The use of the cancellation technique to quantify the Hermann grid illusion.

Abstract

When observers view a grid of mid-gray lines superimposed on a black background, they report seeing illusory dark gray smudges at the grid intersections, an effect known as the Hermann grid illusion. The strength of the illusion is often measured using the cancellation technique: A white disk is placed over one of these intersections and the luminance of the disk is reduced until the disk disappears. Its luminance at this point, i.e., the disk's detection threshold, is taken to be a measure of the strength of the illusion. Our experiments showed that some distortions of the Hermann grid, which were sufficient to completely disrupt the illusion, did not reduce the disk's detection threshold. This showed that the cancellation technique is not a valid method for measuring the strength of the Hermann grid illusion. Those studies that attempted to use this technique inadvertently studied a different effect known as the blanking phenomenon. We conclude by presenting an explanation for the latter effect.