Abstract
We previously proposed a quantitative model of achromatic color induction based on the idea that underlying lightness and darkness induction signals fill in from borders in separate neural filling-in networks before being combined to create an achromatic color signal (Rudd, 2001; Rudd & Arrington, 2001; Rudd & Zemach, 2002). A key innovation of this model is the concept of induction signal modulation: Induction signal strength is altered when the signal crosses an intervening border by an amount that depends on the log luminance ratio of the crossed border. We show here that signal modulation depends on the induction signal type (i.e., lightness or darkness) and the contrast polarity of the crossed border, and we investigate the spatial falloff of the induction signals that are generated by borders of different contrast polarities. Our psychophysical experiments were based on a lightness matching paradigm. The stimuli consisted of two disk and ring patterns presented side-by-side on a flat-panel monitor. The luminance of the right disk was fixed and its surround luminance was varied to manipulate the disk lightness. Three observers adjusted the luminance of the left disk to achieve a lightness match to the right disk. Four experiments were performed in which the rings had either higher or lower luminance than the background and the disks had either higher or lower luminance than the rings. Each experiment was repeated with rings of different widths to study the effects of changing the distance between the test disk and the outer border of the ring. The results both support and generalize our earlier findings suggesting the existence of separate underlying lightness and darkness induction processes. The spatial falloff in induction strength follows a different function for the two types of induction signals. Lightness induction falls off with distance, whereas darkness induction strength remains roughly constant over the range of distances tested (up to 2.5 degrees).