Figure 1.1 For the stripe configuration,

it is straightforward and easy for fabrication and driving circuit design However, it has a poor colormixing performance for the same display area and resolution For mosaic and delta configurations, theirfabrication and/or driving circuit are more complicated but their image quality is better because of bettercolor mixing capability Also, displays with mosaic and delta configurations exhibit faster response timessince the moving distance between the pixels is shorter Actually, as the resolution gets high enough thesubpixel arrangement becomes less critical For medium and large displays, the stripe configuration istypically used In contrast, for a small-size display which requires high resolution, e.g video cameras,one may use the mosaic or delta configuration

1.3.2 Brightness and color

Luminance and color are two important optical characteristics of an FPD A display with high luminancelooks dazzling in a dark room On the other hand, a display with insufficient brightness appears washedout under high ambient Typically, the luminance of an FPD should be as bright as (or slightly brighterthan) the real object Under an indoor lighting environment, a monitor has a luminance of 200–300 cd m−2(Section 2.3.6) For a large-screen TV, a higher luminance (500–1000 cd m−2) may be needed An FPD

is used to produce or reproduce colors; hence, how many colors of an FPD and how real the color is(color fidelity) between an FPD and a real object are two important characteristics of an FPD Since thecolor of an FPD is mixed by (at least) three primary colors, i.e RGB, more ‘pure’ (saturated) primariesresults in a broader range of the possibly displayed colors, which is called ‘color gamut’ (Section 2.3.4).One can equally divide the stimuli to the eyes from dark to bright with 2, 4, 8 or more spacings, which

is called ‘gray level’ or ‘gray scale’ (Section 2.3.3) For example, an FPD can display 16 million colors(28× 28× 28≈ 16.8 million) when each RGB subpixel is divided into 8 gray scales

in a perfectly dark room However, due to the surface reflection from the ambient, Equation (1.1) should

be modified to

A-CR=Lw+ Lar

Lb+ Lar

where A-CR is the ambient contrast ratio and Laris the luminance from ambient reflection A-CR is used

to specify the ambient contrast ratio, to distinguish from the intrinsic ‘device’ contrast ratio as described

in Equation (1.1) From Equation (1.2), as the ambient reflection increases, A-CR decreases sharply Tokeep a good ambient contrast, one can: (1) increase the on-state luminance, and (2) reduce the reflectivity

of the display surface However, for a very strong ambient, e.g in sunshine outdoors, luminance fromthe direct sun is four orders of magnitude higher than that of an FPD, which severely washes out theinformation content of the FPD Sunlight readability is an important issue especially for mobile displays

In contrast, an adequate ambient light is required for conventional displays, such as books or newspapers

A similar situation applies to reflective displays, such as reflective LCDs

1.3.4 Spatial and temporal characteristics

Uniformity of an FPD means the luminance and color change over a display area Human eyes aresensitive to luminance and color differences For example, a 5 % luminance difference is noticeablebetween two adjacent pixels For a gradual change, human eyes can tolerate up to 20 % luminancechange over the whole display

Optical characteristics (luminance and colors) may also change at different viewing angles ForLambertian emitters, such as CRTs, PDPs and FEDs, viewing angle performances are quite good Theemission profile of LEDs and OLEDs can be engineered by packaging and layer structure However,the viewing angle of LCDs is one of the major issues because LC material is birefringent and crossed

polarizers are no longer crossed when viewed at oblique angles There are several ways to define the ing angle of an FPD For example, to find the viewing cone with: (1) a luminance threshold; (2) minimumcontrast ratio, say 10:1; or (3) maximum value of color shift For some cases that contrast ratio is smallerthan 1; this is called ‘gray level inversion’.

view-Response time is another important metric If an FPD has a slow response time, one may see blurredimages for fast moving objects By switching the pixel from ‘off’ to ‘on’ and from ‘on’ to ‘off’, andcalculating the time required from 10 to 90 % and 90 to 10 % luminance levels, one can obtain rise andfall time, respectively One may also define the response time from one gray level to another, which iscalled the ‘gray-to-gray’ (GTG) response time Most display scenes contain rich grayscales Therefore,GTG response time is more meaningful For LCDs, this GTG response time can be much longer than theblack-to-white rise and fall time.3A TFT is a holding type of active matrix It is different from the CRT’simpulse type Therefore, a motion picture response time4is commonly used to define the response time

of a TFT LCD

After long-term operation, the luminance of an FPD (especially an emissive display) decays In anemissive display, if a fixed pattern is lit on for a long period of time before all the pixels are turned on forthe full white screen, one can see nonuniformity of the fixed pattern with a lower brightness, which iscalled the ‘residual image’ As mentioned before, the human eye can detect less than 5 % nonuniformitybetween two adjacent pixels Hence the lifetime of an FPD is crucial for static images An alternativesolution is to use moving pictures, rather than static images, for information display Then the luminances

of all pixels decay uniformly, since the average on time for all pixels is the same

1.3.5 Efficiency and power consumption

Power consumption is a key parameter, especially for mobile displays, as it affects battery life Fordisplays with wall-plug electrical input, lower power consumption implies lower heat generation, whichmeans heat dissipation is less serious Typically, one uses the unit lm W−1to describe power efficiency

of an FPD (Section 2.3.6) Lumen (lm) and watt (W) are units for describing light output and electricalinput A portable display with lower power consumption leads to a longer battery life For notebooks andTVs, high optical efficiency also translates into less heat dissipation and a lower electricity bill Thermalmanagement in a small-chassis notebook is an important issue

1.3.6 Flexible displays

An FPD is usually fabricated on thin glass plates Glass is a kind of rigid substrate In contrast, conventionaldisplays are printed on paper, which is flexible An interesting research topic is to fabricate FPDs onflexible substrates, as a ‘paper-like’ display.5Compared to the glass-based FPDs, flexible displays are thinand lightweight Also, flexible displays can be fabricated by the roll-to-roll process, which is potentially oflow cost Substrate selection of flexible FPDs includes ultrathin glass, plastic and stainless steel Bendableultrathin glass substrate is achievable, but the cost is high Plastic substrate is suitable for flexible displays,but the highest durable temperature is typically lower than 200◦C Stainless steel substrate is bendable,and durable for high temperature; however, it is opaque hence not suitable for transmissive displays.There are many technical bottlenecks for flexible FPDs, such as material selection, fabrication processes,device configurations, display package and measurement

1.4 Applications of flat panel displays

The following subsections briefly outline the applications of each technology Detailed mechanisms aredescribed in the related chapters

1.4.1 Liquid crystal displays

Although LC materials were discovered more than a century ago,6,7their useful electro-optic effects andstability were developed only in the late 1960s and 1970s In the early stage, passive matrix LCDs werefound useful in electronic calculators and wristwatches.8With the advance of TFTs,9color filters10andlow-voltage LC effects,11active matrix LCDs have gradually penetrated into the market of notebookcomputers, desktop monitors and TVs Today, LCDs have found widespread uses in everyday life,including (1) mobile applications, such as mobile phones, personal digital assistants, navigation systems,notebook personal computers; (2) office applications, such as desktop computers and video projectors;and (3) home applications, such as large-screen TVs.12

To satisfy these wide-spectrum applications, three types of LCDs have been developed: transmissive,reflective and transflective Transmissive LCDs can be further separated into projection and direct-view

In a small-size, high-resolution LCD, the pixel size is around 40 m× 40 m Here, the apertureratio becomes particularly important because it affects the light throughput.13To enlarge the apertureratio, poly-silicon (p-Si) TFTs are commonly used because their electron mobility is about two orders ofmagnitude higher than that of amorphous silicon (a-Si) High mobility allows a smaller TFT to be usedwhich, in turn, enlarges the aperture ratio For the detailed structure of a TFT LCD, see Figure 4.1.For direct-view transmissive TFT LCDs, the pixel size (∼300 m × 300 m) is much larger thanthat of a microdisplay Thus, a-Si is adequate although its electron mobility is relatively low Amorphous

silicon is easy to fabricate and has good uniformity Thus, a-Si TFTs dominate the large-screen (>10

inches) LCD panel market

Similarly, reflective LCDs can also be divided into projection and direct-view displays In projectiondisplays using liquid-crystal-on-silicon (LCoS),14the pixel size can be as small as∼10 m × 10 mbecause of the high electron mobility of crystalline silicon (c-Si) In a LCoS device, the electronic drivingcircuits are hidden beneath the metallic reflector Therefore, the aperture ratio can reach 90 % and thedisplayed picture is film-like In contrast, most reflective direct-view LCDs use a-Si TFTs and a circularpolarizer Their sunlight readability is excellent, but they are not readable in dark ambient Therefore,the application of reflective direct-view LCDs is rather limited

To maintain high-quality transmissive display and good sunlight readability, a hybrid TR-LCD has beendeveloped In a TR-LCD, each pixel is divided into two subpixels: one for transmissive and another forreflective display.15In dark to normal ambient, the backlight is on and the TR-LCD works as a transmissivedisplay Under direct sunlight, the TR-LCD works in reflective mode Therefore, its dynamic range iswide and its functionality does not depend on the ambient lighting conditions TR-LCDs have beenwidely adopted in portable devices, such as mobile phones For a detailed discussion of TR-LCDs, seeChapter 4

1.4.2 Light-emitting diodes

A LED is an electroluminescent device based on crystalline semiconductors.16To convert electrical tooptical power, one has to inject carriers into the LED through electrodes, and then they recombine togive light The emission wavelength is mainly determined by the semiconductor materials, and can befine tuned by device design

Since it is difficult to grow large-size single crystals, the wafer diameter of LEDs is limited to about

8 inches After device processing, LEDs are diced from the wafer followed by the package process Thedimension of a single LED is typically several millimeters, which means the ‘pixel size’ of the LED

is large Hence, it is difficult to use a LED as a small display or it will have a very low resolution Anexception is to dice LED arrays from a wafer and use as a microdisplay with a size less than 1 to 2 inches.Due to their self-emissive characteristic, LEDs are commonly used for large displays, such as outdoorsignages (single color, multicolor and full color), traffic signals and general lighting to replace light bulbs.Compared to conventional displays enabled by light bulbs, LED displays exhibit the advantages of lower

power consumption, greater robustness, longer lifetime and lower driving voltage (so safer) There arealso lots of outdoor screens with diagonals of over 100 inches which consist of millions of LED pixels.Rather than a display itself, a LED can also be used as the light source, such as the backlight modulefor a LCD, and general lighting Compared to a conventional cold cathode fluorescent lamp (CCFL),which resembles a thin fluorescent tube, a LED exhibits a better color performance, longer lifetime andfaster response Another important driving force to the use of LEDs as LCD backlights is that the mercury

in CCFLs is harmful to the environment When using LEDs for general lighting applications, a broadspectrum is preferred to simulate natural light, such as sunlight, for obtaining a high color rendering

of reflective objects (Section 2.3.5) This is quite different from the requirements for LED displays andLCD backlights, which usually need a narrow spectrum

1.4.3 Plasma display panels

The typical structure and operation principle for PDPs are similar to those of a fluorescent lamp In thestructure of a fluorescent lamp, two filament electrodes are formed in two ends of an inner glass tube.The wall of the inner glass tube is coated with phosphor The cavity of the glass tube is filled with a gasmixture of argon and mercury When a certain voltage is applied to the electrodes, plasma is generatedfrom a gas discharge Due to the energy level system of the plasma, ultraviolet (UV) radiation is generated

with peak wavelength at λ= 254 nm The phosphor of the fluorescent lamp is excited by the UV radiationwhich, in turn, emits light

PDPs use a similar operation mechanism to fluorescent lamps but the gases commonly used in PDPsare neon and xenon instead of the argon and mercury used in fluorescent lamps Neon and xenon gasesgenerate peak wavelengths at 147 and 173 nm which belong to the vacuum ultraviolet (VUV) region.VUV radiation can only propagate in a vacuum because it is strongly absorbed by air Although the PDPstructure is similar to a fluorescent lamp which is composed of two electrodes, phosphor and gases, anadditional barrier rib structure is needed in PDPs to sustain the space between upper plate and lowerplate.17

Because of the structure of the barrier rib, the unit cell size of PDPs cannot be made too small Inaddition, PDP operation voltage is high because a typical plasma generation is needed The high operationvoltage demands a high voltage driver integrated circuit (IC) and results in a high cost of the electronics.However, PDPs exhibit a wider view angle, faster response time and wider temperature range than LCDs

In other words, PDPs remain good candidates for large-panel displays spanning from static pictures tomotion pictures, from cold ambient to hot ambient and from personal use to public use In addition to theseperformance advantages, PDPs can be fabricated with a low-cost and simple manufacturing process Forthese reasons, many different PDP structures intended for a wide spectrum of applications have beendeveloped.18−20

1.4.4 Organic light-emitting devices

An OLED is also an electroluminescent device, like a LED, except its materials are organic thin films withamorphous structures.21Amorphous organic material has a much lower mobility (typically five order ofmagnitude lower) than crystalline semiconductors, which results in a higher driving voltage of OLEDs.Also, the operation lifetime of OLEDs is one order of magnitude shorter than semiconductor LEDs

However, due to the amorphous characteristics, fabrication with large size (>40 inches) is possible.

Since the conductivity of amorphous organic materials is very low, very thin organic films (100–

200 nm in total) are required to reduce the driving voltage to a reasonable value (i.e <10 V) This is

quite a challenge in thin-film formation, especially for large-size substrates There are several fabricationtechnologies proposed, such as physical vapor deposition, spin coating, ink-jet printing and laser-assistedpatterning Prototypes of OLED panels of 40 inches have been demonstrated, and OLEDs of 11 inches

(or less) are also commercially available.22,23Another challenge for large displays is the reliability A

TV must have a longer lifetime than a mobile phone Recently, commercial OLED products for mobiledisplays and monitors have emerged However, for TV applications, OLED panel lifetime still falls short.Two advantages of OLEDs are: (1) low process temperature, and (2) nonselective to the substratematerial, which is suitable for flexible displays One of the strategies for OLED development is toimprove device performance (especially driving voltage and lifetime) so as to be as good as (or not toomuch worse than) LEDs Also, due to the possibility of large-size fabrication, the potential manufacturecost of OLEDs is lower than that of LEDs Because OLEDs have some advantages in performance andfabrication cost over LEDs, they have a chance to replace LEDs in some applications since they areboth electroluminescent devices with similar operation principles Besides, in comparison with LEDs,OLEDs have two unique advantages: larger panel size and higher resolution

1.4.5 Field emission displays

A FED is a display using electrons generated by field emission to excite phosphors and generate ance There are several different approaches to generate electron emission, such as thermionic emission,photoemission and field emission.24,25 Thermionic emission electrons are thermally excited over thepotential energy barrier while photoemission electrons are excited over the potential energy barrier byincoming photons In field emission, the electrons tunnel though the surface potential energy barrier,which has been thinned and shaped by the influence of a strong electric field

lumin-The field emitter plays an important role in the electron emission of FEDs lumin-The structure of the fieldemitter can be in the shape of a cone, a wedge, a cylinder or a tube.26The emitting region is a tip for

a conic-shape emitter while the emitting region is an edge for wedge, cylinder and tube shapes Thereare also many types of emitters such as the Spindt emitter, carbon nanotube (CNT) emitter and surfaceconduction emitter (SCE).27A Spindt emitter uses sharp conic material as an emitter while a CNT emitteruses a carbon tube of nanometric diameter as an emitter A SCE uses a material named PdO as an emitterwith a nano-gap structure to generate surface electrons These types of emitters can be undesirablydamaged by the ions generated from residual gas This undesirable damage usually results in a shortlifetime of operation Therefore, less residual gas and lower operating voltage are strongly demanded inFEDs In order to have a lower operating voltage, an emitter material of low work function with a sharpstructure is desired In addition, vacuum is also required so that residual gas can be eliminated.Both FEDs and CRTs use phosphors to generate visible light which demands a vacuum to ensure longlife of electron emission The structure of a FED consists of an emitter, electrode and phosphor which issimilar to that of a CRT The display performance is also similar to a CRT However, a spacer structure

is needed in a FED to maintain the space between phosphor plate and field emission plate The emissionuniformity which is caused by the process of emitter formation is the major challenge for FEDs Aboveall, FED structure is simple because it does not require backlight, color filters, polarizers or other opticalfilms which are needed in LCDs Furthermore, FEDs have higher luminance efficiency, faster responsetime, a wider view angle and greater temperature range than LCDs.28FEDs can be widely applied fromstatic pictures to motion pictures, from cold ambient to hot ambient and from personal use to public use.29

References

1. Wu, S.T and Yang, D.K (2001) Reflective Liquid Crystal Displays, John Wiley & Sons, Ltd, Chichester.

2 Okamoto, M., Hiraki, H and Mitsui, S (2001) US Patent 6,281,952.

3 Wang, H., Wu, T.X., Zhu, X and Wu, S.T (2004) Correlations between liquid crystal director reorientation and

optical response time of a homeotropic cell J Appl Phys., 95, 5502.

4. Song, W., Li, X., Zhang, Y et al (2008) Motion-blur characterization on liquid-crystal displays J SID, 16, 587.

5. Crawford, G.P (2005) Flexible Flat Panel Displays, John Wiley & Sons, Ltd, Chichester.

6. Reinitzer, F (1888) Monatsh Chem., 9, 421.

7. Lehmann, O (1889) Z Phys Chem., 4, 462.

8. Ishii, Y (2007) The world of liquid crystal display TVs: past, present and future J Display Technol., 3, 351.

9. Lechner, B.J., Marlowe, F.J., Nester, E.O and Tults, J (1971) Liquid crystal matrix displays Proc IEEE, 59,

1566.

10. Fischer, A.G et al (1972) Design of a liquid crystal color TV panel Proceedings of the IEEE Conference on

Display Devices, p 64.

11 Schadt, M and Helfrich, W (1971) Voltage-dependent optical activity of a twisted nematic liquid crystal.

Appl Phys Lett., 18, 127.

12. Liu, C.T (2007) Revolution of the TFT LCD technology J Display Technol., 3, 342.

13. Stupp, E.H and Brennesholtz, M (1998) Projection Displays, John Wiley & Sons, Inc., New York.

14 Armitage, D., Underwood, I and Wu, S.T (2006) Introduction to Microdisplays, John Wiley & Sons, Ltd,

Chichester.

15. Zhu, X., Ge, Z., Wu, T.X and Wu, S.T (2005) J Display Technol., 1, 15.

16. Round, H.J (1907) A note on carborundum Electrical World, 19, 309.

17 Fischer-Cripps, A.C., Collins, R.E., Turner, G.M and Bezzel, E (1995) Stress and fracture probability in

evacuated glazing Building Environ., 30, 41.

18. Oversluizen, G and Dekker, T (2006) High efficacy PDP design SID Symp Dig., 37, 1110.

19. Hirakawa, H., Shinohe, K., Tokai, A et al (2004) Dynamic driving characteristics of plasma tubes array SID

Symp Dig., 35, 810.

20. Sano, Y., Nakamura, T., Numomura, K et al (1998) High-contrast 50-in color ac plasma display with 1365×

768 pixels SID Symp Dig., 29, 275.

21. Tang, C.W and Vanslyke, S.A (1987) Organic electroluminescent diodes Appl Phys Lett., 51, 913.

22. Iino, S and Miyashita, S (2006) Printable OLEDs promise for future TV market SID Symp Dig., 37, 1463.

23. Hirano, T., Matsuo, K., Kohinata, K et al (2007) Novel laser transfer technology for manufacturing large-sized

OLED displays SID Symp Dig., 38, 1592.

24. Gomer, R (1961) Theory of Field Emission: Field Emission and Field Ionization, Harvard University Press,

Cambridge, MA.

25. Dyke, W.P and Dolan, W.W (1956) Field emission, in Advances in Electronics and Electron Physics, Vol 8 (ed.

L Marton), Academic Press, New York.

26. Liu, D., Ravi, T.S., Gmitter, T et al (1991) Fabrication of wedge-shaped silicon field emitters with nm scale

radii Appl Phys Lett., 58, 1042.

27. Okuda, M., Matsutani, S., Asai, A et al (1998) Electron trajectory analysis of surface conduction electron

emitter displays SID Symp Dig., 29, 185.

28. Utsumi, T (1991) Keynote address vacuum microelectronics: what’s new and exciting IEEE Trans Electron

Dev., ED-38, 2276.

29. Itoh, S et al (2007) Development of field-emission display SID Symp Dig., 38, 1297.

As illustrated in Figure 2.1(a), the human eye sees the color of an object under sunlight, which is a

‘white’ light source, because its spectral bandwidth covers the entire visible range and is of relativelyuniform intensity across the range If there were no light source, there would be no photons to stimulatethe human eye and, therefore, no color would be observed When illuminated, an object (e.g the paper

in Figure 2.1(a)) absorbs some of the incident photons and reflects the rest As shown in Figure 2.1(b),there are yellow and green inks on the white paper When the incident white light illuminates the yellowink, the blue component of the white light is absorbed The reflected red and green components result in

a yellow light Similarly, the green ink absorbs red and blue components Wherever there is no ink, thewhite paper reflects all the incident light, so it appears white From the above discussion, we can deducethat the color of an object is also dependent on the spectral distribution of the incident light For example,

if the light source is red, then yellow ink will appear red After the light–object interaction, the reflectedphotons are received by the detector; in this case a human eye The human eye can distinguish lightintensity and color, but not the polarization state and phase of incident light Variation of the intensity

of incident light gives the observer a perception of bright and dark Photons of different wavelengthsstimulate differently the photosensitive cells (cone and rod cells, which are discussed later) of the eyeand this creates the perception of different colors

There are three different cone cells in the human eye, with different spectral sensitivities which make

it possible to use three primary colors (red, green and blue) to generate all colors (trichromatic space) and

to describe colors quantitatively.1In 1931 the Commission Internationale de l’Eclairage (CIE) suggestedthe (X, Y, Z) colorimetric system, which can specify all the colors by their distinct coordinates.2It is aconvenient system for describing colors However, the 1931 CIE system is not suitable for discussing themagnitude of the perceived difference between two colors To solve this problem, uniform color spacesare proposed (e.g the 1976 CIE (L∗u∗v∗)- and (L∗a∗b∗)-spaces).2In such a system, the perceived colordifference for two mismatched colors is nearly identical at different positions of the CIE chart, for equalchanges in the values of the coordinates Since the trichromatic space can be quantitatively described byCIE colorimetric systems, all colors can be produced or reproduced in a display device by mixing three

Introduction to Flat Panel Displays J.-H Lee, D.N Liu and S.-T Wu

c

 2008 John Wiley & Sons, Ltd

P A P E R

yellow ink

green ink white

paper

green light white light

(λR+λG+λB) (λR+λG) (λR+λG+λB) (λ G)

yellow light

white light

white light white light

(b)

Figure 2.1 Formation of colors.

primary emitters Although the reflection spectrum of a real object may be different from that appearing

on the display, the colors look the same to human eyes; this is called ‘metamerism’

In this chapter, we first describe the structure of the human eye and its functionalities This is followed

by the formulation of colorimetry which includes the CIE standards, light sources and photometry Lastly,

we discuss metamerism

2.2 The eye

Figure 2.2(a) shows a schematic diagram of a human eye.3Incoming light passes through the cornea, theaqueous humor, the eye lens and the vitreous humor, and is received by the retina The eye lens, with a

higher refractive index (n = 1.42) than the cornea, the aqueous humor and vitreous humor (n = 1.33 −

1.37), functions to focus a clear image at the retina, as shown in Figures 2.2(b) and (c).4The shape of theeye lens can be adjusted by the ciliary muscle around it Such a system can be approximately described

by the Gaussian lens formula:5

cornea

aqueous

humor

optic nerve ciliary

muscle retina

blind spot

Figure 2.2 (a) Cross-section of the eye 3 (b, c) Formation of images in the human eye.

where d1is the distance from the object to the eye lens, d2is the distance from the eye lens to the retina

(which is 17 mm typically) and f is the focal length The image on the retina is totally reversed (upside

down and right-side left) However, after interpretation by the brain, we can recognize images at theirnormal orientation in real space When the object is farther away, the eye lens becomes flatter, as shown

in Figure 2.2(b) When viewing a closer object, the ciliary muscle will contract the eye lens to increaseits curvature (Figure 2.2(c))

The retina receives the incoming photons and transforms them into bio-potential signals These signalsare then transmitted through the optic nerve to the brain giving rise to the perception of vision The retina

is a multilayer structure which can be divided into three parts: (1) photoreceptor; (2) connecting nervetissue (including outer and inner plexiform and nuclear layers, and ganglion cell layers); and (3) opticnerve.6There are two kinds of photoreceptor cells in the eye, which are called rod and cone cells, namedaccording to their shapes.7The dimensions and quantities of the cone and rod cells are listed in Table 2.1.Rods are more sensitive than cones and, thus, are more easily saturated when the ambient illumination ishigh (e.g indoor lighting) Also, rods can sense light intensity but not colors In contrast, the cone cellsfunction well under brighter ambient conditions and can distinguish colors This explains why peoplecan only see monochrome rather than color images under dark ambient conditions (e.g a moonlessnight)

Figures 2.3(a) and (b) show the spatial distribution of the photoreceptors Solid and dashed linesshow the mean value and one standard deviation away from the mean value of different specimensstudied.8 The symbols show the results obtained from previous reports.9 We can see that the coneshave a maximum distribution near the visual axis of the eye Also, there are almost no rods atthe visual axis, because this area is occupied by the cones The number of rod cells increases andreaches a maximum away from the visual axis As shown in Figure 2.2(a), there is a blind spot inthe human eye where the optic nerve joins the eye and hence there are no cone and rod cells atthis point Typically, in bright ambient the eye is most sensitive within a 10◦ viewing cone Out-side this region, colors are almost indistinguishable At the blind spot we cannot receive opticalsignals However, in actual life, we ‘feel’ that we can see quite a large viewing angle and there is

no blind spot This is because our eyeballs can move and rotate, and because of interpretation by thebrain

Figure 2.4(a) shows the normalized spectral sensitivity of human eyes at ‘scotopic’ and

‘pho-topic’ regions (which mean low- and high-level ambient), in terms of V(λ) and V (λ), respectively.

It is not completely true to regard the scotopic and photopic vision regimes as single contributionsfrom the rod and cone cells, respectively Actually, there is an overlap of the intensity responsivitybetween rod and cone cells When the ambient light is brighter than full moonlight and dimmer thanindoor lighting, both cones and rods can sense light Higher intensity saturates the rod cells, but alower intensity cannot stimulate the cone cells At low and high ambient, the eye is most sensitive

at λ= 507 and 555 nm, respectively Note that there are three different kinds of cone cells, whichhave different spectral responses, as shown in Figure 2.4(b) The S-, M- and L-cones are sensitive toshort, medium and long wavelengths, respectively The bio-potential signals from the photoreceptors(rod cells and three kinds of cone cells) are processed to obtain the signal intensities, as shown inFigure 2.4(a)

Table 2.1 The dimensions and quantities of cone and rod cells.

Cell type Diameter (m) Length (m) Quantity

160(b)

12010080

2 (×1000)6040200

140

2 (×1000)

5101520

25(a)

S-cone

M-coneL-cone

For example, as shown in Figure 2.5, G is larger than its unit vector while R and B are smaller This

means that stimulus Q stimulates green more than red and blue The length of the stimulus vector Q

represents the intensity The greater the length, the higher is the intensity Also, two or more stimuli can

be added linearly to form a new stimulus For example:

R G

B

Q

Figure 2.5 Trichromatic space of (R, G, B) primary colors (redrawn from Wyszecki and Stiles 1 ).

Each point in the triangle represents a distinct color The line linking the blue and red represents the colormixing between these two colors The color inside the triangle is the mixing of these three colors.

2.3.2 CIE 1931 colorimetric observations

Considering the trichromatic color space, it is possible to match an arbitrary color by mixing thered, green and blue primary colors Figure 2.6 shows the setup for color-matching experiments Anarbitrary light of the color under investigation illuminates the lower half of the white screen which pro-duces a stimulus to the human eye through the black shadow A hole in the shadow delimits a smallviewing angle for light passing through, which stimulates only a certain region of the photorecept-ors of the human eye Red, green and blue lights illuminate the upper half of the white screen Theintensities of the red, green and blue lights are adjusted to match the color of the lower light Accord-

ing to the trichromatic color space, it is possible to find a set of R, G and B to fit the color to be

matched

Mathematically, a spectrum can be expressed by superposition of monochromatic units Hence, band light can be viewed as a mixing of many monochromatic components, as shown in Equations(2.3)–(2.5) To obtain the coordinates of each color, the first step is to obtain the RGB stimulus of themonochromatic light In 1931 the CIE used three primary colors with wavelengths of 700, 546.1 and435.8 nm to match all visible monochromatic lights The intensity of the matched monochromatic light

broad-is kept constant in terms of radiometry units (W sr−1m−2), which is called the equal energy condition.This means

(b) (a)

–0.50.00.51.01.52.0

380 475 485 495

505 515 525 535 545 555 565 575 700g

Figure 2.7 (a) Tristimulus values for different wavelengths and (b) CIE 1931 (R, G, B) chromaticity diagram 1

component Hence, any color can be represented by the horseshoe-shaped region, formed by the chromatic light locus and the connection between 380 and 780 nm Note that in this color system, the

mono-r(λ) values are negative when wavelengths are between 435.8 and 546.1 nm, as shown in Figure 2.7(a).

Sometimes, in color-matching experiments as shown in Figure 2.6, one cannot obtain the matched color

no matter how one adjusts the (R, G, B) intensities However, it is possible to move one of the ies, for example the red one, from the opposite to the same side as the color to be matched This can

primar-be expressed as

So, in Figure 2.7(b), we have to use negative r values when describing some colors, which is not entirely

satisfactory To improve on this, the CIE 1931 (X, Y, Z) system was proposed Linear transformationsfrom the CIE 1931 (R, G, B) system are carried out using the following equations:

One important feature of the CIE 1931 (X, Y, Z) color system is that the Y value is set as the luminance

of the stimulus, in terms of lm sr−1m−2or cd m−2(Section 2.3.6):

y V , Y = V, Z = z

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.80.0

0.10.20.30.40.50.60.70.80.9

green

600 580 560

yellowish

- green

yellow green yellow orange raddish-orange

red publish - red publish-pink pink

puple publish

- puple

publish - green grrenish

- blue blue

Figure 2.8 CIE 1931 (X, Y, Z) chromaticity diagram.

where V is the luminance of the stimulus Then, it is straightforward to understand that

where k= 683 lm W−1, which represents the transformation from radiometry units (W) to photometry

units (lm), and P(λ) is the spectral distribution of the stimulus in terms of W sr−1m−2

Answer From Equations (2.23) and (2.24), X/x = Y/y = Z/z = X + Y + Z:

2.3.3 CIE 1976 uniform color system

Although the CIE 1931 (X, Y, Z) color system can describe a color exactly, there is a problem when dealingwith color difference and tolerance Figure 2.9 shows the famous MacAdam ellipses in the CIE 1931(X, Y, Z) chromaticity diagram.11Color differences cannot be discerned by the human eye within theellipses in this figure Note that the ellipses are magnified 10 times for clarity, so the ellipses are verysmall We can see that ellipses in the blue color region are much smaller than those in the green andred ones, which means that a small shift in color coordinates of the CIE 1931 (X, Y, Z) color system

in the blue region results in a serious difference perceived by the human eye In other words, when

the difference of the color coordinate, i.e (x, y), between two color stimuli is kept constant, the

differences of perception by the human eye are largest in the blue region To better illustrate the ‘colordifference’ between two stimuli, e.g a real object and an image from a display, it is necessary to have auniform color system

0.00.10.20.30.40.50.60.70.8

x

Figure 2.9 MacAdam ellipses in the CIE 1931 (X, Y, Z) chromaticity diagram 1

0.0 0.1 0.2 0.3 0.4 0.5 0.60.0

0.10.20.30.40.50.6

u'

470nm 480nm 490nm

Figure 2.10 CIE 1976 (u, v) chromaticity diagram and MacAdam ellipses.

One such system is the CIE 1976 (L∗u∗v∗) color system (Figure 2.10) In this color space, we cansee the sizes of the ellipses vary less in different regions of the color space than those in the CIE 1931(X, Y, Z) color system, although the area occupied by an ellipse in terms of color discrimination is exactlythe same in both systems Equations (2.25)–(2.32) govern the coordinate transform from the CIE 1931(X, Y, Z) to the 1976 (L∗u∗v∗) color system; since this is a linear transformation, equations for colormixing are still valid in this color space with some modifications:

L∗= 116



Y Yn

color without a specified light source, X , Y , and Z values are set to zero

Compared to the CIE 1931 (X, Y, Z) color system, not only does the color difference become uniform,

the brightness difference becomes uniform by nonlinear transformation between the Y and L∗ values

in the CIE 1931 and 1976 color systems, respectively Luminance (or Y value, in terms of cd m−2) isproportional to the optical intensity in terms of radiometric units (W sr−1m−2) However, due to the

nonlinear response of the human eye, it is not appropriate to use the Y value to describe the intensity difference between two stimuli of the same color but different intensities The L value is more uniform in photometric terms than the Y value As is evident from Equation (2.25), L is proportional to Y to the power

of 3, which means that a certain intensity difference distinguished by the human eye requires larger andsmaller physical intensity (i.e watts) changes under brighter and darker stimuli, respectively In otherwords, the human eye can distinguish a smaller change when the luminance is low This means that,

when the stimulus is weak, only a small difference in luminance value (Y ) can result in a large difference

of perception by the eye (L∗) However, when the stimulus is bright, a larger difference in luminance isrequired to obtain the same difference of perception by the eye In a display, we use the term ‘grayscale’

to determine the brightness difference The differences of perception by the eye between two adjacentgray levels are the same Figures 2.11(a) and (b) show the grayscale levels versus the luminance using

linear and logarithmic scales, respectively The slope of the plot in Figure 2.11(b) is called the γ value,

which is 2.157 in this case, i.e greater than 1 and less than 3, as shown in Equation (2.25) For example,

a display has grayscales from 0 (black) to 255 (white) Grayscales 1 to 254 represent different ‘grays’from dark to light The luminance difference between grayscale 0 and 1 is smaller than that between 254

and 255 since the γ value is not equal to 1.

2.3.4 Color saturation and color gamut

As shown in Figures 2.8 and 2.10, the boundary of the horseshoe shape of the chromaticity diagram isformed by the monochromatic line and the purple line, the connection between the shortest and longestwavelengths Any spectrum can be divided into monochromatic components It is straightforward tounderstand that the color coordinates of a stimulus with a broader bandwidth lie nearer to the boundary

of the chromaticity diagram The center of the 1931 (X, Y, Z) chromaticity diagram, i.e (0.33, 0.33), showsthe equal-energy spectrum with a white ‘color’ Color saturation is used to describe the ‘colorfulness’ of astimulus Color saturation increases towards the boundary of the chromaticity diagram A monochromaticstimulus exhibits the highest color saturation When mixing monochromatic light, the color saturationdecreases The white stimulus can be regarded as the ‘achromatic’ color

In a display, three primary colors are used to generate all color images, based on the trichromatic colorspace theory By mixing the three primaries, the colors within the triangle shown in Figure 2.12 can

10100