Purchase this article with an account.
James M. Hillis, David H. Brainard; Color detection and appearance: A non-linear link. Journal of Vision 2004;4(8):57. doi: 10.1167/4.8.57.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Color vision is useful for detecting surface boundaries and identifying objects. Are the signals used to perform these two functions processed by common mechanisms, or has the visual system optimized its processing separately for each task? We measured the effect of adaptation on both color discriminability and appearance under well-matched conditions. In the discrimination experiments, a pedestal spot was presented in one interval and a pedestal + test spot in a second. Observers indicated which interval contained the test. In the appearance experiments, observers matched the appearance of test spots across a change in background. We studied i) the effect of mean chromaticity and luminance of spatially uniform backgrounds and ii) the effect of introducing temporally modulated chromatic contrast. Discrimination thresholds measured over a range of pedestal contrasts determine the response function of the visual system for each adapting condition. The response functions were non-linear, saturating at high contrasts. Changes in the response function across the uniform background conditions can be accounted for by a combination gain control and subtractive adaptation. Such a model cannot account for the effect of contrast adaptation. Contrast adaptation induces a broadening of the response function, which extends the visual system's operating range. The response functions derived from the discrimination experiment allow prediction of the results of the matching experiment. The linking hypothesis is that two test spots in different contexts will match when they produce the same visual response. This prediction gave an excellent account of our matching data, indicating that the same mechanisms of adaptation regulate information processing for both discrimination and appearance. This result validates the use of threshold tasks as a method for studying the effect of context on color appearance, but only when the non-linear nature of the response function is characterized.
This PDF is available to Subscribers Only