Spatial frequency thresholds versus border enhancement: sensitivity to retinal defocus.

Abstract

Two psychophysical methods for monitoring retinal defocus are compared. One methods, spatial frequency thresholds with sine wave grids, is well known. The second methods is relatively new: it uses the effect of defocus on the width of the perceptually enhanced region of brightness or darkness generated next to a luminance edge. Both methods produce characteristic functions as the retinal image is thrown out of focus. However, spatial frequency thresholds appear relatively unchanged until a certain level of defocus has been introduced, whereas border enhancement spread is affected even by very small deviations from focus. It is suggested that variations in the frequency threshold are dampened in the near-focus range by aberrations and by the nonlinear relation between contrast and defocus, but chiefly by the resolution limit of the retina and visual pathways. However, it is found that the apparent lack of response to small deviations from focus is due to a drastic compression rather than absence of threshold changes. If responses through this region are shown on a greatly expanded scale in relation to the rest of the function. fluctuations appear that often match those generated by border enhancement.