Abstract
Viewing a high contrast pattern for a few seconds reduces sensitivity for similar patterns: more contrast is then needed to see a faint test pattern that resembles its predecessor than if the test pattern is unlike its predecessor. Thus after exposure to a high contrast drifting vertical grating, the contrast threshold is higher for vertical than for horizontal test gratings.
If the adapting grating is a rapidly phase-reversed (or counterphasing) grating, it loses perceived contrast, becoming invisible at sufficiently high reversal rates—an expected consequence of visual persistence. We find, however, that rapidly counterphasing adapting gratings remain unexpectedly effective in reducing contrast gain after the adapting exposure. This loss of contrast sensitivity is (as usual) highly orientation-selective and shows considerable interocular transfer.
Unfortunately the perceptual uniformity of rapidly counterphasing gratings is easily disrupted by eye movements. This motivated two further experiments. First, in a “phase contrast’ stimulus, a square-wave counterphasing grating is sampled by a lattice of circular windows: for a vertical grating, dark columns of dots alternate with light ones in space and time. At 80Hz, this looked like a steady field of identical light grey dots on a darker background, and observers performed at chance when asked to identify the phase-contrast boundary. Such a perceptually uniform field of dots nevertheless produced orientation-specific desensitization. Second, when counterphasing gratings are accompanied by masking stimuli to conceal the faint artifacts, forced choice identification of the adapting orientation fails, yet orientation-specific sensitivity loss remains. These observations imply central limitations on the time resolution of perception, not shared by the contrast gain control.