Abstract
We have investigated the properties of binocular colour vision mechanisms using a new measure; the dichoptic colour difference threshold (DCDT). The DCDT is the minimal detectable difference in colour between two dichoptically superimposed stimuli. We examined DCDTs in order to determine the dependency on a) the average colour and b) the colour contrast of the stimuli. The stimuli consisted of two, vertically separated, 4.57-deg diameter, equiluminant, circular colour patches. Both were stereo-pairs presented via a modified Wheatstone stereoscope. In one of the stereo-pairs, there was a between-eye difference in colour whereas the colours of the other stereo-pair were same for both eyes. Subjects had to decide which of the two stimuli was different in colour. Psychometric functions were fitted to the data and used to calculate thresholds for DCDTs. We found that DCDTs were higher than monocular colour difference thresholds and lower than binocular rivalry thresholds (dichoptically superimposed colours seen in alternation). DCDTs were found to be between 10 and 20 deg for test colours, and there were no apparent minima at points intermediated between either of cardinal colors or unique hues. DCDTs were significantly and positively correlated with perceived colour difference suggesting that perceived color difference plays a significant factor for DCDTs. Thus, in contrast to previous findings, our DCDTs data cannot be explained in terms of a limited number of mechanisms sensitive to either the cardinal colours or unique hues, suggesting that multiple binocular colour vision mechanisms exist at the post-receptoral level. DCDTs were constant as a function of chromatic contrast (except at very low contrasts) when measured in terms of distance in colour space.
Supported by Canadian Institute of Health Research grant #MOP-11554 awarded to FK.