Colors are frequencies on a limited spectrum–a line segment, not a circle. So why does a circular chart make such a good way of arranging them? Is it pure cultural conditioning that tells me violet goes next to red?
We perceive color red, green and blue sensors in our eyes. You can think of the color wheel as having the sensors 120 degrees apart on the wheel. So something that stimulates the red sensor and the blue sensor looks purple.
It’s not arbitrary. In subtractive color, (what happens when you mix pigments) the primary colors are red, blue and yellow. So you stick those at 120-degree angles on a circle. Then look at the spot between, say, red and yellow. You mix red and yellow paints together, and you get orange. So you put orange between red and yellow. Then go over one to the yellow/blue spot. You mix yellow and blue, and you get green. So green goes between yellow and blue. Finally, blue + red = violet. These are your secondary colors.
You can repeat the process to get as many intermediate colors as you like.
I hope this isn’t too much of a hijack…
Why does a rapidly rotating color wheel look “white.” I realize that a mix of all the spectrum colors is “white,” but why does this work on a rotating color wheel? Why don’t my eyes perceive that as a blend of all the colors? I.e., let’s say I mixed a bunch of paints of the colors of a color wheel…that result would be hideous.
I learned this in 7th grade. I read that a color wheel rotating produces this effect. I didn’t believe it, so I tried it. Voila.
Hmm. OK, but is there any physical justification for placing them at those positions on the wheel?
Actually, I know the reason for this. Mixing together all the frequencies of light creates white. This is called additive coloring. Mixing paints is subtractive coloring because paints work by selectively absorbing color. In subtractive coloring, mixing all the colors equally creates black because all the colors are absorbed.
Actually, you need 3 dimensions to include all colors; a typical color wheel doesn’t show shades and tints (the addition of varying degrees of black and/or white).
More or less. It’s not a consequence of the nature of light, but of your colour receptors: low-frequency light stimulates the red receptors, middle-frequency the green, high-frequency the blue, but not only do intermediate frequencies stimulate more than one set: very high frequencies stimulate the red receptors as well as blue.
No, they don’t. The wikipedia article on cone cells has a graph showing the response curves of the three receptors.
We commonly think of the colors on the color wheel as being equivalent to a spectrum, but that’s not strictly true. In a spectrum each color is produced by a single pure wavelength of light. But there are many colors on a color wheel that are impossible to achieve if you limit yourself to single wavelengths. For example, the entire band of purple-pink that bridges the gap between red and blue can only be reproduced by mixing different wavelengths of light.
A color spectrum is one-dimensional. As you vary the wavelength of the light you get different pure monochromatic colors.
But human color perception is three-dimensional. We don’t see the actual spectrum. We sample it at three different points. So your brain doesn’t know for sure WHAT wavelengths it’s seeing – all it knows is how the red, green and blue cones in your retina respond. The response of each cone can be translated to XYZ cooridinates and plotted in a 3-D cube. That cube represents the complete gamut of colors that humans can perceive. Some locations in the cube correspond to pure monochromatic colors, and others can only be achieved through a mix of various wavelengths.
(Incidentally, this is why we have three primary colors. If our eyes had four color receptors instead of three, little kids would learn about the four primary colors in kindergarten.)
The color wheel is a compression of the full 3-D color cube into 2-D. This is accomplished by changing the three coordinates you’re using. Instead of RGB, you can also locate a color in the cube by HSL – hue, saturation, and luminosity. If you take a slice through the cube at any constant luminosity level, you’ll get a 2-D image that looks like a color wheel.
Actually magenta, cyan and yellow.
It is true that magenta, cyan and yellow provide a larger gamut than red, yellow and blue but since the set of colors produced by CYMK color printing is far from complete it is really a mistake to think of them as THE primary colors. Attempts to expand the gamut available to printers include things like hexachrome which is capable of producing colors outside the gamut of CYMK.
The problem is people have incorrectly learned that primary color are somehow perfectly defined to produce all colors, when in fact many color sets can serve as primary colors. Red, yellow, blue is not a bad set of primary colors.
I want to amend my post.
After several decades of working with color professionally and “knowing” how colors work, I’ve recently begun to think that you need **FOUR **dimensions to include all colors:
The “pure” colors (RYB, CMYK, RGB, etc.) can all be shown 2-dimensionally.
You need 3 dimension to show tints (degrees of white) and shades (degrees of black).
And you need another dimension to include combinations of the “pure” colors with various shades of gray (adding degrees of white **AND **degrees of black. I don’t entirely understand this, and can’t even begin to visualize how this would work.
You *can * show tints and shades on a two -dimensional wheel.
I’m not sure I trust that graph. I recall seeing one somewhere where red sensitivity extended up close to 400 nm and showed a little secondary peak there. I’m sorry I don’t remember the source. (Might’ve been Foley, van Dam or another computer graphics book.) There’s one here that’s considerably different than the Wiki version, at least.
Anyway, the bottom line is that a normally sighted person’s eye and brain respond the same to pure violet light as they do to a mixture of red and blue light, so a color wheel ends up making sense. I suppose it didn’t necessarily have to turn out that way.
Another thing: it turns out the eye has a fourth receptor type that isn’t visual, but is used in sleep regulation. (This was in Science News a few months back.) The fourth sensor has a peak sensitivity to sky blue. My latest pet peeve has become the number of incredibly bright blue LEDs that have ended up in my bedroom. I worry that’s one reason I have trouble sleeping. More about that in another time and thread, I suspect.
I think that’s actually a three dimensional vector space, but I’m not a math whiz.
Geez, I’m surprised nobody has jumped me for this. Even a non-math-whiz like myself eventually remembered that a plane coordinate system is 2D whether it’s plotted with vectors or coordinates.
I guess all the smart people are sleeping. Now, are they smart because they’re sleeping, or are they sleeping because they’re smart? Dumb question. The answer is yes.
Sure, but you’re just sacrificing a different dimension. Where is gray? What happens if I mix purple and green?
Many years ago. when I started out learning the busiiness of boat repair and fiberglass, it became apparent that I had one major weakness, colour matching, that would be a severe inpediment towards starting my own business.
Knowing that there were three primary colours, I decided to chart various combinations of red, yellow and blue in all the permutations of ratios based on units of 16,8,4,2,1,0 with each sample represented as (x,y,z). Red is x, Blue is y, and z is yellow.
for example, red would be ((16,0,0) and blue would be (0,16,0).
I wanted a chart that would be quick to read , simple and show the relationships of the primary colours as "mathematically as possible. I hit upon the idea that I could present the relationship of two primary colours in complete continuity by placing them on the outside of an equilateral triangle with the pure primary colours on the corners. For example , starting at red towards blue, I would place (16,0,0), (16,1,0), (16,2,0),(16,4,0),(16,8,0), (16,16,0), (8,16,0),(4,16,0), (2,16,0).(1,16,0), (0,16,0) on one side.
Now that worked fine for permutations of two, but it quickly,became apparent that it wasn’t going to work very well when I wanted to represent the permutations of three colours. That’s when I hit upon the idea of a colour wheel. As has been pointed out, it is simple enough to place the primary colours at 0,120, 240 degrees. I could now place (16,16,16) say black in the centre and allowing for 5 concentric circles of samples, represent every permutation of the primary colours. Working with syringes and pigments, I made 9 colour wheels with various ratios of white.
Those colour wheels made my business what it is today. I really don’t need them any more. I now promote myself on my colour matching, all thanks to the colour wheels I made.