Abstract
Purpose: Recently, we showed that colour illusions in half-occlusions result from a feedback mechanism that reduces binocular colour differences and affects the colour appearances of neighbouring monocular objects. Here we examine whether this mechanism can fully explain the phenomenon of binocular colour mixing. Methods: Stereograms were displayed on a colour CRT monitor. Each eye viewed a set of four uniformly coloured discs (2 deg dia) configured in the shape of an upright cross. The horizontally aligned discs of one image and the vertically aligned discs of the other were the same colour. The vertical offset between the two images was such that in a stereoscope the top disc of one image was binocularly fused with the bottom disc of the other image so that in binocular viewing the binocular disc was surrounded by six monocular discs. The discs were viewed against black and white backgrounds. The subjects matched the colours of the top and bottom (monocular) discs to that of the central (binocular) disc. Results: The colour matches expressed in x,y-chromaticity coordinates show that the settings for discs within the same stimuli were different from both their initial values and from each other. Distances between the two settings were larger when the discs were viewed against the black background. Conclusions: Binocular colour processing based on a feedback mechanism predicts that the settings would have been identical to the initial colours of the monocular discs. Binocular colour processing based on a feed-forward mechanism, i.e. binocular colour mixing, predicts identical settings for the two monocular discs. Settings in between these two extremes suggest that binocular colour processing is based on both mechanisms.