Color appearance under chromatic adaptation varied along theoretically significant axes in color space.

Changes in color appearance caused by chromatic adaptation were measured with a wide range of adapting fields. Observers viewed a 39'-55' annular test field composed of an admixture of lights from the red phosphor and the green phosphor of a CRT. The annular mixture field was centered and superimposed upon a 4.7 degrees steady, circular background field. After the observer was completely adapted to the background, the luminance of the red phosphor in the test was held fixed while the observer adjusted the luminance of the green phosphor until the test appeared neither reddish nor greenish. Twenty-two equiluminant backgrounds (4.5 cd/m2, approximately 50 Td) were systematically selected along two axes in Judd chromaticity space. One axis was along tritanopic confusion lines, with middle-wavelength-sensitive- (M-) and long-wavelength-sensitive- (L-) cone stimulation held constant. The other axis maintained constant short-wavelength-sensitive- (S-) cone stimulation. The results show that adapting backgrounds that were varied along tritanopic confusion lines do not have a differential effect on color appearance at high test levels (well above the adapting level). At lower test levels there is a systematic change in color appearance of the test light, which is quantitatively described by additive redness. Along constant S-cone-stimulation lines, adapting backgrounds differentially affect color appearance in a systematic way, reflecting changes in receptoral gain and the additive contribution. The measurements taken with adapting fields throughout color space are described by the two-process model of chromatic adaptation.