The lilac chaser is a visual illusion, also known as the Pac-Man illusion. It consists of 12 lilac (or pink, rose, or magenta), blurred discs arranged in a circle (like the numbers on a clock), around a small black, central cross on a grey background. One of the discs disappears briefly (for about 0.1 seconds), then the next (about 0.125 seconds later), and the next, and so on, in a clockwise direction. When one stares at the cross for about 5 seconds or so, one sees three different things:
The chaser effect results from the phi phenomenon illusion, combined with an afterimage effect in which an opposite color, or complementary color – green – appears when each lilac spot disappears (if the discs were blue, one would see yellow), and color adaptation of the lilac discs.
The illusion was created by Jeremy Hinton some time before 2005. He stumbled across the configuration while devising stimuli for visual motion experiments. In one version of a program to move a disc around a central point, he mistakenly neglected to erase the preceding disc, which created the appearance of a moving gap. On noticing the moving green-disc afterimage, he adjusted foreground and background colours, number of discs, and timing to optimise the effect.
In 2005 Hinton blurred the discs, allowing them to disappear when a viewer looks steadily at the central cross. Hinton entered the illusion in the ECVP Visual Illusion Contest, but was disqualified for not being registered for that year's conference. Hinton approached Michael Bach, who placed an animated GIF of the illusion on his web page of illusions, naming it the "Lilac Chaser", and later presenting a configurable Java version. The illusion became popular on the Internet in 2005.
The lilac chaser illusion combines three simple, well-known effects:
These effects combine to yield the remarkable sight of a green spot running around in a circle on a grey background when only stationary, flashing lilac spots have been presented.
As of May 2011[update], no systematic study of the stimulus properties of the illusion had been published. Hinton optimised the conditions for all three aspects of the illusion before releasing it. He also noted that the colour of the green disc could be outside the colour gamut of the monitor on which it was created (because the monitor never displays the green disc, only lilac ones). Michael Bach's version of the illusion allows viewers to adjust some aspects of the illusion. It is simple to confirm that the illusion occurs with other colours.
It is not necessary to fixate on the black cross for the effects to occur. As long as the eyes are held steadily on any point of the figure, even the centre of one of the discs, the illusion will occur.
If instead of fixating on the black cross, one follows the moving gap with one's eyes, one will see only a moving gap and 12 lilac discs rather than a single green disc. This is because the green disc arises as an afterimage, requiring the eyes to be held steadily to occur.
If after looking at the effect for 5 minutes or so, one moves one's eyes elsewhere (e.g., to another point on the figure or to a blank sheet of white paper), one will see a stationary ring of 12 green discs that will fade after a short time. These green discs are the afterimages of the 12 lilac discs.
If one watches the illusion for long enough to see only the green disc and then moves away from the computer screen while keeping the eyes on the cross, one sees larger green spots outside a ring of lilac spots with a smaller green disc running around them. The smaller green disc may merge briefly with the outer green spots, making the spots appear to be radial blobs. The outer green spots soon fade. These outer green spots are afterimages that appear larger because of Emmert's law: the size of an afterimage becomes larger as its viewing distance is increased. They are outside because moving away from the computer screen has decreased the visual angle of the lilac spots. They fade because the lilac discs that constantly refresh the green afterimages are now projected onto a different part of the retina. If one moves towards the screen, the effects are opposite.
If one closes the right eye and moves close to the stimulus so that the nine-o'clock disc falls in the blind spot, one sees that the movement is no longer smooth. There is a noticeable pause when the disappearance of the disc occurs on the region of the retina having no rods or cones. This suggests there are limits to the filling-in that normally prevents us from noticing a black hole in our visual fields at the location of the blind spot.