I had always thought that purple and violet were very similar, just slightly different shades of each other. But Wikipedia says differently:
I’m having a lot of trouble understanding the difference between a true spectral color, vs. a composite color.
Wikipedia also says that green and orange are spectral colors. Is there a composite color analogous to purple (i.e., not green) that is made by combining blue and yellow? Is there a composite color (not orange) that is made by combining red and yellow?
It depends on what you mean by “color”. You can have two completely different spectra, which will look different through a spectroscope, but which look identical to the human eye. For instance, if you have a pure red light (from a neon tube, say) and mix it with a pure green light (I’m not sure precisely what works for this, but there are sources), you’ll get something that looks to the eye just like a pure yellow light (say, from a sodium vapor lamp).
Mix red and blue together, and you get something that looks just like if you took a single wavelength somewhat shorter than blue. This situation is somewhat unusual, in that violet is not in between red and blue, like most apparent color combinations are. But they still look the same.
And an important difference between purple/violet and green/green, orange/orange as mixes or spectral colours is that those latter mixtures give you the sensation of a color with a wavelength between the two colours mixed, whereas purple is the mixture of opposite ends of the visible spectrum, giving you the sensation of the extreme end of “blue”.
The article on purple also mentions this difference between the spectral colour violet and the mix purple:
Now the Bezold-Brücke shift is described as
So a mixed orange should stay orange, while a spectral orange sensed to be the same hue should shift towards yellow.
It’d be fun to experiment with this, but it would require a rather advanced system of wavelength-adjustable light sources … Or just a lot of lamps and filters.
Thanks! I keep forgetting that violet is at the far end of the spectrum, and is NOT located between red and blue. The reason I keep forgetting that fact is probably because on a circular color wheel, it does end up between them.
So I guess now the question shifts to our perception of these colors: Why do we think of violet as being reddish, when in fact red is at the other end of the spectrum?
it means it’s possible for something to emit pure violet light. Violet light corresponds to a wavelength of roughly 400 nm. there are LEDs available which do this (Indium-gallium-nitride.)
purple is a range of colors, produced by mixing red and blue.
Spectral colors are the ones you find in the rainbow. You’ll notice that there’s nothing there that looks purple. The really intense blue at the end of the rainbow is sometimes named “violet” (hence “ultraviolet”) but it’s not very close to purple.
The white light of the sun consists of (almost) all wavelengths, and a rainbow or spectrum shows them from long to short. So each spectral color consists of a single wavelength of light. Non-spectral colors need multiple wavelengths. This includes all pastels such as pink or powder blue, those add white. Purple is is also a non-spectracl color, it consists of red and blue but no (or little) green.
Orange is a spectral color but you can get pretty good orange by mixing red and green, the same way you can get all colors by mixing the three primary colors are eyes are sensitive to, red, green and blue. (Actually it’s a bit more complex than that.)
What you’ve found is quite correct – purple is a composite color, and not spectral – that is, you can’t separate out white light with a prism or diffraction grating and find Pure Purple. But you can do this with violet.
To answer your last question, about combining two spectral colors to get something not a spectral color. well, speaking technically, it happens all the time.
Color science is a weird thing, partly because it involves the human senses and brain, which form a weird mix. They spent a lot of time screwing around with trying to figure out how color works. In 1931 an international convention , the International Committee on Illumination (CIE, from its French name, Commission internationale de l’éclairage) took a lot of work that had been done and formally recognized it and published it. There have been other meetings and ramifications since, but the basics established then pretty much hold.
Using methods I’m not going to go into now, they represented color on a two-dimensional graph, in which the x and y axis are “normalized” colors. A roughly horseshoe-shaped curve on this represents the “spectral locus” of pure spectral colors. The line joining the ends is “purple”. Al;l colors that can be observed fall inside the horseshoe-plus-line diagram: Chromaticity - Wikipedia International Commission on Illumination - Wikipedia
White is somewhere around (0.3, 0.3), the exact coordinates depending upon how you define “white”. If you have two different colors represented by points on this diagram, you can generate any color on the line joining them. Thuis all putples can be generated by the red at one end and the blue at the other.
Since only points on the actual spectral locus are true spectral colors, the combination of any two spectral colors willgenerate a point NOT on the spectral locus, and therefore ANY combination of spectral colors meets your description.
The problem is that this is only technically true, because people’s definitions of colors are kinda wobbly. On a 12931 CIE Chromaticity diagram, you can’t tell apart colors falling within an ellipse (called “Macadam ellipses”, after the guy who discovered them. His office used to be down the hall from mine). Besides, people are forgiving in their definitions of colors, and even outside the ellipses they’ll still use the same names to describe them. (For what it’s worth, we pretty much use the same name for any color along a line from that “white” center to the spectral locus. We just say that, as you get close to “white”, the color gets lighter and more “washed out”)
So, unfortunately, although you can generate new colors by mixing pure spectral colors, you’re not going to create some whacky new color that you’ll have to come up with a new name for, like mixing pure blue and pure yellow to make “gormwatz”. What you’ll make, unsurprisingly, is some kind of green, slightly lighter than the corresponding pure spectral green.
By the way, if you mix colors made by THREE colors represented by points on the CIE diagram, you can make anything inside that triangle. This is what happens with color monitors, which use three phosphors to make color images. There’s a vast literature on this, which you can lose yourself in iif you want. Try here: Understanding the Color Gamut of an LCD Monitor or sRGB - Wikipedia
The point is that, even if your three basis colors are on the spectral locus (and they aren’t for most monitors, but I suppose you could do it with three lasers), yoiuur mixtures won’t be, and you can’t duplicate the spectrum. This is why the CIE diagrams you see on your monitor look mkind of wonky and not right.
As I recall a color wheel relates to pigments and mixtures thereof. Mixing lights, as in stage lighting, follows a different path, and dealing with wavelengths as mentioned above may be different yet.
Ok, but couldn’t you have violet paint that absorbs all the other visible spectral colors? Or a violet filter that only passes spectral violet from a true white light source?
A related fact is that the color pink does not exist at least as part of the light spectrum. This article explains why and also has relevance to the purple/violet question.
See, to me, and no doubt I’m influenced by Crayola and my alma mater’s school colors (Northwestern), that’s definitely a color I would also call “purple.” I mean, look at the rainbow here. That’s not purple at the end? Or look at the Northwestern logo. That’s called “purple” by everybody connected to the school.
Colloquially, purple and violet are pretty much interchangeable in my dialect. I will tend not to call reddish purples “violet,” but bluish purples can go either way.
This is a good point, indeed if we wanted to reach as many people as possible we probably would want to use the word “purple”. However, if we say the ozone layer absorbs ultrapurple light at a scientific conference, maybe we’d lose a bit of credibility.
Bravo for CalMeacham’s post. I was going to post something on this, but he totally nailed it better than I could hope to. Really, it is all there. Go back and read it if you are still puzzled.
Important point. You cannot post links to pictures of colours and talk about whether there is purple or violet in them. Your screen is only capable or reproducing those colours that live within triangle on the CIE diagram the three basic colours it uses defines. If your screen’s colour gamut does not include violet (and it won’t) what you see on the screen is the three colour approximation your screen provides that does live within that triangle. And, guess what? It will be a mix of your screen’s red, green and blue. Probably not much green, so - a mix of red and blue. Better known as purple. Your screens inability to actually deliver colours out of its gamut will fold down both true spectrally pure and mixes into that gamut. This renders discussion on the actual colour pointless.
A complicating factor in the question of purple versus violet is that if you look at the sensitivities of the eye’s receptors, there is a glitch in the far blue spectrum, where the effective sensitivity to red is non-zero. So violet is perceived as a mix of red and blue, but due to an artefact of the sensor’s spectral response. However, you can’t reproduce the exact stimulus by actually mixing red and blue light, as this will inevitably result in some green stimulus as well, and you won’t get exactly the same values presented to your brain. So violet manages to remain not purple, but still close. (Note, there are not such things are red green and blue sensors, just three different sensors with slightly different spectral responses and some tricky neural processing at the back of the retina that creates an encoding of the colour and brightness to send to the brain. That coding isn’t RGB either.)
