When light mixes, you get white; when paint mixes, you get brown: why?

Short answer

When you mix light, you increase the light reaching your eye: when red, green, and blue light overlap, you get white. When you mix paint, though, each paint subtracts one more portion from the light that falls on it; the more things a mixture absorbs, the darker it gets, and with real paints the result is often a muddy brown. There's also a third way: when colors sit side by side as very small dots or as rapidly spinning slices, the eye sees their average.

Additive mixing: light upon light

Picture three spotlights in a dark scene: one red, one green, one blue. Where the red and green circles of light overlap, you see yellow. Where green and blue intersect, cyan appears; where red and blue intersect, magenta appears. Right in the middle, where all three meet, there is white. The lights are adding to one another, and the total light reaching the eye is increasing.

Why this mixture works so well with exactly these three colors has to do with the eye's three cone types: the three chosen lights stimulate the three cone types as separately from one another as possible. The theory behind this relationship—how every color can be produced from three stimuli—is the subject of BI-10. What matters here is this: the "yellow" in the mixture of red and green light contains no yellow wavelength at all. The eye cannot tell the response that the two lights create in the cones apart from the response of real yellow light.

James Clerk Maxwell was the first to demonstrate this principle with a photograph. In a lecture he gave at the Royal Institution in London in 1861, he projected three black-and-white photographs that the photographer Thomas Sutton had taken through red, green, and blue filters, using three projectors with the same filter placed in front of each one, and overlaid them on a screen. The result was a color image of a tartan ribbon. Today, the screen of the phone in your pocket works on the same idea: every pixel produces color by lighting its red, green, and blue subpixels in different proportions.

Subtractive mixing: paint takes light away

Paint doesn't produce light; as you saw in BI-04, it absorbs part of the light that falls on it and sends the rest back. A yellow paint absorbs blue and leaves red and green. A cyan paint absorbs red and leaves green and blue. When you mix the two, the mixture absorbs both blue and red, and all that's left is the green that both of them leave behind. Every new paint you add "subtracts" one more portion from the light reaching your eye.

In theory, a perfect mixture of cyan, magenta, and yellow should absorb all the light and give black. In reality, things are more complicated. Real pigments have broad, uneven absorption bands; each one also absorbs a little of the wavelengths it isn't supposed to, and fails to fully absorb some of the ones it is. When you mix many paints together, these small imperfections pile up: most of the light gets absorbed, but not evenly. What's left is usually a dull mixture that favors long wavelengths a bit, and to the eye this shows up as brown, a muddy tone.

Brown is actually an interesting color, because it has no counterpart in the spectrum. Brown is an orange or reddish color that stays dark relative to its surroundings. If you saw the same light on its own in a dark room, you would call it an orange or yellowish light; when you see it as a dark surface inside a bright surrounding, you call it brown. That's why paint mixtures "turning muddy" so often lands on brown: a mixture that stays in the orange region but has lost a great deal of light.

A third kind: partitive mixing

Maxwell also had a tool that resembled a toy: a spinning top fitted with slices of colored paper. When the top spins fast, the eye can't pick out the slices one by one and sees their mixture instead. But the result is different from the additive mixing we saw with the spotlights: a top made of two complementary colors looks gray, not white. The eye doesn't add up the light from the slices; it averages it.

Partitive mixing is everywhere in daily life. A fabric that looks like a single color from a distance is woven, up close, from threads of different colors. A magazine photograph, seen under a magnifying glass, breaks apart into the tiny dots of four inks; EK-12 explains this raster layout. In the late 19th century, Seurat and the painters around him sought the same effect by placing colors side by side on the canvas as dots instead of mixing them on a palette; the results of this experiment are discussed in SA-04.

Yellow light and yellow paint

Now you can return to the question on the lesson card: are yellow light and yellow paint produced the same way? No. Yellow light can come about in two ways: either it genuinely consists of wavelengths around 580 nanometers, or, as on screens, it's the sum of two separate lights, red and green. In both cases, yellow is produced by adding something to the eye.

Yellow paint does exactly the opposite: it subtracts blue from the white light falling on it and sends back red and green. The mixture that reaches the eye resembles the mixture of yellow on a screen, but it's arrived at by subtracting, not by adding. The same color name is the result of two separate physics. That's why a yellow that looks bright on screen doesn't always come out equally bright when printed; this difference between screen and print is the subject of EK-08.

Example — skin on screen, skin on the palette

When a makeup artist mixes two foundation shades in the palm of her hand, she's doing subtractive mixing: the pigments in both products absorb their own share of the light, and the mixture usually gives a color between the two shades, but a bit duller. When a photo of the same face appears on a phone screen, though, the skin tone is reconstructed through additive mixing, with the light of red, green, and blue subpixels. Whatever the tone, the skin here has been produced twice, by two different kinds of physics.

This difference matters in practice. The shade you see in a cosmetics brand's product photo may not be identical to the pigment in the jar; the screen's and the camera's settings can create a deviation, especially with dark skin tones and saturated shades. How closely the color in a photo matches reality is PZ-08's question for e-commerce, and cosmetic shade ranges are the subject of MT-08.

Common misconception

Summary

1. In additive mixing, lights are added and the total increases; the sum of red, green, and blue is white, and this is how screens work.
2. In subtractive mixing, every paint or ink subtracts a share from the light; because of the imperfections of real pigments, mixtures often drift toward brown, which is orange that has stayed dark.
3. In partitive mixing, small dots, threads, or spinning slices are averaged by the eye; the result is neither as bright as additive mixing nor as dark as subtractive mixing.

Test yourself

1. (Easy) What color do you see when red and green light are overlapped?
2. (Easy) In which kind of mixing does every material you add reduce the reflected light?
3. (Medium) Are yellow light and yellow paint produced the same way?
4. (Medium) Why does a spinning top made of two complementary colors look gray instead of white?
5. (Hard) In theory, mixing cyan, magenta, and yellow should give black; why does it usually give a brownish mud with real paints?
6. (Hard) Why is there no brown light source, yet there are brown surfaces?

Answers

1. Yellow; the eye interprets the two lights' combined effect on the cones as yellow ("Additive mixing").
2. Subtractive mixing; paint, ink, and filters absorb a share of the light ("Subtractive mixing").
3. No. Yellow light either consists of yellow wavelengths or is the sum of red and green light; it's produced by adding something to the eye. Yellow paint, on the other hand, sends back red and green by subtracting blue from white light; it arrives at the same color by subtracting ("Yellow light and yellow paint").
4. The eye doesn't add up the light of the rapidly spinning slices; it averages it, and the average of complementary colors is an intermediate gray value ("A third kind: partitive mixing").
5. Real pigments have broad, imperfect absorption bands; as they mix, most of the light is absorbed, but not evenly. What's left is a dull mixture that favors long wavelengths, and this looks brown to the eye ("Subtractive mixing").
6. Brown is an orange surface color that stays dark relative to its surroundings. If the same light is seen on its own in the dark, it's perceived as an orange or yellowish light; perceiving brown requires the context of a dark surface inside a bright surrounding ("Subtractive mixing").

Sources

• James Clerk Maxwell Foundation, "First Colour Photographic Image by Maxwell."
• National Museums Scotland, "Edinburgh physicist James Clerk Maxwell's explorations in colour."
• Colour Literacy Project, "Mixing processes" and "Optical mixing spinning disks."
• Scholarpedia, "Color mixture."

Next step

DT-03 will take up the question this lesson raises: why do the red, yellow, and blue primary colors taught in school seem to contradict the red-green-blue and cyan-magenta-yellow systems, and when is each one useful? EK-08, meanwhile, moves on to the printing-press version of subtractive mixing: it covers four-ink CMYK printing, why black ink is added, and why bright screen colors turn dull in print.