Purchase this article with an account.
Lynn Olzak, Patrick Hibbeler; Bandwidths of gain control pools in overlaid and center-surround masking. Journal of Vision 2011;11(11):1178. doi: 10.1167/11.11.1178.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Overlaid sinusoidal gratings which differ greatly from a target grating in spatial frequency or orientation can create a large performance drop in certain suprathreshold hyperacuity discrimination tasks. This between-channel masking has been attributed to pooled gain control processes. A similar masking effect is found when targets and masks are of the same spatial frequency and orientation, but spatially displaced (center-surround masking). These effects have also been attributed to pooled gain control processes. In the current experiments, we measure bandwidths of the putative gain-control pools with overlaid and center-surround stimuli. The target was always a vertical 4 cpd grating. Stimulus configurations and discrimination tasks were designed to isolate masking interactions previously found in the overlaid case: 1) orientation discriminations with an overlaid vertical 16 cpd mask and 2) spatial frequency discriminations with an overlaid 4 cpd horizontal mask. In the first case, orientation bandwidths were measured by changing mask orientation from the vertical until interactions were no longer observed relative to an unmasked control. In the second case, spatial frequency bandwidths were measured. Orientation and spatial frequency bandwidth measurements were then repeated in comparable experiments with center-surround stimuli. On both the orientation and spatial frequency dimensions, bandwidths measured with overlaid components were considerably broader than those measured with spatially displaced masks. To the extent that our masking results actually reflect characteristics of contrast gain control pools, these differences may indicate different pooling processes entirely, or the same process but with a more limited pool in the center-surround case. Finally, we note that an illusory contour separating center and surround becomes increasingly visible as target and mask differ. It is known that a real gap of just a few pixels between center and surround stops masking effects. We speculate that an illusory contour might also be able to restrict the masking effects.
This PDF is available to Subscribers Only