Interactions between modulated luminance patterns and random-dot patterns.

Abstract

It has been suggested that density modulated random-dot patterns can be used to study higher order pattern vision [Van Meeteren and Barlow (1981) Vision Research, 21, 765-777]. The high contrast dots of which the pattern is composed, are assumed to be reliably transduced and transmitted by the lower levels of the visual system. Therefore, such dot patterns could offer a way of by-passing the limits set by these earlier steps in the visual system. So, detection performance should reflect the capacity of more central visual mechanisms to combine and compare groups of dots. We test this assumption by selectively desensitizing the spatial frequency channels which are involved in detecting luminance contrast patterns. The results show a selective decrease in sensitivity for modulations in dot density at the adapting spatial frequency. We conclude that detection of differences in dot density is mediated by the same channels that detect luminance contrast. The conclusion by Van Meeteren and Barlow that dot patterns can be used to study higher-order processing in the visual system appears not to be valid. In addition, we present a new type of modulated dot pattern of which the density modulations are shown to be invisible for the spatial frequency channels. This pattern may therefore be used to study higher order visual processing.