Color Appearance Models, by Mark D. Fairchild, Addi-books that are both authoritative and up-to-date. And even amid the dustiest of material (e.g., in the colorimetry son- Wesley, 1998, 448 pp., $96.79 hardback chapter), there is some new spin. For example, of the 1931 and 1964 CIE standard observers, Fairchild says, The modeling of color appearance is the most important frontier in color science, because of the increasing de-''The differences are computationally significant, but certainly within the variability of color-matching functions mands for a standard way to transport colors among devices on the Internet, and to achieve faithful rendition of found for either 2ะ or 10ะ fields'' (p. 90).
Chapter 4, on color-appearance terminology, is also colors in the new digital television. But modeling color appearance is an arcane subject, prey to ambiguities of lucid, and disambiguates the terms brightness, lightness, colorfulness, chroma, and saturation. (Hue, I think, was terminology, homegrown answers to general problems, and a pervasive tendency of some to dismiss the entire already unambiguous.) Where similar terms can be confused, Fairchild uses simple equations to draw distinc-field as not worth the effort. It is, therefore, important now for practitioners of color management to have in tions.
There follows a surprisingly complete summary (Chap-hand a monograph that summarizes the needs, approaches, and benefits in a rather diverse field. Such a ter 5) on color-order systems, and then an effectively illustrated catalogue of color-appearance phenomena milestone book is Color Appearance Models, by Mark D. Fairchild.
(Chapter 6). Chapter 7 on viewing conditions sets the stage for an introduction to chromatic adaptation (Chapter Perhaps the greatest tribute I can pay to this book is confessing the regret I felt when I thought I had lost my 8). It is here that we first see (in Figure 8-7) the structure of a model (in this case, predicting appearance matches reviewer's copy. With mounting sadness, I recalled what I had lost. Nowhere else is catalogued so completely or across adaptation states). The structure is quickly filled lucidly the color-appearance models of the past 40 years:
in Chapter 9 with several chromatic adaptation models CIELAB, CIELUV, Nayatani's model, Hunt's model, (von Kries, Retinex, Nayatani, Guth, and Fairchild), and Guth's ATD model, LLAB, RLAB, ZLAB, and CIEthen the scene changes to the more general notion of CAM97 (s and c, replete with a Web address to consult a color-appearance model (Chapter 10), in which the for updates). Nowhere else are models addressed so evencoordinates have absolute perceptual significance as well handedly: in the middle of the book is a catalogue of as color-matching status. Fairchild uses CIELAB and models, each one (call it ''X'') allocated sections on CIELUV as simple examples of such models, the former objectives and approach, input data, adaptation model, connected with the Von Kries chromatic adaptation via opponent-color dimensions, perceptual dimensions, ina ''wrong von Kries transform'' (i.e., in the basis X, Y, verse model, phenomena predicted, and (most provoca-Z rather than in the basis of the receptors L, M, S). [I tively) ''Why not just use Model X?'' Nowhere else are wish Fairchild had also traced out the connection between such models compared so precisely with such different CIELUV and Judd's 1940 model 1 -both of which effect formulations as color-order systems and chromatic-adapadaptation by subtracting the white chromaticity. Perhaps, tation theories. How could I function professionally in the in the future, an historical connection as well as a formal next few months without my copy of Fairchild's book? one can be made in this regard.] Fortunately, the book turned up within 24 hours, and I Chapters 11-14 catalogue the models mentioned in the decided to keep better track of it.
second paragraph of this review, and Chapter 15 deals The book begins with a motivational introduction, with with tests of the models (presented pictorially and not a list of visual effects whose explanation is not obvious just through data figures). Chapter 16 introduces the apwithout a model of color appearance. Fairchild offers plications of color-appearance models in fields outside short explanations, or ''clues,'' for these effects, one by color management (i.e., to assess color rendering, color one. These explanations are self-contained, and give an differences, and metamerism). idea of what a color-appearance model is supposed to do.
Then follows a true highlight of the book. Chapter 17 is Next, the book launches into an accelerated (but readily a masterpiece in describing how to use a color-appearance understandable) set of three chapters on human color vimodel in a color-management context (a problem of sasion, psychophysics, and colorimetry. Where detail neceslient importance). First, the chapter discusses the use of sarily must be left out, Fairchild refers to reviews and such a model in device-independent color reproductionwhich requires that the model be run forward to a deviceindependent perceptual space, and then inversely to new