Purchase this article with an account.
Lawrence K. Cormack, Scott B. Stevenson; Illusory reverse-motion from contrast modulation. Journal of Vision 2002;2(7):257. doi: https://doi.org/10.1167/2.7.257.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose: When a drifting grating is contrast modulated over time, a strong illusory percept of reverse motion occurs when the stimulus is near zero contrast. The purpose of this study was to demonstrate and understand this illusion. Methods: The primary stimuli used to explore the illusion were rotating radial gratings. The perceived motion of a contrast-modulated test stimulus was matched with a constant-contrast comparison stimulus containing adjustable motion reversals. We did several experiments exploring the effects of contrast modulation and contrast masking on the illusion, and we measured perceived speed as a function of contrast for test stimuli of constant contrast. Various qualitative observations were also made with other stimuli including radial noise that was uncorrelated across temporal contrast cycles. Results and Conclusions: The illusory percept builds over time but is inconsistent with a traditional motion aftereffect; the percept builds very quickly, and square-wave contrast modulation produces no illusion. An extrapolation of perceived speed vs. contrast functions does not provide a reasonable explanation for the illusion either. Decreasing the depth of the contrast modulation by changing either the maximum or minimum contrast attenuates the illusion. The temporal phase of spatially-interleaved, high-contrast masking noise strongly affects the illusion with maximum attenuation occurring at pi-phase. Noise that is uncorrelated across temporal cycles produces a strong illusion, which obviates an explanation based on phase ambiguity across temporal cycles. The presence of a delayed negative afterimage additively interacting with the stimulus can qualitatively account for many of the observations, but the delay required is much longer than that typically reported between populations of cell types early in the visual pathway.
This PDF is available to Subscribers Only