Item

Abstract

Watson, Barlow and Robson (1983) argued that the visual stimulus which humans detect best specifies the spatio-temporal structure of the receptive field of the most sensitive visual neurons in the human brain. To investigate 'what the eye sees best' they used various achromatic stimuli (squares, spots and gratings) and found that drifting grating patches were seen best by their observers.
A decade later Chaparro, Stromeyer, Huang, Kronauer and Eskew (1993) reexamined this idea using flashed circular spots whose spectral content could be changed. Their best chromatic stimulus was seen 5 - 9 fold better than their best luminance spot and 3-8-fold better than Watson's best stimulus.
Here we report experiments with a rectangular stimulus containing two vertically oriented edges of 20 min of arc height. Chromaticity of the edges was chosen from a set of 16 equally detectable colors which covered the whole gamut of color available on a video monitor. For each observer, these colors were adjusted to be equiluminant with respect to the grey background using flicker fusion photometry. Observers fixated a spot in the center of the stimulus such that the two vertical edges were seen at about 1 deg. eccentricity to the left and to the right. After initial fixation of 0.5 sec, the stimulus was presented in one of the 16 colors. After .5 s., at random, one of the vertical edges (left or right) jumped to the left or right and stayed there for another .5 s. Subjects were instructed to report (1) on which side relative to the fixation spot the edge had moved, irrespective of its direction and (2) its direction of motion.
For all observers 16 colors were found, for which they correctly reported the direction of motion of the edge, while failing to reliably detect the location of motion irrespective of its direction. Quantitatively, we found that contrast thresholds for identification of the direction of motion were about half of those for detecting the side at which motion was presented or for flashed stimuli of the same chromatic content.
As a control we also ran experiments where achromatic edges of various contrasts were presented on the same grey background to the same observers. Now all subjects performed equally well in both tasks.
Thus, the high sensitivity for the direction of motion of an edge found in our experiments is due to both the chromatic contents of the stimuli, as well as the spatio-temporal interactions extracting the direction of motion. Comparable results of experiments where the moving edge was presented to the central fovea were described by Kremers, Wachtler, Wehrhahn, Lee Zrenner (1993). We conclude that the stimulus that the eye sees (so far) best, is the direction of a chromatic moving edge.