| Abstract: | Vos demonstrated that the Judd 1951 color-mixture diagram is a projective transformation of the CIE 1931 mixture diagram. He has provided transformation formulas for computing values of x′ and y′ from x and y, but he does not explain how the one set of data is derived from the other. Neither does Judd. Judd wanted the CIE to replace the 1931 V(Λ) curve with a new V(Λ) curve and to replace the 1931 XYZ diagram with a new X′Y′Z′ diagram. I have assumed that Judd started with the CIE 1931 RGB diagram. His X′Y′Z′ diagram can be derived from the RGB diagram if we use a different set of luminosity coefficients than those specified by the CIE. This change in luminosity coefficients was no doubt intended to compensate for the use of a new V(Λ) curve, but Judd has ignored the effect that this change in the luminosity coefficients should have on the alychne of the RGB diagram and the locus of spectral colors. This makes Judd's X′Y′Z′ diagram inappropriate for use in colorimetry and for use in building color-vision models. I have proposed a new solution to this problem. |
