Abstract

At the 1991 OSA Annual Meeting, I described a model of TVI curves based on the assumption that vision is limited by photon fluctuations at high background levels.¹ The theory predicts that a de Vries-Rose region or an asymptotic Weber region—or both—may be observed, depending on the area (or spatial frequency) and the duration of the test stimulus. The model accounts for changes in the slopes of TVI curves as a function of background illuminance without the need for postulating a gain control mechanism. However, certain types of gain control are allowed within the model. Many aspects of the time course of visual response are also accounted for by the model. To account for these latter data, a gain control mechanism (response compression of the output of a dc-suppressing receptive field mechanism) must be assumed. Many alternative models that account for TVI curve shape entirely on the basis of gain control assume that response compression depends on the background, as well as the incremental, intensity. Predictions of the two classes of models will be compared to psychophysical data.

© 1992 Optical Society of America