Abstract
Overlaid masks of very different spatial frequency or orientation can disrupt fine pattern discriminations. Lateral or surround masks can produce similar disruptions in performance. The effects have both been modeled as pooled divisive contrast gain control, but it is not clear a) whether the processes are identical in the overlaid and lateral conditions, or b) whether response gain control is involved as well as contrast gain control. We measured contrast response functions to evaluate whether curves shifted horizontally (contrast gain) or vertically (response gain) in the presence of masks. Observers viewed one of a pair of at-or near vertical 15 cpd sinusoidal gratings on each trial of a session, upon which spatial frequency or orientation discriminations were made in different sessions. Also in different sessions, masks were either overlaid or positioned as an annulus, and were presented either at 2% or 25% contrast. The orientation and spatial frequency of the mask depended upon both the task and the spatial configuration and was chosen because significant masking had been previously observed in the chosen conditions. Within each session, target contrast was randomly varied in equal linear steps from 3.1% contrast to 25% contrast (3.1%, 6.3%, 9.4%, 12.5%, 15.6%, 18.8%, 21.9%, and 25%) to generate contrast response functions. We fit Naka-Rushton functions to the data, with parameters C50 and Rmax free to vary. The functions accounted for the data well, and no significant differences were found between overlaid and surround masks, nor between judgment types. There was no contribution of contrast gain in the 2% mask conditions, but at the higher mask contrast, combined contributions from both types of gain were required.
Meeting abstract presented at VSS 2012