What to see
Two striped disks rotate slowly over a bipartite blue background grating. This is just the eye catcher…
What to do
Grab the horizontal slider and move it such that the gratings on the two disks are aligned (collinear). Then do one of two things:
• click the ± button. That inverts the background grating direction, now they are no longer aligned
• click the “hide” button. When the background grating disappears, again the disks are no longer aligned, and you can read the misalignment angle in the box at the left.
The demonstration is set up to allow quantitative measurements on the strength of the illusion depending on the angle disparity of the background gratings.
This is now generally known as the “tilt illusion” and was first described by Gibson (1937). Blakemore and colleagues developed an explanatory hypothesis that the tilt illusion is caused by lateral inhibition between cortical orientation detectors. A more recent suggestion by Schwartz traces the effect to optimized processing of natural image statistics (which doesn't contradict the inhibition explanation, but puts it into a Bayesian context). Rather than further paraphrase, I point to the excellent Wikipedia entry on this subject.
Blakemore C, Carpenter RHS, Georgeson MA (1970) Lateral inhibition between orientation detectors in the human visual system. Nature 228:37–39
Gibson JJ (1937) Adaptation, after-effect, and contrast in the perception of tilted lines. II. Simultaneous contrast and the areal restriction of the after-effect. J Exp Psychol:553–569