Purchase this article with an account.
Marco Puts, Joel Pokorny, Vivianne C. Smith; Inferred retinal mechanisms mediating illusory distortions. Journal of Vision 2003;3(12):62. doi: 10.1167/3.12.62.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Certain geometrical illusions disappear under equiluminance. Here we report data on the Zoellner illusion, in which a pair of parallel lines appear to diverge from each other when crossed by short diagonal lines. The diagonals on each line are parallel to each other, but are angled opposite to those on the adjacent line. The Zoellner, along with the Hering and Wundt illusions are examples of illusions in which non-orthogonal line crossings induce distorted percepts.
Why these illusions disappear under equiluminance is not clear. One possibility is that the illusion is established in the magnocellular pathway, which responds poorly to equiluminant stimuli. Alternately, the spatial resolution at equiluminance is substantially lower and may not support the illusory percept. A third alternative is that available cone contrast at equiluminance is not sufficient to support the illusory percept. We studied the origin of the Zoellner illusion in an achromatic experiment in order to separate these explanations. Two clearly visible parallel lines were presented along with diagonal lines for which the contrast sensitivity was measured. This was done using two paradigms: a steady-pedestal paradigm, in which the pattern was briefly presented against a continuously presented pedestal, and a pulsed-pedestal paradigm, in which the pattern and pedestal were briefly and simultaneously presented. Whereas depending upon stimulus conditions, thresholds in the first paradigm could be mediated within either the magnocellular or parvocellular pathways, the second paradigm favors the parvocellular pathway, because the magnocellular cells saturate due to their high contrast gain. The subjects performed two different tasks: 1) detect the presence of the diagonal lines, 2) judge if the illusion was present. The two pairs of sensitivity functions were compared to see if the illusion was established in the parvocellular or the magnocellular pathway, and how much contrast was needed.
This PDF is available to Subscribers Only