Abstract

Saturation of Fechner–Benham subjective colors depends on the spatiotemporal properties of the inducing stimulus. This dependence can be explained by the properties of the Type I cell signal. Type I cells multiplex hue and luminance by adding and subtracting the center and surround spatiotemporal filters. Because the resulting color and luminance filters overlap in both frequency and phase space, the results of any demultiplexing process are ambiguous for some spatiotemporal stimuli. Therefore, according to signal detection theory, the saturation of an illusory color inferred from that ambiguity should be proportional to the degree of overlap between the encoding filters. Assuming equal sensitivity for color and luminance at the decoder, Billock (1989) used Kelly's (1983) color and luminance CSFs to predict the temporal-saturation tuning of subjective red. These results are improved for subjective red and extended to green and blue by using Grigsby and Ingling's (1989) and Kelly's (1974) estimates of R, G, and B Type I-like lateral inhibitory mechanisms. With no free parameters, the quality of the fit provides further evidence that the Type I hue and luminance multiplexing/demultiplexing operation is the locus of Fechner–Benham subjective color.

© 1990 Optical Society of America