Wednesday, February 24, 2010

The Color Wheel, Part 4

There are a few problems with the traditional artist’s color wheel, and its concept of primary, secondary and tertiary colors.

First of all, no color from the original spectrum has any higher claim to be a primary color than any other. Each hue occupies an equally legitimate place on the outer rim of the hue circle and can claim full status as a primary color. Nor are any particular hues by their nature secondary colors. Green is not a composite color any more than blue is.

You could set up a palette with high-chroma orange, violet, and green as primaries and paint a satisfactory image from them. It’s a good painting exercise to do so, and it can result in a perfectly acceptable painting.*

Secondly, it turns out that the traditional YRB wheel is out of proportion, like a clock face with some of the numbers bunched up in one corner (see center of wheel). It expands the yellow-orange-red section of the spectrum too much, so that red is at 4 o’clock instead of 2, and blue is at 8 o’clock instead of 6.

This uneven distribution came about partly because our eyes are more sensitive to small differences among the yellow, orange, and red hues, and partly because pigments are more numerous for warm colors, compared to cool ones. The precious pigments Vermilion and Ultramarine became our mental image for red and blue. There have always been many available pigments for the oranges and reds, but few for the violets and greens.

So, are primaries all relative? Can we set up the color circle in a different way? The answer is an emphatic yes. Tomorrow we’ll look at the Munsell system, which has served as the color map for many great realist painters.
-------
*In the case of mixing colors from OVG secondaries, you’ll have a hard time mixing a pure yellow, because yellow is a special case: it’s purest form is much lighter than the pure form of other colors, so it isn’t easy to mix yellow as a secondary in pigments. But the OVG colors have been used as primaries for the autochrome photo process.

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

19 comments:

etc, etc said...
This comment has been removed by the author.
Jeremy Elder said...

Your excellent color mapping posts are proof of what is possible with primaries other than RYB. Maybe you can't mix a pure yellow from OVG, but as you have pointed out in the past, everything mixed from those colors will be in harmony.

I am looking forward to what you have to say about the Munsell system

Roberto said...

So we are trying to create a coherent and accurate map for understanding the medium of Light, it’s effects on our brain/eye, and our response to those effects; and we want our map to guide us as we attempt to reproduce those effects with pigments, dyes, inks and photons … they also say ‘there is a pot of gold at the end of the rainbow.’ It may be folly to try and find it, but I’m up for the Journey! I'm looking forward to your insights, and those of your blogosphere. -RQ

Tyler J said...

I am enjoying this series very much. I look forward to the rest of it.

Might you be planning on tackling additive vs. subtractive?

Thanks again for the knowledge and inspiration. My fires of curiosity are constantly stoked here.

Roberto, nice analogy =)

René PleinAir. said...

This is getting more and more interesting. lots to think over again. Thanks for these nice postings.

nemecsek said...

Thank you for your very interesting posts!
I'll stay tuned for the following "lessons".

jokergirl@wererabbits said...

...This uneven distribution came about partly because our eyes are more sensitive to small differences among the yellow, orange, and red hues...

That's not quite correct. If we are going by physical breadth of spectrum, the human eye is actually most sensitive in the green range. If we're going by psychological perception, it is quite culture dependent, with some cultures putting higher value in differentiating hues of green, and some of red.
Culture dependent colour groups and colour perception is an entirely differen chapter in itself, though...

Don Cox said...

I think the uneven distribution comes from the fact that in the 18C (when the idea of "primary" colours for printing first appeared), red, yellow and blue were the only high chroma pigments available. Magenta was invented in the 19C and cyan in the 1930s.

In my opinion, primary colours are completely irrelevant when you are using traditional media (paints, inks, pastels). They are needed only for colour photography, printing and digital graphics.

D said...

In a certain limited sense, magenta, yellow, and cyan are the "best" primary colors because they are the lightest and farthest spread apart in LAB color space, so you can make the widest range of colors from them. But you could cover an even larger area of possible colors by using four or more primaries.

Richard said...

I am enjoying all of the color discussions. I don't know if it has already been posted, but here are some color system overviews I found by Adobe.

http://dba.med.sc.edu/price/irf/Adobe_tg/models/main.html

RM said...

The clock spacing analogy seems to space the pure colors by frequency. If wavelength is used the spacing comes out closer to the traditional color wheel.

etc, etc said...

If the traditional YRB wheel is out of proportion, then the Munsell wheel is even farther out of proportion. The Munsell wheel contracts orange and expands blue-green, converse to our perceptions.

jeff said...

etc made some comments on Munsell which do not seem to hold true.

Munsell sought to build his model on equal perceived color differences on each color making attribute, linked in some cases to measurable changes in color stimulus composition, in a freely branching geometry called a color tree.

Geometry of the Munsell System. The backbone of the Munsell system, the "trunk" of the color tree, is a vertical dimension of lightness or value, which was developed through human perceptual judgments of equal differences in lightness across visual (color top) mixtures of white and black paints. The Munsell value scale ranges from pure black (0) to pure white (10); each step is divided into decimal increments, resulting in a 100 step lightness scale.

OK I know James you're going to cover this next. So I'll stop.

By the way this is information is from the web site handprint.com
which is an excellent site for this subject and a historical breakdown of how color wheels and theory developed through the centuries.

etc, etc said...

Jeff,
I am trying to be concise. Munsell did indeed build his model on equal perceived color differences. He did make an exception, however. The Munsell "Hue circle" has 10 basic hues as opposed to the standard 12 hues. That's because ease of calculations (far easier in a 10 unit system) was Munsell's priority, not equal perceived color differences.

jeff said...

etc. Have you studied Munsell?
He developed this in 1906 and he did not use ten because it was easy to calculate on a calculator.

Well that's the point of Munsell is it not? To show the real space each hue has in relation to it's chroma.

It's the first modern color wheel.

etc, etc said...

Jeff,
I am not referring to calculators. Here is a quote from "The New Munsell Student Color Set" 2ed:

"Even finer distinctions can be made between similar hues through the use of decimals. This is why Munsell chose 10 hues as the basis for the system rather than the 12 hues often used in artists' color wheels. A system based on tens can be further divided into decimals, making it possible to designate colors with precision."

Allen said...

As always Jim, your comments are interesting and instructive. I am interested to see your comments about Munsell. I am vaguely familiar with it but have not given that much attention to it because for 30 years the traditional color wheel and system has served me well. Also, the Munsell system seems rather complicated as opposed to the tradional system where, with a warm and cool yellow, w & c blue and a w & c red I get every color, tint, shade, intensity, neutral, gray, compliment, etc that I want. At least I thought so until reading what you have said here. So I am interested now to learn a bit more about Munsell.

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

another thought: Your idea of calling orange, purple and green primary colors sounds wonderfully heretical. Polifka, I think, would skewer you. I totally agree that it is a triad that creates wonderful colors and harmonies, as long as you don't need a saturated red, blue or yellow