On the right, you see some (rather irritating) pattern flicker. Inspect the top grating row – does it look a little reddish? The row below a little greenish (very little)? The third row a little blueish?
The colours are just a little stronger (more saturated) if you speed up to zero frame delay between each flicker step.
Slow down the flicker with the stepper at the bottom left – e.g. delay to 10. Then you see that there is only black and white in the display, no colour. The sequence is full black, full white with the upper grating row, second grating row, third grating row, and full white alone. The colours are illusory.
Another parameter to change is the number of stripes, technically the spatial frequency of the grating. For me, the colours are present without difference from 1 stripe to ≈50 stripes.
Maybe you already saw Benham’s Top on the preceding page. There, I had strived to imitate an actual spinning top. But the rotation is not necessary – the local flicker and pattern are the relevant stimuli. So I “unrolled” the Benham Top into the present demonstration. It is just as strong (or weak, as the case may be for you).
What was unexpected for me: the spatial frequency of the grating seems irrelevant. This is in contradiction to the explanation (“pattern induced flicker colors”) that there is center-surround opponency in retinal colour-specific cells at play – when the bars are wide, they certainly cover center + surround.
See Benham’s Top
