Abstract
Orientation thresholds of cued targets increase with set size, but only when target and distractors are in close proximity (Morgan et al., 1998; Felisberti & Morgan, 2001). Such impairment is generally called ‘crowding’ and cannot be explained solely on the basis of spatial uncertainty or lateral masking. Since orientation averaging in a central feature integrator seems to occur during crowding (Parkes et al., 2001), how will the weighting of afferent stimuli be affected by their contrast?
Stimuli were Gabor patches (3.6 c/deg) at 4 deg eccentricity. Observers had to judge whether a patch flanked by 2–4 neutral distractors was tilted clockwise or anticlockwise. The relative contrast of target and distractors was varied as well as the target position, which was either random or fixed to the centre of the array.
Lower orientation thresholds were observed when the contrast of the distractors was lower than the target contrast, and higher thresholds when the contrast of the distractors was higher than the target contrast, independently of whether target position was known or not.
Our results so far indicate that crowding does not depend on grouping-by-contrast, in agreement with Chung et al. (2001) and point to a second-stage averaging process where contrast information from target and distract0rs is weighted by contrast.