The effect of spatial frequency on the white, shifted white and checkerboard illusions: Data and modeling

In White's effect (White, 1979) equiluminant gray test patches placed on the black and white bars of a square-wave grating appear different in brightness. The gray patch on the black bar is brighter than the gray patch on the white bar, independent of the aspect ratio of the test patch. Unlike other contrast and assimilation effects, this means that the White effect does not correlate with the amount of black or white border in contact with the test patch or in its general vicinity. White (1979) concluded that the effect must therefore depend on the directional (orientation) properties of the inducing grating. The effect of spatial frequency on the White effect, the shifted-White effect (White, 1981) and the checkerboard illusion (DeValois & DeValois, 1988) was measured for four subjects using a brightness matching paradigm. The data indicate that the White effect is similar at all spatial frequencies and confirm White's observation (White, 1979) that the magnitude of the effect is greater at high spatial frequencies. The shifted-White and the checkerboard stimuli, however, produce brightness effects that reverse direction as a function of spatial frequency. These reversals represent a difficult challenge for brightness explanations based on mid-level and high-level grouping characteristics since such explanations should not be sensitive to simple changes in spatial scale. Interestingly, the brightness effects are predicted by an oriented-difference-of-Gaussians (ODOG) model (Blakeslee & McCourt, 1999) and, therefore, may not require an explanation beyond low-level multiscale spatial filtering and response normalization.