Abstract
In the White illusion, test bars colinear with an inducing grating’s darker bars appear lighter than when the same test bars are flanked by the inducing grating’s darker bars. We previously reported that the function relating illusion magnitude to the contrast of the illusion inducers changes from nearly linear to clearly nonlinear as one progresses from low-level sensory to high-level perceptual illusions. Here we investigate additionally how the White illusion depends on 1) the inducer grating’s contrast (0.00 – 0.96), 2) the contrast of the test bar (0.12, 0.26) and 3) location of the test bar (colinear, flanked). Observers were required to adjust the luminance of a comparison bar until it matched the test bar’s perceived lightness. We define the contrast-dependent changes of the illusion magnitude as the inducer’s contrast-response function (CRF). Based on prior findings pointing to the important roles of high-level perceptual effects (e.g., grouping, surface transparency) in modulating the White effect, we expected 1) that the CRF is nonlinear for both test-bar contrasts and for both test-bar locations, and 2) that the CRF shifts toward a higher inducer contrast when the test-bar contrast increased from 0.12 to 0.26. Additionally, variation of test bar location allowed us to see if the contribution of the colinear and flanked test bars to the White illusion were equal or not. Results confirmed our predictions and support prior findings indicating that the White illusion is not produced by low-level (e.g. lateral inhibitory, simultaneous-contrast) effects but instead relies on higher-level perceptual effects, thus implicating correspondingly higher-level cortical processing. Our results also showed that the colinear test bars contributed more strongly to the White illusion than the flanked test bars did.