Abstract
Superimposed grating masks of very different spatial frequency (SF) and/or orientation (O) drastically reduce performance in fine spatial discrimination tasks. Surround masks, however, must be highly similar in both SF and O to be effective. Here, we asked a) how the two types of masks interacted, b) whether the effects depended upon the dimension of the judgment (SF or O), and c) whether segmenting boundary cues disrupted lateral masking effects. The task was to discriminate between two 40-minute, circular grating patches at or near 4 cpd that differed slightly in SF or O using an SDT rating paradigm. In each block of 80 trials, observers rated their certainty that Pattern A or Pattern B had been presented, and daily d′ values were computed. In different blocks, test components were masked by superimposed components, surround components, or both. Surround masks were simple or complex gratings, and in some cases were differentiated from the center patch by the presence of a gap, a change in mean luminance, or relative phase shifts. Bother superimposed and lateral simple masks were effective. In general, masking was always greatest when center and surround were nearly identical in all respects. Segmentation cues reduced or eliminated lateral masking. In SF tasks, complex surrounds reduced masking relative to a single-component surround, suggesting disinhibition of the lateral masking effect. Such disinhibition was not found when the task was an orientation discrimination. The results suggest that a) superimposed and lateral masking effects have different properties, suggesting different underlying processes; b) properties of masking differ for SF and O judgments, suggesting that different higher-level pathways subserve these two tasks; and c) perceptually segmenting a scene reduces or eliminates lateral masking effects in the normal visual system, suggesting that object segmentation processes play a role in lateral masking.