EFFECT OF SPATIAL SAMPLING ON GRATING RESOLUTION AND LETTER IDENTIFICATION

The purpose of the study was to determine whether spatial sampling has equivalent effects on visual performance for grating and letter stimuli, two optotypes that are frequently used in the clinical assessment of visual function. Test targets consisted of five-bar square-wave gratings, spatial D6 (sixth derivative of a Gaussian) patterns, or individual Sloan letters, presented for 255 ms in the center of an adapting field. Spatial sampling was introduced by replacing random regions of the targets with occluding elements that had the same luminance as the adapting field, so that only random samples of the test target were visible on each trial. The occluding elements consisted of 4 x 4, 8 x 8, or 16 x 16 pixel arrays (subtending 0.67, 1.33, and 2.67 arcmin, respectively), and the percentage of the test target that remained unoccluded ranged from 3% to 100%. Visual acuity and contrast sensitivity for the spatially sampled targets were measured in two visually normal subjects. Results were compared with the predicted effect of the reduction in effective contrast that was introduced by the sampling paradigm. For grating stimuli the effect of spatial sampling was equivalent to the reduction in effective stimulus contrast produced by the sampling, regardless of occluding-element size. For letter stimuli the reduction in effective stimulus contrast predicted most, but not all, of the loss of visual acuity and contrast sensitivity for sampled targets, and the deficit in performance increased when larger occluding elements were used. These results indicate that spatial sampling does not have comparable effects on visual acuity and contrast sensitivity for grating stimuli and letters and therefore these stimuli do not provide equivalent measures of visual function.