Daniel R Coates¹ (), Susana TL Chung²; ¹University of Houston College of Optometry, ²UC Berkeley

Precisely determining anisotropies and eccentricity-dependencies remains a crucial method to characterize phenomena in spatial vision. Visual crowding yields numerous such quantities, such as the change in critical spacing (CS) vs. eccentricity, and differing effects of radial and tangential flanker arrangements. Six observers identified the orientation of a Tumbling E in the fovea and in the right visual field at 2, 5, 10, 20, and 30 degrees from the fovea. Targets were presented alone or flanked by two radial or tangential flankers in a variety of sizes and spacings controlled by an adaptive psychophysical procedure. The resultant threshold size vs. flanker spacing curves provided measures of the critical spacing as well as unflanked acuity. As expected, both flanker arrangements yielded monotonically increasing CS (radial = ~0.2-0.3 X eccentricity, tangential = ~0.125-0.16 X eccentricity), but neither curve had a constant change vs. eccentricity. The slope of the radial CS near the fovea was 0.175 degrees of visual angle (dva) for each degree of eccentricity, peaked between 5&10 degrees at approximately 0.3 dva/deg eccentricity. before decreasing to ~0.2 dva/deg at 30 degrees. The slope of the tangential CS near the fovea was 0.1 dva/deg, growing monotonically to ~0.15 dva/deg at 30 degrees. The consequence of these differing rates is a non-constant, non-monotonic radial vs. tangential anisotropy of ~1 near the fovea that grows to a maximum of ~2.2 at 5-10 degrees, and decreases back down to ~1.5 at 30 degrees. These findings help constrain theories of the substrate for crowding, whether governed by cortical receptive fields, anatomical architecture, or other explanations.