Abstract

Spatial frequency discrimination can be either facilitated or reduced by the addition of a grating mask to each test component, depending on the separation between mask and test frequencies. At small separations, performance is facilitated, probably due to unique beat patterns. At larger separations, performance is reduced. Reductions in performance might be explained if the mask introduces noise into the system; alternatively, the mask may alter the apparent spatial frequency of the test gratings, reducing the apparent difference between them. These possibilities were explored in a two-interval spatial frequency matching task. In the first interval, the test stimulus was repeatedly presented. A given test stimulus was composed of a vertical sinusoidal grating, either alone or masked by another frequency component. In the second interval, a sinusoidal grating of variable spatial frequency appeared. For each block of trials, the observer chose one component of the test stimulus to match. Blocks were repeated until the observer was satisfied that all distinct components had been matched for a given test stimulus. Matches were determined by averaging matches obtained by two randomly interleaved, independent staircases. The results provide information about the cues used in the complex discrimination task and are interpreted in relation to the discrimination data by means of a simple model.

© 1986 Optical Society of America