Abstract
Nishida et al. (2007) showed that when red and green bars alternate along an apparent motion trajectory, a single moving yellow bar is often perceived. They suggest this effect could be the visual system's attempt to integrate colors belonging to the same moving object. This conclusion predicts that other common bar features, e.g., luminance contrast, should contribute to the robustness of mixing, and that therefore less mixing should be observed if the stimulus is isoluminant relative to the background compared to if common luminance contrast is present. To test this we used a stimulus composed of spatiotemporally alternating red and green annular sectors that were presented circularly around the central fixation point. The stimulus was presented on a mid-grey background to enable isoluminant stimuli to be used. Perceived color mixing was measured using a single interval procedure, in which observers reported on each trial if the hues were perceived as mixed or as separate reds and greens. The independent variables were the angular subtense of the sectors, their presentation duration and the amount of added luminance contrast. Results indicated that the ranges of angular subtense and presentation duration over which perceived color mixing occurred decreased rather than increased with luminance contrast, by a factor of about 3.0 and 2.6 respectively. A control experiment measured discrimination thresholds for mixed versus non-mixed static red-green sectors, and revealed that the color mixing was not simply due to the chromatic system's reduced spatial acuity. These results contradict the object commonality hypothesis and points towards a lower level process in which luminance contrast suppresses spatiotemporal color blurring, which in turn facilitates the color mixing.
Meeting abstract presented at VSS 2014