yea, there’s also the afterimage / auto white balance factor

also those are fun optical illusions

mildly unrelated, but have you seen https://en.wikipedia.org/wiki/Café_wall_illusion and https://www.shadertoy.com/view/4dBfWK

IDK, I can’t tell from looking at the 2015 CIE CMFs (I think these are the most accurate? also I used the firefox plugin “unpaywall” to see them as sci-hub wasn’t working) if there are any completely identifiable red colors or not. I initially assumed there were, but I guess I don’t really know (I had assumed any perceived color could be made from a standard red green and blue, but now I also don’t know if that’s true).

edit: if that assumption is true than there would be no way to produce photons of different wavelengths in a way that looks like a fully saturated red

also the falloff at the end of the spectrum might mess with that a little, it looks like there is a continuously varying ratio of red to green along the end of the spectrum, but I can’t really tell

edit2: it also varies somewhat with age and among individuals apparently, so that might complicate things further

If you have multiple ports driven off the same internal hub, they will share bandwidth.

I’m usually happy with increased efficiency as it represents an increase in performance in the future. Cost is something that seems much more inevitable to go down than performance is to go up, so the two metrics I look for in the state of the CPU market are peak single core performance and performance per watt. Of course, this only applies to observing the industry from outside, I’m sure if I was actually in the market for a new CPU right now I’d probably be happier with a worse performance per watt chip as long as it was cheaper.

yea I know thats why I called it a slippery slope, because if you know for certain that there’s only one wavelength in the scene then you could tell its cyan, but if it could be any spectrum then you would have no idea

I called black a real color because if we see black, we know for certain there’s actually no light.

And yes I know all that other stuff.

I guess you could call this pink but I would normally think of it more in the magenta range of or

in the c/196 banner there is more red than green, meaning the fully saturated version wouldn’t correspond to a real wavelength

anyways, I’m not trying to say that those colors don’t exist because obviously we can see them, just that they show the weaknesses of human color perception

I’ll start by saying the computer rgb and hsl models are an abstraction matching how we perceive light that doesn’t include any other spectral information that we can’t see

in terms of wavelengths of light, pink is not a hue, unlike other fully saturated colors it does not have a wavelength to go along with it

yes, you can combine red and blue to make pink, but then you’re looking at the wavelengths of red and blue and not those of pink (because pink doesn’t have a wavelength)

any color that isn’t red, green, or blue could be any one of an unlimited number of spectrums that would produce the same perceived result, but any fully saturated color other than pink does have one single wavelength that goes along with it

like I said above, hue is analogous to wavelength, saturation is analogous to the ‘unpure-ness’ of the wavelengths, and value/luminance is analogous to the quantity of those wavelengths - and pink is one that breaks this pattern

look at this chart, and look for the pink. It isn’t there, because pink isn’t real

I should have excluded king blue and violet tho, those are actually real, mb