Crowding under scotopic conditions.

Under certain circumstances, a subject's ability to discriminate spatial features of a target may be hampered by neighbouring contours. This phenomenon is popularly known as the "crowding effect", and it has been intensely studied for photopic vision: little attention has been paid to the effect at lower light levels. The underlying basis of the crowding effect has recently provoked some conjecture, with Hess and colleagues claiming that a passive "physical" phenomenon may either wholly [Vis. Res. 40 (2000) 365], or partially [J. Opt. Soc. Am. A--Opt. Image Sci. Vis. 17 (2000) 1516], account for the effect. In order to investigate the crowding effect under scotopic conditions, we conducted scotopic frequency of seeing experiments for Landolt C targets presented both with, and without, flanking bars; the size of the targets was varied so that frequency of seeing curves could be derived for each stimulus condition. Our results suggest that the spatial extent of crowding is significantly less for scotopic vision than for photopic vision at the same eccentricity--furthermore the effect does not seem to scale in proportion to target size. We also compared the resulting empirical curves to those that would be predicted by the hypothesis of Hess and colleagues. Our results do not support the hypothesis that the scotopic crowding effect is caused by a passive physical process.