Purchase this article with an account.
Timothy Vickery, Won Mok Shim, Yuhong Jiang, Ramakrishna Chakravarthi, Robert Luedeman; Supercrowding: Weakly masking a target greatly enhances crowding. Journal of Vision 2008;8(6):971. doi: 10.1167/8.6.971.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Crowding describes the impairment in identifying targets in the periphery that are flanked by similar distracters, which typically occurs when the target-distracter distance is less than half the target's eccentricity. Here we report a surprising finding, whereby the range of crowding is dramatically enhanced by weak masks surrounding or trailing a target. Observers identified the orientation of a T-shaped target placed 12° in the periphery. The T was either unflanked or flanked by distracter Ts appearing at target-distracter distances equal to 30%, 50%, or 70% of the target's eccentricity. The target was also either not masked, or enclosed by a white square contour, or followed by a weak backwards pattern mask. When no flankers were presented, performance was near ceiling, and there was only a small effect of adding the mask: with a surrounding contour performance dropped from 100% to 87% and with a backwards pattern mask performance reduced from 99% to 93% on average. Flankers at a target-distracter distance of 50% and 70% of the target's eccentricity did not produce crowding when the target was not masked, as expected, but dramatically impaired target identification when the target was weakly masked. The effects of crowding and masking were super-additive, and thus could not be explained by their simple combination. Further experiments showed that these flanker effects obeyed some basic properties of crowding: the outermost flanker was most effective and decreased target-distracter similarity reduced crowding. Critically, “weakening” the target by reducing its contrast did not have this dramatic effect. Although the identification of the target in isolation was impaired by reduced contrast, distracter-target distances of 50% and 70% of target eccentricity did not induce crowding for a low-contrast target. Weakly masking a target reveals “supercrowding,” suggesting that feature integration in the periphery occurs over much larger distances than previously observed.
This PDF is available to Subscribers Only