wavelength = 1/frequency
if you’re at the beach, freqency is the number of waves that splash you per minute, wavelength is the distance between each wave.
Colour is said to be the primary wavelength component of a light.
A colour is made up of a range of different wavelengths (frequencies) of a whole host of photons impinging on the retina. (I will ignore the interpretation by the brain of the signal for now).
Imagine each photon interacting with a molecule of the receptor chemical in the cone. Each cone will be stimulated to a certain degree depending on the number of photons that have a wavelength that falls close to the wavelength that that cone responds to. A photon can have an infinite (close enough) number of wavelengths and there are a massive number of them impinging on the retina to create a colour. The limiting factor is the equipment that measures it - the eye.
There are 6 or 7 million cones (each about 0.006 mm in diameter). Each cone will have a who lot of each chemical that responds to each frequency range, but each cone is just generating one electrical signal (i think).
So, if there are say 2 million of each of the 3 types, then there are 2 million to the power of 3 possible colours.
8E18 or 8,000,000,000,000,000,000 colours.
But the brain distinguishes much less.
Does that sound right?
Even on the limited palette available on the SDMB, there are a wide range of possi bilities, as this post may help to make clear. With HTML enabled, even more are possible.
ok people, there seems to be a confusion between light and colour here.
In colour (solid), there are 3 primary colours: red, yellow and blue. By combining these 3 primary colours, you can get all other colours. The first generation of mixed colours is called secundary colours, or “complimentary colours”. For example: If you put the three colours in a triangular shape
red
yellow blue
then red and blue mixed would get yu purple, which is the complimentary colour of yellow.
blue and yellow would get you green, which is the complimentary colour of red, and so on and so on. The colour black does not really exist, black is all colours put together, so our eyes can no longer make a distinction between the colours, but see it as “dark”. You’ll notice that with clothes: black clothes never stay black, they either fade with a brown tinge, a blue tinge, or a green tinge. meaning the piece of clothing wasn’t really black, but Very Dark Green or Blue of Brown. So: all colours = black, absence of colour = white
With light (rays of light, sunshine), it’s the exact opposite: white light is the combination of all lights, and black is literally the absence of light, ie shadow. When using a prism, we can differentiate the lightbeam into it’s main 7 colours (rainbow colours) red, orange, yellow, green, blue, indigo and violet.
so: all light = white
no light = black
Incidentally, these two “aspects” of colour (colour solid and colour light) are very closely related. In flowers: we see the colour that the flower rejects: all light (the whole spectrum) is absorbed by the flower, apart from one, which it reflects. And that is the colour we “see”.
Very interesting stuff.
read: Dear God, this is Anna, by Flynn.
it’s a must 
Hmmm…a quibble here:
- “By combining these three primary colors we can get all other colors.” - No. Get me magenta. Or pure black (the best you’ll do is a muddy brown). Or even a good, solid midnight blue. True, many colors can be produced by using the “artist’s primaries,” but plenty lie outside the color gamut.
A bit more accurate are the true subtractive primaries: cyan, magenta and yellow. In the printing process black is also added, since CMY together don’t produce true black. This is what’s used in 4-color printing in magazines, newspapers, most printed material. You got a color printer? These are the primaries it’s using. Even CMYK has plenty of colors outside its range. Ever try printing a saturated orange on your Epson? Not a chance, unless you can use custom inks.
yes, pulykamell, i actually work a lot with printers, and printing, so I know all about CMYK colours, or full process colours, but I didn’t think everyone would know.
That is why I opted to keep it reasonably simple, by using the “normal” words for blue, red and yellow 
Hey, it’s the SDMB…you could use big words all you want 'round these parts. It’s takes a bit of effort to be overtechnical here.
Thanks, mate, I reckon i need to get used to having intelligent people to converse with.
Note that “color” is a very abstract term. There are a lot of aspects of light that the human eye can detect that are considered “color” related but don’t always have to do with the wavelength of photons. “Silver” is a color to most people but is mainly about reflectivity. Human skin has a depth to it that is very hard to reproduce using computer graphics (or wax museum models). Such artificial skin looks wrong to us because “the color isn’t right”. Note also things like “neon” paints have additives that absorb light in one wavelength and emit it in another. Such a shift from the ambient light is interpreted as coloring. All 3 of these examples cannot be done well on a computer screen.
There’s more to color than just mixing primary colors. “Infinite” is as good an answer as any.
I used to work for a company that sold clothing via catalog. You might be surprised at the time and energy invested in dreaming up names for colors. (“Does this look like ‘fog’ or more like ‘mist’”?)
LOL! I always wondered who sits in a room and comes up with color names for clothes and cars. Any special training required???
Are you sure; wouldn’t the discrete nature of photons make it only possible to mix in integer proportions? (also ignoring that wavelengths are not discrete)
Even if the mixtures could only be made in integer proportions, the integers go on for infinity.
True enough, but then I wouldn’t want to be directly looking at some of the large-value mixtures (I suppose the point is moot as human perception isn’t sensitive to one-photon-difference changes in colour).
If one believes that the absolute upper limit to wavelengths is the Planck Scale…one Planck Length between crests…and simply adding Planck Lengths to that (2PLs, 3PLs, 10PLs, XPLs, PL^X, etc.) .
This goes from Cosmic rays, Gamma rays, X-rays, Ultraviolet, Visible Spectrum, Infra-red, Microwaves, Radio Waves, and keep on going into infinity…or at leas to the wavelength equal to the size of the entire universe.
I say, if they made it into a crayon, then it’s a color. If it’s not a crayon, then it’s not a color. Like the whole “is clear a color?” thing. You don’t see any clear crayons, now do ya?
When it all boils down, kindergarten logic works for me.
Well, this work by Berlin & Kay is about colour names - and for a colour name to be ‘valid’, it has to fulfil some specific criteria (e.g., it cannot apply to a restricted set of objects, which discounts the term ‘blond(e)’). This sort of work says nothing about the number of colours we can discriminate.
Regarding how many colours we can ‘perceive’, the answer depends on how we measure the perception! For example, side-by-side discrimination of colour samples leads to larger estimates of the number of discriminable colours than successive comparisons.
It’s sort of vaguely right in parts! But - outputs from cones are pooled in particular ways. That is, there are fewer outputs from the retina than there are cones. Plus, the number of ‘colours’ we can perceive is not a function of the number of cones at all. It’s a function of the magnitude of the outputs of the ‘pooled’ signals (well, among other things).
We can see many colours that aren’t shown on a computer screen though… we can see darker colours - if the black on the screen is lightened by light in the room or adjusting the brightness controls… we can see far brighter colours - like light-globes.