I knew when I saw the title I was in trouble. I've got stuff to do tonight, and don't have time to write a real response. And a real response isn't even really

I knew when I saw the title I was in trouble. I've got stuff to do tonight, and don't have time to write a real response. And a real response isn't even really necessary; there's nothing wrong with the all the various color theories and models that paper the floor of the world. They are all imperfect, but often useful, approximations generally designed by trial and error usually for specific use cases. So the notion that those wacky, touchy feely artists (aren't they cute!) don't know jack about color is specious. Their knowledge is specific to their domain and tools they use. While many of them could use some touch up here and there, it's really not necessary because ... dramatic pause ... It doesn't matter! No artist is trying to produce a particular color found in nature. They are trying to produce the appearance of a perceived color of a certain local color source under certain lighting conditions, and, here's the kicker, under certain psychological conditions. A good artist can make you think you see any color they want to using nearly any available palette. The more colors the better, up to a point; limited palettes can be a very powerful tool. And the color interactions you get when mixing pigment paint is not just an optical problem, it's a chemical problems as well. Artists do some very odd things to get the colors they need. I'm reminded of a story about the MIT hackers back in the 50s/60s who found out that a radio produced interesting sounds depending on what the computer next to it was doing. The electronics would produce radio noise that would be played as sound. With a little experimentation the hackers found out what kind of operations produced what kinds of noises, and at what frequencies. So they wrote one of the first computer music programs using this system. Needless to say, they needed to make the computers do some pretty odd and complicated things to get the radio "noise" they needed. The chemicals we use for pigments are the ones that "happen" to reflect, re-emit, or transmit light at particular frequencies. Finding the colors is hard enough, but you also have to worry about how they interact chemically with each other, for both stability (boom!) light fastness, or toxicity. In the early days, toxicity wasn't too big a considerations, and many, many artists died from rather nasty chemical poisoning. If you are interested, get a book on oil painting. Creative Illustration by Andrew Loomis is a pretty good place to start. I also highly recommend John Gurney's books on painting or artistic color usage. These are good things to read even if you have no intention of every picking up a brush, because they are not at all about the physics of color. They are about using color to communicate and create a certain kind of perceptual experience. If optics and physics were what was needed to understand the mechanics of picture and image making, anyone with a good camera could do the job. But a good photographer knows that their job is not to perfectly transcribe whatever may be happening in front their lens at any old arbitrary time. (As an alternative, you can learn a lot about practical color theory by studying older photography books.) As for CMYK and RGB color spaces, these are simple but quite effective tools for dealing with color in a purely mechanical way. And they dovetail rather nicely with the artistic color models used in the past by painters and photographers. It is important to know that these are abstract color models intended to produce an approximation that is almost always good enough for most everyday usage by non-specialists. Specialists use tools that are often similar, but with more special cases and provisions for calibration and feed back situations. If you really need a color model more closely tied to how human visual physiology works, LAB was designed as a good approximation of that. Specialists sometimes use other, often proprietary systems. If you need this level of knowledge, you got some book larnin' to do. No blog posts or comments on HN is going to help you too much. Finally, speaking of human physiology, the model of filters he describes seems to make it impossible to distinguish red from green. In other words it seems to be describing a system that exhibits red/green color blindness. I have a suspicion that the author read a complicated article, got hung up on the part that talked about r/g color blindness, and thought that was the general case. When I've talked to specialists about this kind of thing, the answers I get are all the equivalent to "It's complicated ..." Human visual physiology is real complicated, even the purely mechanical aspects of it, and frankly not a lot is known about some aspects of it. What I've been told is that when you reduce the complicated aspects of what is known into a workable set of approximations, for specific use cases, you get something very similar to ... wait for it ... Artistic color wheels and paint mixing systems.