Tracing a Metameric Match to Individual Variations of Colour Vision

Abstract

A metameric match may be accepted by one observer but rejected by another, indicating that the colour vision of the two observers is different. We analysed the variations of normal colour vision using metameric surfaces. First, we modelled the matches of a theoretical colour-normal observer on the Davidson and Hemmendinger (D&H) colour rule (Kaiser and Hemmendinger, 1980 Color Research and Applications5 65 – 71), using spectroradiometric measurements and a set of fundamentals (Stockman, Macleod, and Johnson, 1993 Journal of the Optical Society of America A10 2491 – 2521). We also derived deviate fundamentals by changing the macular pigment density, the lens density, the photopigment density, and by shifting the long-wave sensitive photopigment. Modelling the deviate normal observer matches yields shifts of no more than 1 sample on the D&H colour rule. The largest shifts are produced by changes in lens density. Second, we asked six observers to perform a match on the D&H colour rule. We also estimated their macular pigment densities and their lens densities by using heterochromatic flicker photometry matches at 466 nm and 413 nm versus a 558 nm reference, and we recorded their Nagel anomaloscope setting. Then, we computed their personalised fundamentals in order to predict their D&H colour matches. As the most frequent match in a sample of young observers is F-15 and the modelling for the theoretical observer yields G-15, a systematic error occurs in the predictions. After correction for this error, the results show that the metameric match of five out of six observers is better predicted by using the personalised correction of the lens and macular pigment optical density only, than by using the theoretical colour observer data.