I've always accepted the idea that color spaces like RGB and CMYK somehow represented reality and an enriching abstraction... that is, until recently. Due to my failed efforts to get proper color from digital printing, it dawned on me that the only reason I accepted this nonsense at all was because I used a lot of image processing software. All computer programs force their users to work in some color space or another that the programmers decided to work in. In other words, if you want to accomplish anything with a computer, you must accept the fraud or be left behind.
But no more...
True color exists only as a subjective experience of an individual human. You can pretend that some coordinate system can be applied to color to make a "space" if you want, but this idiom only goes so far before it runs into trouble.
The typical human eye perceives three color bands corresponding to the three types of opsin in the retinal cone cells. Their maximum sensitivity to light occurs (roughly speaking) in blue, green, and orange light (respectively). These proteins, when exposed to light, generate cyclic GMP which closes an ion channel on the cone cell. This change in ion current becomes the raw material of the signalling cascade which gives us color perception.
Everyone who hears about the color receptors in the human eye assumes that each of these opsins are isolated from the others into primary color channels (blue, yellow, red), but this is an unwarranted simplification. The receptors are referred to as S, M, & L for the short, medium, and long wavelengths they respond to. There is actually a huge overlap in sensitivity between the M & L receptor. Stimulation of the multiple receptors generates the "spectrum" of colors we see.
The M & L receptors are also sensitive to blue light, but not as sensitive as the S receptor. If you look at the overlap between the different opsin sensitivities, you might wonder how all of this works. Yellow is a pure color. You can make green from yellow and blue, but you can't make yellow from any other colors. YET, our M receptor has maximum sensitivity in the green.
The answer is that the two receptors, M & L, don't appear to be detecting intensity of each color (green & orange), so much as the balance between the two, and which side of the overlap point (yellow) a particular color occurs on. So, instead of two intensities coming out of these two receptors, it appears that we have a composite intensity with an error signal, either high or low. The high side signal indicates Green, the low side signal indicates Red. Here we have another "pure" color, Red, which is actually the composite of two detectors.
But as I said, human perception of color is the only relevant color space. Red, and yellow are pure colors because our visual system ignores the absolute intensities and pays attention, instead, to the error signal, the deviation from yellow, which is at the overlap point of the M & L receptors.
The blue receptor appears to constitute its own system. This may be why people frequently have a difficult time agreeing on which colors are blue-green, as opposed to being either blue or green. Human perception of this color range varies widely. I suggest that this is because not every person's signal processing system mixes the Green/Orange receptor system with the Blue detection system (S) in the same way.
The fact that these receptors respond to a range of light wavelengths guarantees that there will be ambiguity in color perception between individuals, and possibly ambiguity for the same individual under different circumstances. The bottom line is that, if there is to be any "Color Space" at all, the only one that matters is the one that your visual system generates for you, and that is the Red, Yellow, Blue (RYB) space.
To a certain extent, I would say that we have 3 real color channels due to the chemistry of the eye (blue, green, orange) which are reduced to one real color channel (blue) and 3 synthetic colors (green, yellow, red). The fact that physical pigments can be mixed to produce green in no way should detract from the fact that Green perception is going to be, necessarily, synthetic due to the overlap in detection molecules.
And there you have your primer on human color vision. See Adler's "Physiology of the Eye" or any other advanced ophthalmology textbook for more details.
RGB Colorspace Fraud:
At the dawn of color displays, apparently they couldn't make a suitable long-lasting yellow phosphor, so instead they made greenish yellow, and the RGB display was born. It does a reasonable impression of yellow, but more importantly, it does a great job of displaying green (think lush vegetation regularly found in nature). So, it's a great choice.
The problems with this paradigm begin only when we force all of computer science, then technology, then science itself to start adopting this fiction. The only color space that has a chance of existing is the RYB space.
But, enter the corporate marketing departments, promoting super-accurate color display for artists, etc. They NEED the RGB color space to work. They have an investment. They invented the idea of color "gamut", which is the range of colors one can supposedly see and tailored it to their need to promote RGB. Never mind that ophthalmology has no need for their input. The RGB space must be defended at all costs as the standard in color, at least as far as the monitor manufacturers go.
Never mind that the range of color perception and the intensity of color varies from person to person, from situation to situation. All that egghead talk gets in the way of selling high end monitors, so it has to go.
And then, after the visual display companies, various printing companies buy into the scam and start advertizing the incredible color gamut of their inks, pigments, paints, etc. You can see how the lie snowballs. Once you accept the fraud of colorspace, and then the idea that we can have different gamuts, the actual mechanism of human perception of color must necessarily take a back seat to marketing.
Needless to say, there is no objective colorspace, not RGB for sure, because all color perception that counts for anything is subjective, i.e. it depends on the individual human. Why should we accept a mixture of colors (green) that triggers two or three photoreceptors as if it were a pure effect? Because we were taught to do so, and now we can only think in those terms.
CMYK Colorspace Fraud:
The CMYK color "space" is actually perpetrating multiple frauds based around printing technology. It's a marketing ploy that allows companies to sell defective printing technology.
Let's cover the "K" (black) portion of the fraud. You might be excused for thinking that black ink in a CMYK device would actually be black... but you would be mistaken. It's actually dark grey. In order to get actual black, you need to add in other colors. The advertizers for this broken technology will tell you to add extra inks (a small amount of Yellow and Cyan) to make 'rich black", but don't add "too much", or the ink won't dry.
There are a number of nuances to black ink. Some people like a nice shiny surface so that it sort of gleams under directed light. And for other purposes, this effect is rather annoying. This particular effect of specular reflection at a surface exists for all colors or all pigments or dyes, but given the extensive use of black on white writing, it becomes important for readability. Thus, we have CMYK which doesn't actually represent a single thing, because the black ink can be swapped with other black inks to produce different surface properties. And yet, it's just CMYK. It seems like there's more to this color space than the fraudsters would like to admit. The newer high end printers also admit that control over this K-black isn't quite as good as once advertized.... there are now several versions of grey dye installed on the same high end printers. I'm guessing it looks better in some way, but the fact remains that the K isn't everything we would want in black ink. It's obviously defective, but being hyped by "science".
The next part of the fraud (the major part, in my opinion) is the idea of using mixtures of colors (cyan and magenta) and expecting to be able to generate anything like the colors we see in the world or on the printed pages of properly printed books. It's simply not possible to unmix the colors by putting them on paper.
If you need to print a real red, like apple or fire engine red, you can't get it. Period. I bet you don't see that in any of the self-congratulatory promotional material describing how "vivid" the latest CMYK inks are.
The same is true for blue printing. As everyone should know by now, "blue" always prints darker than the image on the screen. The reason blue always comes out dark is because it can't come out any other way. The printer simply can't mix Cyan together with other colors and arrive at a brilliant blue color. The dye mixing process always darkens things a bit. That's not really something you want in a technology... at all.
It's also certainly not something you want to be teaching the entire computing industry. it's a massive, massive fraud.
Have you ever noticed that paper money is printed using bright primary colors?
If your CMYK printer can't print primary colors, and all your paper money in circulation uses primary colors, guess what you can't do at home? Go ahead.... guess!
There is no such thing as a conspiracy theory. The current year should be enough to show everyone that understanding the goings on of the world involves understanding government funded deception and shenanigans.
We already know that governments around the world have made alliances to protect money from being scanned, all without the knowledge of their citizens. You could hardly do it any other way, but there it is, just sitting out there for anyone to see how far governments will go without telling people what they're doing. We already know that governments around the world are recording as much of our communications as they can. We already know that there are laws on the books to prevent duplication of currency. We already know that all commercial color printers have hidden watermarks so that printouts can be tracked.
Why is it a stretch to think governments have been interfering in printing technology to prevent any easily accessible technology that can reproduce money? I've been describing the problem up until now as corporate BS, but I strongly suspect the government has stepped in to make sure that printing technology will never be good enough to do what we pay the corporations to accomplish. I mean, corporations are pig-headed, but they're not stupid enough for me to believe they haven't thought of adding the right color inks to their printers.
There Aint No Color Spaces... Period.
I was very surprised to see people hyping famous physicists as the inventors of color spaces: Isaac Newton and James Clark Maxwell, to be succinct. However, these men did no such thing. They proposed theories of color, but color spaces as a true mathematical parameterization of the subjective color experience never became a going concern until the 20th century, when large scale color reproduction became possible. Ergo, these famous icons had nothing to do with it. Isn't is amazing how every crackpot, corporate-sponsored BS was discovered by famous, well-respected scientists? Yes?
Human color perception is subjective and highly context dependent. There are many color-based optical illusions that remove any doubt, such as the Munker effect to the right. As you can see, the heart in the middle appears to change colors, just by being in proximity to another color. The RGB values on the screen don't change. So, what good are color spaces if they don't deal with the things that are obviously going on in human cognition?
In any case, my 21st century knowledge of anatomy and physiology gives me the edge when I declare that printer companies and display companies are incorrect, and that our entire industry needs a re-think about what exists and what is merely corporate propaganda!