Observers differ in the stimuli they choose for unique hues (i.e. pure red, green, blue, and yellow) or binary hues (e.g. orange, purple, yellow-green, and blue-green). Yet there is little correlation between these choices, suggesting that different color categories are influenced by largely independent factors (Malkoc et al., JOV 2002). We examined variations in color appearance both within and between color categories, by sampling finer gradations of color with a hue scaling task. Stimuli were 24 equiluminant hues of moderate contrast spaced in 15-deg steps along a circle within the LM and S chromatic plane, and were shown in a 2-deg field centered in a gray background on a monitor. 59 subjects rated the hue of each stimulus by pressing buttons to indicate the perceived proportion of red, green, blue, or yellow. In a separate task they also chose a unique or binary label for each stimulus. Settings for nearby stimuli were strongly positively correlated while more independent for distant angles. These local correlations were centered near the poles of the LM and S cone-opponent axes, as revealed by a factor analysis of the settings, but are not predicted by scaling differences in the cone-opponent or unique-hue axes. The lack of a global influence on hue scaling confirms the relative independence of different color categories, and suggests that variations in both unique and binary hues may reflect qualitatively different clusters that cannot be related by a simple metrical color space.

Supported by EY-10834

Malkoc, G., Kay, P., & Webster, M.A. (2002). Individual differences in color categories. Journal of Vision, 2(7), 144.