Abstract
Purpose :
Color appearance arises from a complex neural interpretation of the pattern of quantal absorptions spatially distributed across the cone mosaic. Understanding how the appearance of a small colored stimulus in one region of retina is affected by proximal stimuli helps identify the spatial grain of the underlying mechanisms. Precisely characterizing such effects has been technically challenging, especially due to the influence of monochromatic and chromatic aberrations. Here, we investigated the 2-line Rayleigh limits for perceiving hue and for spatial resolution under adaptive optics correction.
Methods :
Stimuli were displayed using custom LED illumination (centered at 526nm and 661nm) and a DMD in an AO vision simulator with a 900nm wavefront sensing beacon. Subjects’ (n=2) pupils were dilated and accommodation was paralyzed while viewing the display through a 6mm artificial pupil. Longitudinal and transverse chromatic aberrations were compensated using a modified Badal optometer and a Vernier alignment task, respectively. The psychophysical procedure involved discrimination of pairs of closely spaced vertical lines (height: 14min, width: 0.3min). The lines varied in horizontal separation (0 to 3min) and were colored in one of three ways: red/green, green/red, or yellow/yellow. In each trial, subjects indicated both the perceived colors and whether the lines were abutting or separated.
Results :
Subjects rarely switched the colors of the red/green lines, nor did they make red/green false alarms for yellow pairs, confirming the fidelity of the TCA correction. At large separations, subjects had no difficulty correctly identifying the color and spatial separation. However, at closer separations perceptually distinct colored lines remarkably "lost" their hues. Well described by sigmoidal functions, spatial gap detection reached asymptotic performance at 1.2-1.5min, while identifying the chromatic content reached asymptote at approximately 2.25min for each subject.
Conclusions :
The region of integration for spatial vision versus chromatic vision differed: the color appearance of a thin line was impacted by a line that was more than 1.5X as distant (~2.25 min) as one that could reliably be identified as spatially distinct (1.2-1.5 min). The inability to properly disambiguate chromatic information from nearby patterns reveals a fundamental bottleneck of the parvocellular pathway, dissociable from the finer mechanisms limiting spatial resolution.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.