Perception: Colours

This is a fragment of a conversation in a newsgroup some years ago. I included parts of three prior posts (in chronological order) in the thread to show the concepts I commented on.

[A]
I was once asked "How many colors are there?". A difficult question, which many people can't answer, they've confused "how many colors" with "how many WORDS for colors". 2^24 (16,777,216) is a better answer than that.

[B]
I agree with 2^24, that seems to be as much as the eye can distinguish.

[C]
Which doesn't mean that reality is so limited. Note that some animals see better than humans, which isn't relevant either.

[Me]
There's a difference between colours as measured by a spectrometer and colours as perceived by a human. Eg, there is no such colour as "brown" in the spectrum. Or "pink". Or "grey". Or etc.

The 2^24 number of colours are the combinations of colour data used to display colours on a screen. Whether there are actually that many colours displayable on a given screen is another issue: screens vary quite a bit in quality, though much less nowadays than they did in the Olden Days. And whether a human can distinguish them all is another issue. And whether they can replicate natural colours in all weathers is another issue again. As anyone who's tried to make a photo "look right" knows.

As for "see better", that's not a clear concept either.

When it comes to perception, the only thing we can objectively measure or observe is what colour (or other sensory) differences the animal can distinguish. While it's true that bees can distinguish ultraviolet wavelengths, that doesn't mean they "see better". They see well enough for their survival, and that's what counts.

Or take frogs. Judging by their behaviour, they can't see fly-sized blobs unless those blobs move. I surmise that's similar to human peripheral vision, which is much better at distinguishing moving blobs of light than still ones.

Bottom line: what's "out there" isn't what we think it is.