hide
Free keywords:
-
Abstract:
Surfaces covering the majority of a visual scene qualify for the role of an anchor--the value used by the visual system to compute the lightness of all other surfaces in the scene. In natural scenes, such surfaces can be a composition of different shades (eg sky, water). However, the visual system often treats such surfaces as having a single shade for lightness computations. In three experiments we used surfaces composed of gray shades as a background in a simultaneous lightness contrast display (SLC). We investigated the articulation of the background, and the duration of the stimuli. The observers' task was to judge the lightness of a comparison patch relative to a target patch in SLC, either by adjusting the comparison patch (experiments 1 and 2) or in a forced-choice task (experiment 3). In experiment 1, we placed the target patch on either uniform backgrounds--static, spatially filtered noise backgrounds; or dynamic noise backgrounds. Both types of noise background strengthened the SLC illusion. The effect was measurable as increased lightness of the target patch, indicating stronger articulation. This articulation effect was further tested in experiment 2 by increasing the spatial frequency of the noise. Although we found the usual SLC trends for the manipulation of target and background luminance levels, the size of the illusion was reduced in comparison to experiment 1. The spatial frequency did not simply change the articulation but also enhanced constancy. Finally, we manipulated the stimulus duration in experiment 3. We found that the illusion was weaker for short duration (320 ms) than for long durations (960 ms or more). This was the case for static and dynamic backgrounds, suggesting that the number of different frames does not play a role in lightness computations
