Retinal inhomogeneity. II. Spatial summation.

Effects of spatial probability summation were measured with concentric cosine forming a set of contiguous annular zones having mean eccentricities of 0 degrees, 2.8 degrees, and 6.1 degrees. To detect these small effects reliably we measured contrast thresholds for all eccentricities at one spatial frequency in each experimental session by an interleaved-staircase method. Sharp edges were eliminated by truncating each zone at a zero crossing of the pattern, and fixation was controlled by stabilizing the retinal image. Under these conditions, each local region of the retina contributes to the sensitivity of larger regions by a standard, fourth-power summation rule, regardless of the number or contiguity of the component regions involved. This was experimentally confirmed with three nested zones for spatial frequencies from 0.5 to 5 cycles/degree. Together with the results of Part I of this series [D. H. Kelly, J. Opt. Soc. Am. A 1, 107 (1984)], the summation rule predicts that, over a large range, target size should have no effect on contrast sensitivity at high spatial frequencies, and this was also confirmed. Similar predictions were extrapolated to other contrast-sensitivity data for which empirical tests are not available.