The effect of spatial adaptation on perceived contrast.

Perceived contrast was assessed by contrast-matching between adjacent sinusoidal gratings of the same spatial frequency (3 cycles/degree), before and after adaptation to gratings of various contrasts. On logarithmic axes, the effect of adaptation was large for test contrasts below the adapting contrast, but absent for test contrasts higher than the adapting contrast. This result rules out a simple gain-reduction hypothesis, in which adaptation would attenuate all test contrasts by the same proportion. Instead, results for all combinations of adapting and test contrast levels (including threshold elevation) conformed fairly closely to a simple subtractive rule: any test contrast presented after adaptation behaved as if one-third of the adapting contrast were subtracted from it. Though not exact, this may be a useful descriptive rule. Deviations from the simple rule may be explained by increased variance in the visual response at high adapting contrasts, combined with nonlinearity at low test contrasts. A subtractive effect at the psychophysical level does not necessarily conflict with evidence for contrast gain reduction at the single-cell level.