Abstract
When a target in the periphery is displayed next to a matching mask, contrast sensitivity is suppressed, producing elevated contrast increment thresholds. This surround suppression is strong only when the mask is located close to the target and, literally, surrounds it (Petrov & McKee, 2006). For example, a mask of the same size as the target produces negligible suppression unless juxtaposed with the target, even in this case the suppression is weak (Petrov, Popple, & McKee, 2007). The target in the present study was a uniform luminance disk displayed on a gray background 10 deg to the left and right of fixation. From trial-to trial, observers were asked to identify which location contained an incremental change in contrast. A mask (50 % contrast disk of the same size as the target) was presented in both locations 5 deg inward of the target; it was displayed for the same duration as the contrast increment (8 – 150 msec). We found that for brief durations ([[lt]]30 msec) surround suppression was amplified five-fold if the contrast increment lagged the target pedestal onset by 100 – 1000 msec. For latencies shorter than 100 msec the increment fused perceptually with the pedestal onset, and the suppression amplification effect promptly disappeared. For longer latencies the effect became less pronounced. The effect grew with the pedestal contrast, but saturated above 20%. We found similar results for Gabor targets and masks. Surprisingly, we found no significant effect of mask orientation on suppression strength. Conventional surround suppression disappears if mask and target orientations differ by more than 45 deg (Petrov, Carandini, & McKee, 2006). To summarize, we observed a new powerful type of surround suppression arising, apparently due to some form of synergy between pedestal masking and surround masking.