Monday, February 22, 2010

The Color Wheel, Part 2

Artists generally regard red, yellow, and blue as the most basic, or primary, colors. If you ask most artists to select three tubes of paint to match their mental image of the primary colors, they will most likely pick something like cadmium red, cadmium yellow, and ultramarine blue.

Why those three colors? From Greek and Roman times to the Renaissance, most people thought green should be included as a primary, too. As we’ve noted on a previous post, green actually has more psychological salience than yellow. It’s mentioned much more often in the English language.

What is a primary color? The idea is that you should be possible to mix every other color out of the three primaries. You may have noticed that with the traditional artist’s primaries you can mix clear oranges, but the greens and violets are on the dull side.

The traditional artist’s color wheel,” above, presents yellow, red, and blue spaced at even thirds around the circle, in the position of 12 o’ clock, 4 o’ clock, and 8 o’clock.

Mixtures of the red, blue, and yellow primaries create secondaries. The secondary colors are violet, green and orange. They appear at 2, 6, and 10 o’clock on the traditional color wheel. But in truth when I painted this wheel, I didn't paint those secondaries from the primaries. If I had tried, they would have come out much duller.

As we’ll see in future posts, there’s nothing written in stone about any of those colors being primary or secondary, and we can take a fresh look at the whole arrangement.

Reviewing the posts in this series:
Part 1: Wrapping the Spectrum
Part 2: Primaries and Secondaries
Part 3: Complements, Afterimages, and Chroma
Part 4: Problems with the Traditional Wheel
Part 5: The Munsell System
Part 6: Cyan, Magenta, and Yellow
Part 7: The Yurmby Wheel


D said...

With computer printers, the primary colors are cyan, magenta, and yellow. These are a lot brighter than the colors you mention here, so mixtures of them wouldn't be so dull, I would think. Maybe this will be in part three.

My Pen Name said...

cad yellow light and cad red light make a pretty vibrant orange. There is a book 'making color sing" J. Dobbie - it's more about water colors but she says to make vibrant colors to mix an opaque with a transparent but never two of the same - I have always found, in watercolors the mixed from primary greens much more vibrant that the tube greens.

also it could be pretty tricky to mix the secondaries with the same blue and same yellows - maybe a cooler blue for the violet and a warmer for the green? just guessing at this point :)

Juanma P. R. said...

very interesting about the secondary colors and green.
also thx "my pen name" for the book recomendation.

Unknown said...

What about Hue, Value, Chroma as in Albert Munsell's therory
on color? I have been reading
"The New Munsell Student Color Set"
and find it interesting. I would
like to get your thoughts. The guys at the "rational painiting" site swear by Munsell and they
mention your site time to time.

David Still said...

The website has a lot to say on the subject as well. I't vary academic though, or as David Briggs says himself: Colour made difficult.

Unknown said...
This comment has been removed by the author.
Unknown said...

As those in the printing industry would tell you, cyan, yellow and magenta are the primary colors of pigment or "subtractive color" (meaning you subtract colors to get white). Light on the other hand behaves in just the opposite fashion. Adding all the colors of light together gives you white. (We all know what happens when you add all the colors on your palette together, you get dark grey or brown, or maybe even black, but definitely not white.) With light the primary colors are red, green and blue.

So we can say that
* red is the opposite of cyan
* yellow is the opposite of blue
* and green is the opposite of magenta.

Super Villain said...

this is an awesome post, i especially like what d and walter have to say about printer cartragie, and how they are able to make a massive specturm of colors with just cyan magenta and yellow,

i have never thought of that before!

Gregory Becker said...

This is a subject of constant consideration for me. There is something mysterious about color and how it responds to light that will always prove fascinating.
I love reading your color posts.

Alex said...

You are making a very fundamental mistake with this color series. You are focusing exclusively on the properties of light, and ignoring the most important aspect of how we perceive color -- the human eye.

Cyan, magenta, and yellow are primary colors not because of their relation to an arbitrary set of pleasing shades, but because they correspond to three kinds of receptors in the eye. This is where color theory must start.

There is a broad spectrum of single-wavelength visible spectrum of light. We perceive this as a rainbow of all the colors. What really happens is that our three kinds of cones create overlapping pulses.

Violet wavelengths are more extreme than either red or blue. But the sensitivities of our cones are such that we don't have anything the responds directly to it, though the magenta and cyan cones each have a little sensitivity in that range. This is why mixing colors can simulate shifting wavelengths.

It's also why there are three colors that look unrelated to each other. That's the limit of unique signals we can perceive. Other animals have other numbers of receptors. Many fish, for instance, have a fourth, purplish one. For such animals there are four primary colors. Purple is no longer a reddish blue or bluish red. It's as unique as red is to green.

Such organisms also have 6 secondary colors, mixtures of two primaries. We only can make shades of 3. They also perceive another 4 complementary colors, colors missing one of the primaries -- these being entirely distinct from the secondary colors.

Such an organism would draw not a triangle-based color-wheel but a tetragonal color-sphere. Each corner would be a primary color. Each edge a secondary (with the red-blue edge being nearly unrelated to pure purple), and each face a complement.

Always remember that color is biology, not physics.

jeff said...

Can I suggest a look at the Munsell color wheel. It has 10 hues and is based on the three dimensions of color, hue, value and chroma.

Here are few links on Munsell:

This site is amazing, David Briggs is the man:

This site is great as well:

Christopher Thornock said...

I am sure you are going to cover this, but what I find most troubling for students or for us that teach painting is the following. The scientific aspects of color (color wheel, etc.) are often difficult to translate for students because you do not buy spectrum colors in tubes. As a student at Art Center, I sat in painting classes with instructors who would only relate to color temperature when talking about how to fix my paintings. (ie., you need to warm up that red) As a student, color theory class did not correspond to actual painting practice. I made a great resource discovery. The book "How to see color and paint it", by Arthur Stern. Now out of print. The simple projects and discussions on how to translate the ideas of spectrum color (your color wheel) to tubes of paint (cadmium, thalo, etc.) has been an invaluable tool. I recommend looking it up, especially if you are inclined towards a naturalist style.
I am looking forward to your book coming out this fall on this very subject.

James Gurney said...

Thanks, everybody.
Jeff: I'll take a look at the Munsell wheel on a future post.

Alex: I agree with you up to a point. But: color is not all perceptual; there are mathematical and physical realities beyond the effects on our senses (wavelengths of light being one example). The opponent process model of perception doesn't map exactly to cyan/magenta/yellow as you suggest, it's a bit more complex than that. And I think it's a bit difficult to know exactly how a non-human tetrachromat perceives light and color, though it's an intriguing possibility.

Walter and everyone else: I'll get to cyan, magenta, and yellow soon, I promise.

And I second your recommendation of and for a very thorough analysis of all these topics. I'm trying to boil down some of that material, knowing that simplifying sometimes distorts reality a bit, and I'm really interested in the practical realities for working artists.

David Still said...

I also think that while the stuff discussed at huevaluechroma and other resources are very interesting, it's more wise to, as you say, boil those things down and focus on color/pigment theory for the artist. At least that's what I'm the most interested in!

Casey Klahn said...

I wholeheartedly support your approach, James. As a painter, and not a scientist, computer guy or printmaker, I rely on pigments and wish to mix colors. The devil with the CRT!

Well, to be more modest, I am fascinated by the optical qualities of color and light. I enjoy reading the differences between so-called additive and subtractive systems. I love the fact that there are competing color wheels.

In point of fact (maybe you say this later) the range of hues available via pigment is greater than the range of hues available via printing inks and (I believe) the CRT as well.

I have a cousin who is a printer engineer for a major computer firm, and he and I have a great time talking color. He's taught me a lot, as have posts like this.

The dominant paradigm nowadays is the computer, and so it is funny to see remarks about how the a traditional artist's use of the color wheel and paint is "wrong," when in fact it makes good sense of how we mix pigments. I am finding, in my own studio, that the theories of mixing are theories, and yet the elements/pigments that create color behave the way they want to. But, the "additive" theories don't solve this any more than the color wheel does.

Blog on!