Abstract
Purpose:
We investigated the impact of the cone mosaic on spatiochromatic reconstruction of two-point stimuli by measuring detection and appearance with stimuli made smaller than retinal cones with adaptive optics. We sought to determine whether detection and perception of these points are independent near the fovea, as has been previously hypothesized, and if not the extent of summation for detection and color. We also investigated whether two-point resolution is consistent with a photoreceptor limit under these conditions.
Methods:
We performed two experiments in which subjects viewed 550nm two-point stimuli through a 6mm pupil with adaptive optics aberration correction. Stimuli were viewed at ~1 degree eccentricity under a rod bleaching protocol to isolate cone responses. In one experiment two subjects were shown either one or two spots (25 ms, 0.5 arcmin estimated retinal full-width at half maximum (fwhm)) of different intensities and reported the number seen and color appearance. Separation was varied from 1.5 to 15 arcmin. In the second experiment two subjects were shown either one or two dim, yet suprathrehsold, spots (15 ms, 0.5 and 1.1 arcmin estimated fwhm) and reported whether they believed one or two spots were present. Separation was varied from 0.5 to 12 arcmin.
Results:
Detection thresholds in the first experiment were consistent with complete spatial summation up to 3-4 arcmin. Detection, but not color appearance, was independent for separations of 8 arcmin or more and consistent with at least 15 arcmin of spatial uncertainty. For two-point stimuli color appearance was highly variable, as for single points, but also highly coupled. Both points were reported to have the same color appearance on over 97% of trials, regardless of separation. Resolution thresholds were similar in both experiments, 2.5-5 arcmin, with a slight tendency towards lower thresholds with the larger spots.
Conclusions:
Cone detection at 1 degree is consistent with complete spatial summation over an extent including at least a cone and its nearest neighbors, creating challenges in recovering ‘single cone’ sensations from the responses to tiny stimuli. Despite this extent, two-point resolution is roughly consistent with expectations based on a photoreceptor limit. The large variation, but lack of independence, in the color appearance of the two-point stimuli, implies nonlocality in the color reconstruction mechanisms.
Keywords: 496 detection •
471 color vision •
649 photoreceptors: visual performance