Which colors does a painter use most frequently?

When a child draws a landscape, she or he paints the sun bright yellow, the sky bright blue, the house’s roof bright red, the tree-trunks bright brown and the leaves bright dark green, the grass bright light green and the flowers bright magenta, bright violet, etc. The result is a childish picture, because all the colors are too saturated (see figure #101 as an example).

Figure #102 shows the same drawing, but with less saturated colors (i.e. colored grays in place of bright colors). The impression is less childish. But the result is rather dull and drab: there isn’t enough contrast between the different colors of the picture.

#101 #102 #103 #104 #105 #106 #107

In figure #105, I have repainted the same drawing taking into account both issues at the same time: saturation and values. I’ve done it with less saturated colors and with more contrast between the lights and the darks. The result is decidedly better.

The proof is figure #106, which represents what would be a black and white photograph of the last image. However, the general impression of such a B&W image is too dull, sad and gloomy — because the colored areas are flat. That’s why figure #107 shows the same monochromatic picture, but treated in a sepia version, which is much better aesthetically.

This demonstrates that for a good painting the values of the colors — i.e. the contrast between the light and the dark tones — are more important than the colors themselves.

Figure #105 could be obtained by mixing pure colors (= bright pigments), not with neutral grays, but with earth pigments like yellow ochre, burnt sienna, and burnt umber. The upshot is rather good because the colors are warmer than the ones which could be obtained by mixing pure colors with neutral grays and because the values are good. This last picture # 105 could be used for a book destinated to children, because however simple it is aesthetically acceptable.

Untill now, my examples were painted with blocked in areas. In reality, nobody paints like that. Furthermore, when an area is painted with several hues of the same color, a sort of competitiveness between all theses hues makes that they seem much brighter than they actually are. And the result is yet better — see figure #108. A black and white photograph of this last image (figure #109) shows that the values (dark and light tones) are correct.

#108 #109 #110

The last image (fig. #110) is the same as fig. #108, but with much more saturated colors. The upshot is a very “sunny” picture, perhaps even too bright. But in a few moments we’ll see that it is very useful for better understanding the best possible use of colors in an oil painting.

#111

For example, fig #111 shows two red circles. Both are of the same color (215-12-24), but the left one looks brighter than the right one, which seems brownish. The green square makes the red circle look brighter, but the red one shows that the color of the circle actually is a dull red.

That’s why the most utilized colors for oil paintings aren’t the brighter ones, but the attenuated colors like earth colors, mixed with white, dark blues and dark greens, like phthalo blue, ultramarine, viridian and phthalo green. It’s only for painting the blue sky, flowers or butterflies — or for some other particuliar intention of the painter  ¹ — that the bright colors become necessary.

#111 bis

Figure #112 shows a comparison between the first figure #101 and the last one #110, which demonstrates clearly that the colors of the fist image are oversaturated in a nonnatural manner. Perhaps you will consider that the right image is too nonsaturated? Let’s thus compare the flat nonsaturated figure #105 (with blocked in areas) and the last one. It’s figure #113. The saturation is approximately identical in both images — maybe this is better seen on the thumbnail than on the bigger image, particularly for the tree leaves —, but the second one seems to have much brighter colors. This is due to the fact that no color is a flat one: there are no blocked in areas left. This is a additional evidence that two closely related hues of the same color look brighter when they are beside eachother.

#112 #113 #114

The last image (fig. 114) compares the first and the tenth ones, i.e. the two most saturated drawings. This is a lesson to be learned: in fact, the last drawing in less saturated than the first, and yet its colors seem brighter! The general impression is the one of an image with rather bright colors, but nearly none of its colors is really bright — except the blue sky.

This explains why the Masters of the Middle Ages and the beginning of the Renaissance could make pictures of which the colors seem to have remained very bright while most pigments they used have faded or discolored.

For example they considered the best way of painting was to begin with a nearly monochrome underpainting, preferably with earths and white than with neutral grays — something like figure #107, except for the red draperies, of which the underpainting was made with bright red. (See Methods of the Old Masters.) Correct values were here the main purpose of the underpainting.

After that, the definitive picture could begin. If you try this way of working, you’ll immediatly see that not very bright colors, put on a brownish underpainting, will look much brighter than they actually are.

Conclusion

For landscapes, a bright yellow (e.g. Cadmium Yellow Light) is needed for making greens by mixing it with Viridian (or Phthalo Green), the blues or some black (for Olive Greens); but here too, the earths will be necessary to avoid making too bright colors.

(Back to An Example)

Copyright © 2001 Sea Ooh See All Rights Reserved