an open letter to michael bach

November 11, 2020
The other day I linked to Michael Bach's "Optical Illusions & Visual Phenomena" page, today I finally got a chance to go through them all, and I shot him this note:

Hello! Thank you for your site on optical illusions, I enjoyed taking the tour.
I think the one most glaring absence (from the "face" section) was this:

The mutual distortions are truly remarkable, especially how the viewer can focus on one face or the other at will and it quickly returns to normal.

It probably wouldn't be too difficult to reconstruct that, I suspect any set of faces might work.

On more well covered territory:

On Structure from motion at equiluminance - I find I am relatively insensitive to the "equiluminance" aspect.

I first saw a version of the effect at the (sadly defunct) Boston Computer Museum - I think the visitor could add one dot at a time that would slide back and forth, and then as more dots were added, the rotation would become more obvious.

Years ago I recreated the effect for a "P5 advent calendar":treemergent and frostyspin

(I use trivial math - each dot has a yspeed, and if it's to the left of the center, a constant is added to the yspeed, otherwise subtracted - I always wondered how close that was to actual "sine" numbers and today wrote a little program to empirically demonstrate the similarity... pretty close! - I have forgotten too much Calculus to see if this is an equivalency via integration...)

The old 16bit computer game "Star Control" also used the effect for a rotating star map:


That shows the illusion would work even for "interior" objects, not just as points on the surface of a rotating solid.

Thanks again! I am amazed at how much "new stuff" has been discovered in terms of illusions: I feel like the books for kids I grew up with in the 80s were stuck with "these two lines are the same length!" and "look at this Escher thing!" - but the "rotating snakes" are truly mind bending.