Wednesday, April 8, 2009

Observed color


Above: Rembrandt etching, The Three Crosses,

I would like to talk a little about color in this post. I can list three different "sorts" of color. I will call them, observed color, formulaic color, and artistic or expressive color. Let me define each of these for you and discuss their uses.

The first is observed color. When I was taught to paint one of the things I needed to learn to do was "see" color. That is the ability to recognize and place on my canvas a color as it appeared in front of me. Now I don't mean the color of the object itself if I were holding it in my hand, but the perceived color from my vantage point and in a particular lighting condition. I worked long and hard to learn this and it was necessary to have an experienced teacher correct my work in order to tune my ability to arrive at the proper color. I can still hear Robert Douglas Hunter saying, I see this note here as being a little warmer than you've got it Stape, as he hit the note with uncanny accuracy. He would then turn to me and patiently ask "do you see that?"
I would compare the color he had laid on my painting with nature and the color I had put there and he was righter than I was. I could not make the color as well as Hunter, but I could agree that it WAS a certain color. We agreed on what it looked like, and I could see that Hunter was closer to it than I.

Now there may be variations in individual color perception, and in some cases they are pathologies, such as color blindness. But absent a perceptual problem I believe we see pretty much the same color.

Here comes the BUT. ......There was a school of painting led for many, many, years by Henry Hensche called the Cape school of art in Provincetown, Massachusetts. I saw Henry do a demo of a head outdoors in the mid 70's and it was an amazing experience. Hensche taught people to "see" color that was brilliant, nearly neon . I never saw it, but his students said they did. I don't think they were lying to me. They would use color to turn the form in their paintings and often used a knife to force themselves to mix the note and put it down cleanly before mixing the next one to set next to it. A number of very good painters passed through Hensches' studios and are out there working today. He died many years ago so you missed that opportunity.

I am going to throw out another qualifier here. There are limited palettes that theoretically will make any color. You get a red, a yellow and a blue. I suppose in theory that is true. However in practice you can get a whole lot more colors with a broader palette .This is the downside of a limited palette, you can approximate a lot of colors but you cant get as close as you can with more. The narrow palette does give you nice color harmony though. There are degrees of "matching" a color. If something in the landscape is redder for instance than your red, there is nothing on your palette that you can add to make your painting red enough. Here's a story to illustrate that.

It must have been close to 20 years ago I was living in Maine. It was late autumn and that was a beautiful time to paint outside there. My wife, the keeper of schedules told me that if I was going to have a piece to put before the National Academies' biennial jury for that year, I would have to make it now as the deadline was approaching.

I had been in that show once several years before and made a point of going down to New York to see it. Walking around I realized that the only way I was going to get another painting in the show was to do again what I had unwittingly done the first time. That was to make something truly weird. The jury was mostly modern guys and it looked to me that they would accept a traditional painting, but only if it had a bit of the outrageous to it. I became really sure of that as I stood before a giant painting of a dead bride. If you want to put traditional painting by a "modern" jury, ain't nothin like a dead bride.

My old friend and painting buddy Stefan Pastuhov and I knew exactly where to find strange landscape in autumn in Maine. We set up in the blueberry barrens. The blueberry barrens are about as odd a place as you could imagine. The scrubby plants cover the ground about a foot high and in the fall they turn a bright crimson color. There are a few tufts of grass and the occasional white birch but otherwise there is nothing but the red barrens, rocks and the sky. These barrens are on the tops of bare windswept hills and often cover enormous areas. Scattered about are strangely shaped rocks from the size of refrigerators to the size of small houses that were left there by retreating glaciers. It is like going painting on the moon. Unless you have actually been in a blueberry barren in the fall its hard to believe such a place really exists when you see a painting of one.

So I am set up and working away. I am using a three color palette that I fooled with for about a year. It was cadmium yellow light, cobalt blue and genuine rose madder. It cost a fortune to paint with, but I got a cool look as the cadmium and the cobalt were such clear and clean colors and rose madder has that warm glow and is transparent. Either way I just could not get the color of those bushes. I muttered and fought with it until finally Stefan came over and added alizirin and some cadmium red to my palette. I was immediately able to hit the color. Stefan explained to me that I was an idiot and he may have been right. He explained that he had worked as a carpenter and would never have tried to do a job without the proper tools .

I was able to finish the painting and it was just as weird as could be. It had a pretty good design and to make it even a little stranger it was a 26x29. Not quite square, but almost.The painting did go by the jury . It may have been the oddest painting I ever made. Although it had nothing on the dead bride.


Mike Thompson said...


Color perception is half hardware and half software with a weird mix of neural nets thrown in to confuse everything by doing some preprocessing. Color perception is affected by its neighboring color, by its neighboring values, and by the light that bathes a scene, and possibly the phases of the moon. On top of this, your brain works feverishly to color correct the scene so that you still perceive colors in harsh noonday sun and mellow setting sun sunlight.

There is a book - Blue and Yellow Don't Make Green - which has a pretty cool explanation of color mixing and endless color combination charts from a 12 pigment pallette. But from an engineering perspective, the color spectra of the 12 colors at the back of the book explain why limited palettes really are limited.

All pigments reflect a broad spectra of differing amounts of various colors, ABSORBING the rest, and our marvelous brains resolve that into a single color. (Guys like you could probably describe how much of each color is in a pigment because you have trained so long but you still nail it down to a single color in your mind.) Since pigments are subtractive, mixing pigments subtracts light from the mixture and mixtures end up darker, only reflecting light both pigments have in common. If you look at the spectra, you can also see that some combinations absorb the color you are trying to hit or add color(s) you don't want in the mixture and that is why more starting pigments give more ending options but not every option.

Some manufacturers will tout a certain pigment by saying that it cannot be duplicated by a mixture because it reflects parts of the spectrum other mixtures absorb or leaves out color(s) other mixtures contain.

There are a lot more characteristics to pigments and mediums that make them endlessly fascinating and these newfangled interference and dual color pigments open a whole new set of characteristics that old pigments don't have.

All this makes me wonder, if we ever encounter space aliens, whether or not they will perceive our great art as art or junk since so much of it involves tickling the uniquely human perceptions.

One last observation. Photographing something I have just painted is often disturbing because it looks so different than the original, colorwise. This works the other way, too. I will sometimes paint from a photograph and most of the time now I only refer to it for details - departing pretty far from the original once the picture is laid in. Later, when I go back through my photograph pile, I am often amazed at how dull the photo is compared to the painting after the painting process is forgotten and it is the painting I see every day hanging on the wall.

jeff f said...

Good post. I use a palette with at least 9 colors on it. 3 yellows, 3 reds, and 3 blues. I also have high and low chroma colors.

By the way I have read that Rembrandt was a good teacher, he's dead too, we all missed that chance...

Stapleton Kearns said...


I own a bunch of books on color,but I have not read that one. I have leafed through it. Since you recommend it I guess I will buy it. I like the old Arthur Guptill text, Color in Sketching and Rendering.. That is now unfortunately out of print. But being a dinosaur I like that.The next few posts are going to go into some different ways of dealing with color that alter the whole game entirely. Also, many of my best clients are, I believe, space aliens ........Stape

Stapleton Kearns said...

Would you comment again and tell me what those colors are? My palette is posted back in the archives if anyone is interested. I do sometimes use restricted palettes as I have said lately..Jeff has a blog it is..

he also has a nice list of other blogs on his sidebar......Stape

jeff f said...

Sorry about that.
My palette:
Cad Yellow Lt, Yellow Ocher, Cad Orange.

Cad Red lt I also use Vermilion, Indian Red or Terra Rosa, Permanent Crimson or Florentine Lake (Blue Ridge Paints) Burnt Senna a reddish earth color I find very useful. So sometimes I'll have a palette of ten colors.

Cerulean Blue, Cobalt Blue and Ultramarine Blue.

Lead and or Titanium White and Mars Black or Ivory Black.

I also have 9 values of neutral grays tubed up that are based on the Munsell gray scales. I find these to be a great way to lower the chroma.

I will also add Cad Yellow, Cad Red medium, Viridian to the palette if I want to use a larger palette of 12. I like thinking in multiple of threes, seems to work.

I also use Burnt and Raw Umbra, Green Earths from time to time.

For still life painting I will sometimes mix a controlled palette based on the local colors of what is in the set up. I then mix strings based of the middle tones of each object.

jeff f said...

I also have the Carslon book which is a great book. I wish there were more color repoductions.

The other book I like is Edgar Payne's book on composition.

The New Munsell Student color book is an excellent book on how we perceive color and the best thing I like about Munsell is how he thinks about color in three dimensions; hue, value and chroma.

Principles of Color by Faber Birren is also good, a lot of historical information starting with Newton and working up to the 20 century.

Mike Thompson said...

Gamblin has a short DVD that describes his color space 3 dimentionally and how he recommends moving through it to match color. He offers Chromatic Black (a red that complements a green) to darken the color mixture to change the value without tweaking the hue. Since I haven't ever painted so meticulously as to do this (add black to darken), what is different about chromatic black as opposed to say ivory black, lamp black, or another black? The several watercolor neutral tints that I have looked at the composition data also use red/green complements. I know some blacks interact with yellow to make green. The little I played with this, I got the most dramatic changes with a cadmium yellow and lamp black.

Stapleton Kearns said...

That seems like a pretty useful palette. Nothing strange there and it covers all the bases real well.
I wonder how much longer any of us will have viridian as it seems to grow ever more costly and of lessening quality. Cerulean is getting costly too. I use cobalt violet,thats an expensive but wonderful color. I have and know the Faber Birren book and the Itten books also. I can't say that either of them lit me up.I intend to do a book report on the Payne book soon.There's a great book....Stape

Stapleton Kearns said...

I think any formula method of forming shadow color using a black is a problem. The short answer is that is inferior to mixing in the compliment. However that is inferior to observing the actual event in nature. Finally that is inferior to choosing a shadow color that looks cool.That last point is where I am going to go with this whole series of posts.

jeff f said...

I never use black for a shadow by itself.

I sometimes mix a cobalt blue or ultramarine into the the black and add a red like a crimson which alters the black enough to use.

I also like mixing Burnt Senna and Ultramarine Blue for a shadow value when sketching things out.

I have a tube of Cobalt Violet which is very nice.

Your right about Viridian. I have tube of Vasari which is very expensive but it's tinting strength is the best. Compared to Gamblin it's almost a different color.

I also recently discovered this company called Blue Ridge paints who make very good paint for a decent price. I use their Cobalt Violet. I also have a color that they make called Emerald Green which is very intense. They also have a decent Viridian.

Stapleton Kearns said...


In the 19th century there was a color called emerald green it was copper aceto-arsenite. It was fabulously poisonous and was also used as an insecticide called Paris green. Your emerald green is probably made from emeralds or maybe pthalo. I am guessing the latter. I have been fooling with pthalo green to get away from the viridian problem. I think I can make it work. I don.t just hate it......Stape

jeff f said...

Yeah I think it's a pthalo green as well. Real Emerald as you said was poisonous.

I find pthalo green way to intense and I have to mess with it a lot to get it to work. I do mix it with a lot of white and it does work well for the green haze you see at dusk close to the horizon.