Posted by theogravity 20 hours ago
It would be much funnier, and also more insightful, if it didn't do this and let you contradict yourself.
Also, I found that sometimes it looked like there were two colors. The top was green and bottom was blue. Maybe my monitor?
I noticed that what I was actually seeing later on was 'is this more blue or green than the last colour' due to my eyes adjusting to the previous screen and just seeing the difference.
First I shifted the app to use P3 `oklch(.7066 .1611 $hue)` with range (150..210) centered on cyan at 180°, same as sRGB. No change, so it's not some sort of artifact of colorspaces. Then I upped it to 16 steps instead of 8. The window narrowed slightly, but the same first-then-the-rest shift kept happening. Finally I raised the random color static mask duration from 200ms to 5000ms. Scored 180 +/- 1. Huh. Makes sense, given the image persistence stuff I deal with.
So, for those seeing that same variability I'd recommend editing that first (local response override index-blah.js, search `, 200` replace `, 5000` by hand, reload page) to get a more stable result.
No, turquoise is turquoise!!1
But with both eyes I got
> Your boundary is at hue 174, just like the population median. You're a true neutral.
I should test with one eye.
Meaning, there is no absolute color, the brain just learns what things have the same color, and how similar or dissimilar they are in hue to other objects. And for example “cold” colors are cold because we associate them with cold things, not because of some independent “qualia”.
If I'm looking at a certain color of green illumination and then cover one eye then the other, my perception of that color shifts slightly. It's still green, but with one eye it is "brighter" than the other eye.
Cataract implant technology is moving very fast, and my data is about 5 years old, so YMMV.