Abstract
Natural visual scenes typically contain both luminance and chromatic contrast, but spatiotemporal resolution is notably better for luminance variations than for color variations. This raises the question of how the presence and analysis of color differences might be used by the visual system. Do color differences in the presence of luminance differences help or hinder spatiotemporal discriminations? We have asked how changing color over time affects one's ability to discriminate the direction of moving luminance patterns.
Luminance motion direction can be correctly determined at contrasts near detection threshold when no color contrast is present. However, if the color changes over time (even though each frame is isochromatic and the pattern is defined solely by luminance variations), the luminance contrast required to discriminate the direction of motion may increase significantly (motion interference). The magnitude of the motion interference is proportional to the temporal color contrast in the LM-opponent system, but color changes restricted to the S-opponent system do not interfere with motion direction discrimination at all. Thus, the motion interference is not simply related to the salience of the color differences or to the overall magnitude of color contrast. Our data suggest, rather, that the parvocellular system is involved in the analysis of first-order luminance motion, but the koniocellular system is not. Control studies imply that it is the color variation that is responsible for these effects, not luminance artifacts. Our results suggest a model of luminance motion analysis that incorporates chromatic selectivity. This is consistent with models that postulate a parvocellular input into first-order motion mechanisms and suggests a possible role of chromatic selectivity in object segregation by motion