The contribution of color to motion in normal and color-deficient observers

By opposing drifting luminance and color gratings, we have measured the "equivalent luminance contrast" of color, the contribution that color makes to motion. We found that this equivalent contrast was highest (greater than 10%) for low spatial and temporal frequencies and was higher for red/green than for blue/yellow stimuli. Equivalent luminance contrast was about 4% for a green/purple stimulus that fell along the tritan confusion line, indicating a modest input to the motion pathway from the short wavelength-sensitive cones (B-cones). Contrast thresholds for the discrimination of the direction of motion showed that the contribution of color to motion was about the same (within a factor of 2) as that for luminance in terms of multiples of threshold contrast. These responses to moving, chromatic gratings could be mediated by any of several factors that can create a residual response in a luminance pathway: temporal phase lag between the responses to the colors of the stimuli, second harmonic distortion in the response and variability in equiluminance points across units. Each of these factors was evaluated experimentally and their combined effect could account for only a small portion of the contribution of color to motion. As a result, we attribute the perception of the motion of equiluminous stimuli to an opponent-color input to directionally selective cortical units. Chromatic stimuli had little or no equivalent contrast for color-deficient observers, whether the stimulus was red/green, which they discriminate less well than normals, or blue/yellow, which they discriminate almost as well as normals. The equivalent contrast measure provided an excellent basis for classifying normal, protan and deutan observers.