Anomalous Motion Illusion
Anomalous motion occurs when some parts of a moving image appear to lag behind other parts of the moving image. That is, some parts of a moving image will appear to continue moving for a short period of time when the rest of the image stops. When the image first starts moving in the opposite direction, a part of the image will momentarily remain still. For an example of anomalous motion, scroll down to the activity.
Kitaoka and Ashida (2007) performed four experiments which support their explanation of the anomalous motion illusion: that random dots with low contrast (close to the same shade of gray) are processed more slowly (have a longer latency) by the visual system than random dots with higher contrast.
This example of anomalous motion works best if viewed in the periphery. Notice the image moving left and right. When the image reaches the left or right edge of its movement, and momentarily stops, the central circle appears to continue moving for just a little bit longer. When the image first starts moving in the opposite direction, the central circle appears to stay stationary for just a little bit of time. That is, the motion of the central circle appears to lag slightly behind the motion of the rest of the image. In reality, the motion of the central and surround parts of the image are identical.
Notice that the small random dots that make up the central circle have a lower contrast (smaller difference between the brightest and darkest dots) than do the large random dots that make up the background of the image. The low contrast dots take a little longer to be processed than do the high contrast dots and this increased latency leads to the perceptual lag of the central circle compared to the rest of the image.
Kitaoka and Ashida (2007) found that low contrast random dots were more likely to produce the illusion than higher contrast dots. My personal experience is that you want a moderate amount of contrast for the illusion -- too little or too much contrast seems to destroy the illusion. Adjust the contrast slider to find a contrast that produces the strongest illusion for you.
If contrast is the key to the illusion, then the overall brightness of the dots should not influence the illusion as long as contrast (the difference between the lightest and darkest areas) remains constant. That is, having the darkest dot at an intensity of 50 and the brightest dot at an intensity of 100 (contast of 100 - 50 = 50), should have the same strength of illusion as a darkest dot at intenstiy 200 and the brightest at intenstiy 250 (250 - 200 = 50). When the Brighten checkbox is checked, all of the dots will have a constant value added to their intensities so that the brightest dot will have the maximum intensity permited by your monitor. Thus, the contrast should remain constant. "Should" depends on your monitor -- many modern monitors and graphics subsystems have a linear relation between the RGB values of the pixles and their brightness. If yours does not, then the contrast is unlikely to remain constant.
Try checking and unchecking the Brighten checkbox for a given contrast and see whether brightening influences the strength of the illusion.
Kitaoka and Ashida (2007) also report that the effect occurs for rotations. That is, the center circle will lag in its rotation relative to the rest of the figure. You can print the image and rotate the printed page back and forth to see if the effect occurs.
Kitaoka, A. & Ashida, H. (2007). A variant of the anomalous motion illusion based upon contrast and visual latency. Perception, 36, 1019-1035. http://dx.doi.org/10.1068/p5362