Craik-O'Brien-Cornsweet Illusion

From Michael’s “Optical Illusions & Visual Phenomena”

What to see

At top right the is a yellow­ish ring, we need that later. At top left and bot­tom right there are two grey disks in whose cen­ter there is a dar­ker disk. Com­pare the two large disks, they are quite alike, no?

What to do

You can grab the yellow ring with the mouse and move it around. Place it on the cen­ter of the large top left disk – the dar­ker cen­ter is visible in the hole. Now place it over the bot­tom right disk. Unex­pec­tedly, the cen­ter shows the same light­ness than the sur­round!

Using the slider at bottom left you can increase the con­trast. With high levels of con­trast, it becomes imme­di­ately obvi­ous that the bot­tom right cen­tral disk has no con­stant grey level.

Comment

The retinal ganglion cells encode the incoming luminance profile via their centre-surround luminance profile in a sort of “delta code”. In the cortex this is integrated to perceive the veridical square luminance profile.

The profile of the Craik-O’Brien-Cornsweet disk on the right is (almost) a fixpoint for this encoding, thus sending a nearly identical spatial code up the cortex. The cortex thinks: “Oh, I know, that’s the ganglion cells doing their usual spiel”, integrates it and arrives at the same perceptual result (yes, I know this is a simplification).

Sources

O’Brien V (1959) Contrast by contour-enhancement. Am J Psychol 72:299–300

Craik KJW (1966) The nature of psychology (Sherwood SL, ed). Cambridge, UK: Cambridge UP

Cornsweet TN (1970) Visual perception. Academic, New York

Dale Purves’ demonstration

Demo from Ted Adelson’s site