Skip to content


More on color

In my last post about color, I dis­cussed the inad­e­qua­cies of the stan­dard color wheel and explained why we’re going to have to replace it with two things: (1) a more accu­rate way to describe color; and (2) a sys­tem for approx­i­mat­ing real-world paint mix­ing. In this post, I’ll talk about describ­ing color. As I do so, I’ll refer you to cer­tain sec­tions of the Hand­print web site (from which I have stolen shame­lessly) in case you want more detail.

Although we often still see the stan­dard three-primary color wheel in books about paint­ing and color mix­ing, it really went out of date in the late 19th cen­tury, when guys like Ogden Rood demon­strated that it pretty much stinks for describ­ing color accu­rately. There is noth­ing in the way humans per­ceive light to sup­port the idea of three unmix­able pri­mary col­ors (red, yel­low, blue), each of which is com­ple­men­tary to a spe­cific mix­able sec­ondary color (red and green, yel­low and vio­let, blue and orange). In fact, it makes sense to me that there are no spe­cial pri­mary col­ors at all, whether the tra­di­tional artist’s pri­maries (red, yel­low, blue), the printer’s pri­maries (cyan, magenta, yel­low), or any­thing else.

A num­ber of more accu­rate ways of describ­ing color have been devel­oped. Many of them are designed pri­mar­ily to sup­port the needs of the print indus­try, the dye indus­try, man­u­fac­tur­ers of video equip­ment, and other com­mer­cial ven­tures. They are need­lessly com­plex for our pur­poses. The best sys­tem that is com­pre­hen­sive enough, but not too com­plex to be eas­ily under­stood, is the Mun­sell color sys­tem. It was first devel­oped in the early part of the 20th cen­tury and has been updated a few times since then, although the orig­i­nal struc­ture remains. Any such sys­tem rep­re­sents a series of com­pro­mises, so there are ways in which Mun­sell is imper­fect, but over­all it suits our pur­poses bet­ter than any other that I am aware of.

Rather than a color wheel, Mun­sell is built around a three-dimensional color space. This space takes the shape of an irreg­u­lar cylin­der. Mun­sell uses three prop­er­ties of color: hue, chroma, and value. I described those prop­er­ties in detail in a pre­vi­ous post.

Value

Run­ning up the cen­ter axis of the cylin­der is the prop­erty of value. At the bot­tom of the cylin­der is value 0 (pure black); at the top is value 10 (pure white). So, for exam­ple, a value 6.5 gray is fairly light, while a value 1.0 gray is almost black.

Hue

Run­ning around the out­side of the cylin­der is a hue cir­cle. It is defined by five prin­ci­pal col­ors (there are no pri­maries in Mun­sell). These col­ors are red ®, yel­low (Y), green (G), blue (B), and pur­ple ℗. These are gen­er­ally rep­re­sented in clock­wise order, start­ing with yel­low at the top. The five prin­ci­ple hues have five inter­me­di­ate hues in between them: yel­low red (YR), green yel­low (GY), blue green (BG), pur­ple blue (PB), and red pur­ple (RP). Within each of the hues are ten sub­di­vi­sions, with 5 at the cen­ter. So 5BG is a pure blue green, while 2BG is more blue and 9.3BG is more green.

Each of the prin­ci­pal hues has a visual com­ple­ment that is the inter­me­di­ate hue directly across from it on the cir­cle. So the com­ple­ment of red is blue green, the com­ple­ment of yel­low is pur­ple blue, the com­ple­ment of green is red pur­ple, the com­ple­ment of blue is yel­low red, and the com­ple­ment of pur­ple is green yel­low. These com­ple­ments cor­re­spond (approx­i­mately) to how humans see color. Mun­sell com­ple­ments are rea­son­ably close to actual data on after­im­ages. If you stare at a spot of green for a long time, and then look at a neu­tral gray sur­face, most peo­ple who are not color blind report that they see an after­im­age within Munsell’s red pur­ple range. The same goes for each of the other com­ple­men­tary pairs on the hue cir­cle. You see these after­im­ages because of the way that cone cells on the retina work. I’m not going to describe the phys­i­ol­ogy, but it helps to know that these are real phe­nom­ena (which I am over­sim­pli­fy­ing dras­ti­cally here), not arbi­trary or aes­thetic conventions.

Chroma

If the value para­me­ter is a line down the cen­ter of the cylin­der, then the chroma para­me­ter radi­ates out­ward from that cen­ter line to the edges of the cylin­der. Zero chroma (gray/black/white) is at the cen­ter. Mov­ing out­ward are increas­ingly chro­matic (intense) col­ors. So a value 6 yel­low at chroma 1.5 is basi­cally a warm grey (not very chro­matic), while a value 6 yel­low at chroma 15 is very intense.

There is no arbi­trary max­i­mum chroma, so the chroma scale for each hue runs from 0 to how­ever intense that hue can get. As new, brighter pig­ments are devel­oped, they are sim­ply placed at higher chroma lev­els than those of older pig­ments. Any pig­ment can there­fore be placed upon the Mun­sell color tree. Because of the physics of light and the nature of color vision, the max­i­mum pos­si­ble chroma is dif­fer­ent for dif­fer­ent hues. For exam­ple, the max­i­mum pos­si­ble chroma of a light-valued yel­low is much higher than that of a light-valued pur­ple. Max­i­mum chroma for a given hue is also dif­fer­ent depend­ing on value. So the Mun­sell color space is a bumpy, uneven cylin­der (when Mun­sell first invented this sys­tem, his real­iza­tion that the color space couldn’t be sym­met­ri­cal was a big improve­ment over pre­vi­ous sys­tems that had tried to cram a messy real­ity into an ide­al­ized cir­cle or triangle).

Color nota­tion in Munsell

Col­ors are named in Mun­sell in the stan­dard nota­tion of hue value/chroma. So ver­mil­ion is noted as 8.5R 5.5/12. That means that, within the hue of red, it is at posi­tion 8.5 (closer to yel­low red than a pure red), with a value of 5.5 (right in the mid­dle) and a chroma of 12 (fairly intense). Some paint man­u­fac­tur­ers, such as Liq­ui­tex, put these num­bers on every tube of paint. Unfor­tu­nately, that’s rare.

The Mun­sell sys­tem has been updated sev­eral times to make it more tech­ni­cally accu­rate, but none of those updates is sig­nif­i­cant for our pur­poses. You can buy color sets from the Mun­sell com­pany. They con­sist of a book describ­ing the sys­tem, a bunch of color chips (sets have either glossy chips or matte chips), and pages with lit­tle pock­ets that the color chips fit into. The idea is that you learn the sys­tem by fit­ting each chip into its appro­pri­ate pocket. I haven’t bought a color study set, not because I’m unin­ter­ested but because they cost hun­dreds of dol­lars. Get­ting a set and plac­ing all of the chips would not be a waste of time for a seri­ous stu­dent of painting.

That’s Mun­sell. Boy, that was a lot of expla­na­tion, even though I picked the sim­plest use­ful color sys­tem that I know of and avoided extra­ne­ous detail. Color is really complicated.

What’s it good for?

So how is Mun­sell more use­ful to a painter than the old three-primary color wheel? First, it dis­penses with the con­fus­ing idea of pri­maries and sec­on­daries while more accu­rately iden­ti­fy­ing use­ful com­ple­men­tary color rela­tion­ships. Sec­ond, as you become more famil­iar with Mun­sell, you can begin to think about col­ors in terms of how they relate to each other within the color space. If you are look­ing at a blue wall, for exam­ple, and you are think­ing in Mun­sell terms, you can fig­ure out where the color lies and how to accu­rately describe it. What is its hue? How chro­matic is it? What value is it? How do those para­me­ters com­pare to other col­ors you are try­ing to work with? How do the hue, chroma, and value of the wall relate to the hue, chroma, and value of the blob of paint you are try­ing to use to rep­re­sent it? Some artists pre-mix a set of col­ors on their palette in Mun­sell value steps. Sev­eral com­pa­nies sell paints that are graded accord­ing to Mun­sell; Stu­dio prod­ucts sells a set of neu­tral grays and another set of greens, all the same hue and chroma, of dif­fer­ent Mun­sell val­ues. These are par­tic­u­larly use­ful for underpainting.

Lots more on color in future posts.

Posted in art technique, color.

Tagged with , .


5 Responses

Stay in touch with the conversation, subscribe to the RSS feed for comments on this post.

  1. painterdog says

    this is very inter­est­ing but way to complicated.

    A gray value sys­tem with going from white to black should be sim­p­ley based on color values.

    That is if I have a range from cad yel­low light, cad yel­low med, cad orange(middle value for flesh tone, or low­est tone before shadow value),yellow ochre, cad red lt, cad red med, cad red dark, raw umber, burnt unber, cobalt blue, ultra-marine blue, aliziran crimz­ian, black.

    If you mix gary val­ues to these col­ors you will have a very good con­trol and an easy way of mix­ing hues from all the col­ors, giv­ing you a way of under­stand­ing value and color relationships.

  2. David says

    Jef­frey,

    Thanks for the com­ment. I’d like to empha­size that Mun­sell is a sys­tem for describ­ing color. It isn’t a good model for mix­ing pig­ments. I’ll get into color mix­ing, includ­ing (as you sug­gest) strate­gies based on pig­ment val­ues and adjust­ing chroma with grays of equal value, as I con­tinue this series of posts.

  3. Amanda says

    Hi there,

    I’ve been spend­ing a lot of time study­ing colour the­ory. Includ­ing Mun­sell — work­ing through the Stu­dent Set and paint­ing grey cubes. They are great exer­cises for get­ting hue, value and chroma thor­oughly understood.

    Beyond that Ost­wald (a con­tem­po­rary of Mun­sell) is also has some inter­est­ing the­o­ries and points to prac­ti­cal appli­ca­tion by the some of the old masters.

    I achieved my colour mix­ing prowess by mix­ing all 2500 colours sug­gested in “Blue and Yel­low don’t make Green”. It was no good just read­ing it.

    Recently though I came across a thread at WC, of all places, that really made the red, yel­low, blue and CYM issue come together for me. You need to keep an open mind while you fol­low his exam­ples. None of it negates any of the other the­o­ries — just makes it make sense.

    Try it: http://​www​.wet​can​vas​.com/​f​o​r​u​m​s​/​s​h​o​w​t​h​r​e​a​d​.​p​h​p​?​t​=​330741

    Amanda

  4. David says

    Amanda,

    Bill (the guy who wrote the post you refer to) is a smart guy, and excel­lent painter, and a real gen­tle­man. He was sup­port­ive to me in pri­vate emails dur­ing a time when I was get­ting a lot of hurt­ful flak on an inter­net forum.

    He is, if I recall cor­rectly, a retired pro­fes­sional in the print indus­try. Per­son­ally, I’m not too inter­ested in color the­o­ries that come from print­ers spray­ing tiny droplets of a few col­ors of ink in pat­terns designed to trick the eye into per­ceiv­ing a wide range of col­ors. I don’t find prin­ci­ples of ink halftone print­ing par­tic­u­larly applic­a­ble to mix­ing paint.

    If you found it illu­mi­nat­ing, how­ever (no pun intended), then that’s great.

    I def­i­nitely found “Blue and Yel­low Don’t Make Green,” aside from its irri­tat­ing writ­ing style, to cre­ate more con­fu­sion than it solved.

Continuing the Discussion

  1. Amanda Williams - Portfolio » Blog Archive » Colour theory unscrambled linked to this post on 18 July 2007

    […] just fin­ished leav­ing a com­ment over at All the Strange Hours in response to a great arti­cle on Mun­sell and colour the­ory in gen­eral. Then realised I could […]



Some HTML is OK

or, reply to this post via trackback.