The demonstration is an extension of the 2009 Curveball illusion (http://illusionoftheyear.com/cat/top-10-finalists/2009/).
In the ZigZag, the internal spin moves back and forth, while in the Curveball, the internal phase moves in a single direction with a constant speed. I think the alternating directions in the ZigZag create a much stronger effect than the constant drift in the Curveball (but I have not yet measured the difference).
Here is my take on the ZigZag illusion:
I am interested in how the visual system resolves conflicting aspects of the stimulus. The ZigZag demo has two types of motion: spinning motion (which is luminance-defined) and dropping motion (which is contrast-defined). Central vision processes are capable of keeping these types of motion separate from each other. Peripheral vision seems to lack these processes, and that is one reason why perception in peripheral vision can seem so strange.
Cavanagh & Tse refer to this class of stimuli as Double Drift phenomena (which is a better scientific term). They study the Double Drift to examine mechanisms that the brain uses to predict spatial location. Their latest work shows that the drift effect persists in the presence of smooth-pursuit eye movement. This research is really quite spectacular, both scientifically and for those who like cool demos.
Also, just found a new paper by Liu, Yu, Tse, and Cavanagh, in which they examine double-drift stimuli in fMRI. They report correlations between the perception of the drifting motion and areas of the brain that lie outside of the visual system.
Other vision scientists have worked on double drift (read the intro in Cavanagh and Tse’s paper for a current list). David Knill’s approach to the problem should be mentioned: https://www.pnas.org/content/112/26/8142.
And now… on to the demo:
Below is an Amazon link to Patrick Cavanagh’s new book The Visual World of Shadows. Support this blog: click on the link, and get a good book!