Colors - primary vs. secondary and the rainbow

Factual Questions
1

teaching my daughter about primary and secondary colors, something occurred to me that I never thought of before… Why is there a 7th color in the rainbow?

Red, yellow and blue are the primary colors
Orange, green, purple are the secondary colors

Red + yellow = orange
Yellow + blue = green
Red + blue = purple

Looking at the order of a rainbow, it (mostly) fits to what one would expect, except where blue and violet are broken down further into indigo.

red, orange, yellow, green, blue, indigo, violet

Was this just a random decision made by whoever decided the “official” colors of the rainbow? There are gradient shades between all colors, correct? So why identify indigo as a seperate color?

Is there more to this (I have no idea about color theory, or even if such a thing exists), or is it just one of those random acts of science that was defined long ago and there is no need to change it? Or, are there actually seven, and only seven, distinct colors of the rainbow? And if so, does indigo have a specific label, like “primary” or “secondary”?

I know that when you mix all colors of light the result is white… When you mix all pigmented colors, black is the result. Do the designations “primary” and “secondary” only work with pigmented colors, and not with colors of the rainbow (light)?

Clearly, I have thought way too much about this, and just as clearly, I don’t know the answer. I am hoping the TM will help me out!

2

You have not thought too much about this at all!

The colors in a spectrum from sunlight vary continuously, and their names vary a bit from language to language. Rainbows are pretty nearly spectral, though there’s some other stuff going on (like the alternating purple and green supernumerary bands).

You have cells in your eye to detect red, and green, and blue. They are not very tightly specific, and there’s a great deal of overlap, especially between the green and the red.

If you want to create light sources of all possible colors by mixing three kinds of lamps together, the closest you can come is to use red, green and blue. This doesn’t work perfectly because of the way the cell sensitivities overlap. Still, it’s not bad. Thus you can think of red and green and blue lights as primary because you can combine them to make other colors.

Pigments absorb light and (except for fluorescent ones) don’t create any. Yellow pigment absorbs blue light, magenta pigment absorbs green light, and cyan pigment absorbs red light. Thus by mixing these together you can create all the other colors colors though darkness (if you follow me) the way the primary colors create all the other colors through lightness. They call these secondary colors.

How primary and secondary relate to the named colors of a spectrum or rainbow is a bit of a mishmash because the Ministry of Naming Stuff hadn’t been appointed yet, so you’re right about the random decision thing.

“Color science” is the name of the field you’re interested in. It’s neato. There are, for example, a different set of primary colors called X, Y and Z that truly can be mixed to create any color you can see, including all the colors of the spectrum. Trouble is, X, Y and Z are imaginary, and to create them out of real physical lamps you need to be able to contribute negative brightnesses at some wavelengths. Then there are gamuts, which are ranges of color you can create by mixing any real group of physical colored lamps or pigments. Here’s another nifty bit - your eye measures the red and green and blue, but then it converts them to get the ratio of green to red, and also the ratio of blue to (green+red). So, this weirdness is closer to what’s on the wires between your eye and your brain.

3

Pretty much. Newton, for all of his genius, was also quite subject to woo, and thought that seven colors was more mystically appropriate than six.

4

Yeah, there was really no good reason for Newton to stick indigo in his spectrum, thereby confusing school children forever, and I wish he hadn’t done it. (Although I suppose that some of his other ideas were half-decent. ;))

It’s not entirely clear which colors Newton meant to indicate by each label, anyway. The way I’ve most often seen his spectrum printed, his “indigo” is blue, and his “blue” looks cyan. Like this.

5

Color naming is completely arbitrary. Personally, I have no idea what indigo is supposed to be or the difference between violet and purple, but I find the difference between blue and cyan rather fundamental.

Compare color to music. A single tone maps to a specific frequency/wavelength of light, which can range from red to very deep blue. However, there is no wavelength of light that maps to purple. To get purple, you have to mix red and blue. Add green and you get white. That’s like a chord in music.

6

The primary colors (when mixing pigments or paints) are not red, yellow and blue.

They’re magenta, yellow, and cyan.

7

I suppose that if anyone would know, it’s you.

8

:wink:

9

RYB is a set of primary colors. It’s just not the one with the widest gamut for subtractive color mixing.

10

Maybe you could try to explain the idea that what we ‘see’ is always light; either reflected or refracted. The colours are how our brain interprets the signal from our eyes. Some people are colour blind and confuse colours, which is why traffic lights always follow the same sequence.

11

Except for (at least) this one.

12

Yeah–if he was going to make 7 colors, it should have been Red, Orange, Yellow, Green, Cyan, Blue, Violet. While it’s hard to see looking at an actual rainbow (as they are so desaturated), every picture I’ve seen where they do something to make it brighter has a significant cyan band. (I don’t have a prism to do it myself.)

I have seen some bands with 8 colors, though, and it’s the two colors after blue that I think may have been indigo and violet. Problem is, the colors are what painters would call violet and purple, and crayola users would call blue-violet and red-violet. Here’s an example

13

To me, that’s the difference. Violet is–or can be–a single wavelength, on the short side of the range that the human blue receptor will register. (Not a single wavelength in the sense that there’s only one frequency that’s true violet, but if you emit light on a single frequency within that range, it would be violet by itself.) It’s between blue and UV.

14

For the old rear projection TV’s, they were (are) red, green, and blue. Why no yellow in this case? Why wouldn’t they be the same colors as for mixing paints?

15

The short answer is that mixing light doesn’t work the same way as mixing pigments. A red pigment mixed with a green pigment will give you an ugly brownish mess, while red light mixed with green light will give you yellow.

16

It’s not just your old TV, it’s your swanky new computer screen, too.

This is just your good old additive versus subtractive color mixing. You want a limited set of primaries that combine to produce as many colors as possible. For light, you add wavelengths together (additive mixing). Pigments and dyes, on the other hand, absorb wavelengths (subtractive mixing). It turns out that RGB are pretty good primaries for the former, while CMY are good for the latter.

Note that the secondary colors you get for additive mixing are the same as the primary colors for subtractive mixing, and vice versa. So it’s two sides of the same sort of thing.

For a higher level “why”, as in “why does it work this way”… well, it involves nature, your brain and God. Three complicated things.

17

You are confusing additive colors with subtractive. One is for pigment mixing, the other is for light mixing. They are opposites, color-wise.

18

Thank you for the reply and the identification of “color science” as the term I was looking for. And thank you for the sanity check. I really thought to myself “you are thinking WAY too much about this”, but it kept nagging at my noodle. I am always happy when I ask a question out here and discover I’m not as crazy as I (sometimes) think I am!

[Quote=Chronos]

Pretty much. Newton, for all of his genius, was also quite subject to woo, and thought that seven colors was more mystically appropriate than six.
[/QUOTE]

I did a quick google, and was directed to a wikipedia page, where I found this nugget:

I did not check the sources for accuracy, but I’ll assume they are close enough for the purposes of this discussion.

I wonder… Why has there been no change in the identification of the colors of the “rainbow” since Newton? Is it because Newton is credited with the identification and naming of the seven colors we all know now, and questioning/altering anything Newton came up with is sacrilege? Is it just tradition?

There seems to be no scientific reason for the 7th color (no moreso than naming 8, 9 or any number of shades visible in the spectrum), and the choosing of the number 7 was based on arbitrary (and it turns out, faulty) reasoning.

Would there be a major problem/objection to changing the rainbow to reflect the six “standard” colors I mentioned in the OP (red, orange, yellow, green, blue, purple)? I don’t want to get into a debate about the names of these 6 colors (i.e. If cyan is a more accurate label than blue, fine). Having 3 primary colors and 3 secondary colors to represent the “rainbow” seems to make a lot more sense. It would (to me, at least) be a better way to represent the basic concept of visible color. We could argue about the maximum number of colors to include in any rainbow, but I don’t believe we could reduce any rainbow to less than the 6 (3 primary, 3 secondary) colors already discussed.

When the world’s scientific community made the decision to remove Pluto from the list of planetary bodies orbiting the sun, the world managed to cope. A recent set of planets I puchased for my daughter’s ceiling doesn’t include a “Pluto” :eek: - the world has clearly moved on. So why have we not updated our concept of the rainbow? Are there viable reasons we stick with the current model/naming convention, or is there simply no compelling reason to update it?

No Pluto? No Indigo!

19

I say keep the number seven, but label them red, orange, yellow, green, cyan, blue and violet, like BigT said. The cyan band is pretty prominent, and there’s no reason to pretend that it’s not there. Indigo gets the shaft, that’s fine.

Roy G. Cbv. Rolls right off the tongue, doesn’t it? :wink:

There’s nothing “primary” or “secondary” about the colors of the spectrum anyway. Any color in the rainbow corresponds to a wavelength of light, there’s no mixing going on. Sure, the colors red, yellow and blue happen to correspond to the primary colors in the typical artist’s color wheel, with orange, green and violet/purple as secondary. But this color wheel is just one way (and as we have seen not the only way, with RGB and CMYK working differently) of creating a color space and systematize colors.

Also, unlike your color wheel, there’s nothing “circular” about the rainbow or the electromagnetic spectrum. If you mix red and blue in your artist’s color wheel, you get a purple color that your brain interprets as looking more or less the same as the violet end of the spectrum. But this is your brain (and the color receptors in your eyes) playing tricks. It’s an optical illusion. The wavelengths corresponding to violet is way on the other side of the visible spectrum from red. They don’t meet up, they just bugger off into ultraviolet and infrared.

BTW, another fun optical illusion: Magenta, a primary color in your friendly CMYK model, isn’t even in the rainbow. It’s a non-spectral color, and doesn’t correspond to a specific wavelength of light. In a sense it exists only in your brain. (Although I’m note sure if you should mention that to cmyk, he may get some existential angst…)

20

:smiley: I always tell my students (jokingly) that Newton included indigo because it would be awkward to say Roy G Bv.