Abstract
The visual system can reliably extract the average value of stimulus distributions that vary along many dimensions, from orientation to facial expressions. We examined how sensitive observers were to the average chromaticity of color distributions. This average has special significance for color coding (e.g. to estimate the illuminant) but how it might depend on the level of representation (e.g. perceptual vs. cone-opponent) or nonlinearities (e.g. categorical coding) is unknown. Stimuli were 11 by 11 arrays of 0.75-deg spots that alternated between 2 component colors differing in hue or contrast. In one task, observers set the mean to a balanced blue-green by adjusting the hue angle of bluish and greenish components. These settings were as reliable as blue-green boundaries in uniform color arrays, even when the components differed by up to 80 deg in the color space (~2.5 fold difference in S cone excitation). However, the two-color mean was biased toward blue, consistent with a compressive response to S cone signals. In a second task, observers instead adjusted the hue angle or contrast of one component until the two hues appeared opposite or complementary (a mean of gray). The chosen complementary hues did not clearly favor a specific (e.g. red vs. green or +LM vs. -LM) representation, and instead were characterized by high variability, far greater than for saturation matches between actual complements. These results for larger color differences suggest that color may not be perceptually represented in a space that assigns metrical relations among different color percepts.
Meeting abstract presented at VSS 2013