Abstract

The variability normal human subjects exhibit in repeatedly adjusting the position of one spatially discrete visual stimulus so as to place it in apparently perfect horizontal or vertical alignment with another is referred to as their vernier acuity. The magnitude of this variability corresponds to a visual angle of about one-fourth that subtended by the eye’s smallest photoreceptors. The mean error of a subject’s vernier alignment settings reflects any directional bias of his judgements of relative stimulus location and this may vary independently of one’s vernier acuity (see Baker & Bryan, 1912; Guild, 1930; and Fendick, 1984). Vernier acuity is a member of the class of human visual capabilities referred to as the hyperacuities (see Westheimer, 1981 for a review). Like other hyperacuities, it is remarkably robust to various stimulus parameters, including luminance (Baker, 1949 and Leibowitz, 1955), duration (Westheimer & McKee, 1977), and blur (Stigmar, 1971). This, as well the obvious fact that such fine spatial discrimination must necessarily require post-receptor neural processing of one form or another makes vernier acuity a superb candidate for clinically assessing spatial visual function. Although normal subjects viewing nearly abutting stimulus elements demonstrate directional biases that are small relative to their vernier acuity, the same is not true for patients exhibiting clinical disorders in which the retina is stretched or otherwise distorted (Fendick, unpublished data). Alignment biases in such cases are consistent with perceptual spatial distortions evidenced by the Amsler Grid test.

© 1985 Optical Society of America