The proposed experiments are designed to examine how accurately and reliably channels selectively tuned to spatial frequency and orientation convey information about the appearance of visual stimuli and how possible nonlinearities in the visual system may change perceived spatial frequency and/or orientation. Specifically, the aims of a given set of experiments are to (1) study the high spatial frequency cutoff as a function of eccentricity and determine whether distortions due to aliasing can occur in the peripheral visual field; (2) determine if Mueller's labeled line concept is valid for spatial frequency and/or orientation channels using visual patterns presented near threshold and, if so, the number of lines (channels) and the labels attached to those lines at a particular location of the retina; (3) determine how perceived spatial frequency (size) changes as a function of contrast, eccentricity, or orientation of visual patterns; (4) determine how perceived orientation changes as a function of contrast, eccentricity, or orientation; or (5) determine how the abnormally-functioning eyes of strabismic amblyopes or the reduced apparent contrast in anisometropic eyes of amblyopes may influence perceived spatial frequency as a function of eccentricity or contrast, respectively. The magnitude and direction of shifts in perceived spatial frequency (or orientation) will be compared to different predictions of several versions of multiple-channels models which differ in their assumptions about: (1) noise in the channels; (2) form of the contrast transfer function following the output of a given channel; and (3) how the outputs of different channels are labeled, weighted, and combined. These experiments will help extend models of multiple channels for detection and discrimination of visual patterns to models for the perception of the appearance of visual patterns. The proposed experiments utilize psychophysical methods with human observers. The observer is asked to (1) detect the presence of a visual pattern; (2) discriminate or match one visual pattern to another along a specified dimension (e.g., spatial frequency); and (3) estimate how different two visual patterns are along a specified dimension.
