I think I once asked this quetion on the old AOL SDMB, but it either wasn’t answered, or I don’t remember the answer, so here goes:
You often see purple or violet described as being a mixture of red and blue. When arranging colors, it seems natural to arrange them in a wheel, with red adjacent to orange, orange to yellow, yellow to green, green to blue, and purple or violet lying between blue and red to complete the circle. My dictionary defines violet as “any of a group of colors, reddish blue in hue, that may vary in lightness and saturation; the hue of that portion of the spectrum that may be evoked in the normal observer by radiant energy of wavelengths of approximately 420 nanometers.” (My emphasis.) It’s that last part that makes me realize I don’t understand something here: violet light is the shortest-wavelength, highest-frequency electromagnetic radiation we can perceive. Any shorter wavelength, and it would be ultraviolet, and invisible to the normal unaided human eye. Blue light is (according to my dictionary) light of ca. 475 nm. Red isn’t defined in nanometers, but is at the “long-wave end of the spectrum”; orange is merely defined as being between red and yellow; yellow is given as being about 580 nm. Other sources seem to indicate that infrared (i.e., EM radiation waves so long they can no longer be seen) begins at maybe 750 nm. So, my question is, if you average blue (475 nm) and red (call it 680 nm), shouldn’t you get about 578–in other words, yellow? Put another way, violet light is as far from red light as you can get (in terms of frequency and wavelength) and still be visible to us. Why do we perceive the opposite ends of the visible spectrum as being “next to” each other? Why do we perceive color as being a “wheel”, when in fact it is objectively an absolute high-low scale? (Extremely low frequency radio waves never “wrap around” and blend into gamma rays.)
I’ve wondered about this too. Why should a mixture of red and blue appear to the eye the same a purple. I think it is not a matter of wavelength, but probably has something about the way the human eye preceives color. This is just my WAG, though.
"Scientists have known for years how color is seen by the human eye.
Different colors are perceived when light interacts with an object.
The wavelength of the light determines which color we see.
However, the idea of color constancy has never been satisfactorily
explained by any previous mathematical models. Color constancy is
the idea that the colors of objects stay the same despite changes in
lighting conditions. For example, an object doesn’t have a different
color in midday sun than it does at sunset.
A mathematical model is a representation of reality that is easier to
work with and explore than reality itself. In addition to addressing
color constancy, this new model also explains the phenomenon
known as the McCollough Effect, which demonstrates how our
brains work to perceive color. The McCollough Effect can be tested
by looking at a pattern of black and red vertical stripes, followed by a
pattern of black and white vertical stripes. When looking at the black
and white pattern, you are likely to see green in the white areas of
the pattern. The brain creates this color illusion by attempting to
correct for false color signals at the eye, known as chromatic
aberration. In the white lines, the eye thinks it is supposed to be
perceiving red; in the confusion, the eye ends up perceiving green."
Interesting topic. I think you are right, handy and DrMatrix, the answer has to do more with how our three different types of cones in our eyes each see only a portion of the visible spectrum and how ours brains combine this information. Our brains do not know that we are seeing a portion of a linear spectrum. They only know that they are receiving three inputs that they must put together and to which they must assign a color. It appears to be quite an elegant system even if it does not particularly follow a simple mathematical formula. (It is definitely more efficient for the brain to wrap the spectra around rather that having blue (470nm) + red (650nm) = yellow-green (560nm).
I saw one page that suggested that color-blindness may be due the poor-functioning in one of the types of cones. One wonders how our color palette might change if we had four types of cones. Would our brains still come up with the same colors? Would they be slightly shifted?
And, although it seems like it, does the average of the wavelength of two colors really equal the wavelength of the perceived colors (as alluded to above)? I don’t know, but I don’t think so. (We know it is not true for color combinations closer to the ends of the spectrum rather that closer to the center, per MEBuckner.) Does anyone know of a page where one can enter a color and get a generally accepted wavelength? or vice versa? (If blue = 470nm and green = 520nm, does 495nm = cyan? Or is it just a blue-green?) Maybe none of the color combinations follow this formula. Maybe we cannot reach agreement on which hues are cyan and which ones aren’t, so we can’t prove it anyway.
My WAG is that because violet light is almost a harmonic of red light, it primarily stimulates our eye’s blue receptors, but will also by resonance partially stimulate the red receptors. Thus violet is interpreted by our eyes as blue with some red in it.
What I’ve never understood about color perception is why yellow is perceived as a special separate color. Pure yellow light, like sodium vapor light, has a wavelenght intermediate between red and green light. And our eyes perceive it as light that equally stimulates our red and green receptors. But yellow doesn’t look red-green, it looks yellow.
This is probably going beyond the bounds of General Questions, but I wonder if extraterrestrials (or inteligent squid, for that matter) might think we’re crazy because we describe light of wavelength 420 nm as “reddish blue”, even if they can perceive more or less the same part of the EM spectrum that we do. And what sort of weird aesthetic or subjective ideas about color might they have?
I yearn to put this on a more mathematical footing, but “almost a harmonic” won’t help us out. With red at 650 nm and violet at 410 nm, they are not harmonics. If harmonics are involved, it would require the eye sensors (rods and cones) be affected by out visible spectrum light like 205 nm. This would be a more complex scenario, though, and I suspect that lots of testing has been done to verify that we do not “see” light outside of the visual spectrum.
While the intensity of addition and sutraction are reversed, the colors are the same just shifted. They both wrap.
Red is considered to be 630 to 750 or so. Violet is 380 to 420, according to my dictionary (not my physics book). That 750 to 380 is pretty much a full “octave”.
Dr. Paprika is right, and I’ll 'splain things a little more.
Mixing pigments is additave, and mixing light is subtractive. If that makes sense.
With pigments, the three primary colors are red yellow and blue, and mixing each of them with another primary gives rise to the three secondary colors. Mix all three and you get black, or something close anyway.
But with light the primaries are red, green and blue, hence the RGB monitor. I forget the different secondary combinations, but when you mix all three, you get white light.
This page widens the visible spectrum even farther out (“about 350-800 nm”), but I’m not sure that I believe it: http://www.bartleby.org/65/li/light.html
(It also says that UV starts at 400 nm and IR starts at 750 nm, so perhaps they are plagued by to many editors.)
Lastly, this page also has some inconsistencies, but claimed that each cone has about a 100 nm wide filter and purports that the three cones have the highest receive intensity and are centered on 430, 530 and 560 nm. Something here is wrong. We need better specs for these cones to go down the harmonic path. http://mitpress.mit.edu/e-journals/Leonardo/isast/articles/eskinazi.html
(It also shows the visible spectrum to be 390 - 800 nm.)
I was certainly hopeful that individual viewing ranges would not be as variable as hearing ranges. Does anyone have a spectrometer to check the edge of the viewable spectrum? (or a more authoritative cite on the edges?)
Also, there have been a number of threads which mention that the retina of the human eye is sensitive to near-UV, but that the lens blocks it. Possibly this is due to an inability to focus both UV and red to the same point (chromatic aberration), or possibly the UV damages the retina:
What about the wavelength of red-violet? (The combination of red and violet) It is, after all, the color wheel link between the highest and lowest frequency when you include the tertiary colors.
Thanks, Fernmeldetruppe, those links were very enlightening (so to speak). So, basically, the whole “violet is bluish-red” thing is just a quirk of the way our eyes are wired up. Like I said before, it makes you wonder how aliens would see the universe.
I can guarantee that it has nothing to do with harmonics and nothing to do with averaging wavelengths - it does have to do with adding together (superimposing) many wavelengths. Almost all light you see will be some combination of wavelenghts - light of a single wavelength is only produced in very special circumstances with careful filtration of all other wavelengths.
Our eyes perceive whatever wavelengths are NOT absorbed by an object and translate that spectrum (the intensity as a function of wavelength) into a colour (sorry - that’s my canadian spelling coming through). As mentioned above this is achieved by the interaction of those wavelengths with the receptors in our eyes - there are three types and each absorbs a spectrum of wavelengths with the maxima at the wavelengths mentioned in earlier posts.
If an object absorbs mostly red we see green, if it absorbs mostly yellow we see purple - this is the origin of complimentary colours. If the object absorbs both green an purple (which we would observe as red and yellow respectively) our eye interprets this as orange. If the object abosrbs mostly green and orange (red and blue respectively to our eye) we perceive purple. This is the origin of the paint-mixing colour wheel. If the object absorbs both red and green then pretty much the whole visible spectrum has been somewhat absorbed leaving little to interact with our eye and we see something like brown or black depending on the exact absorption spectrum of the object.
I have always wondered whether other people’s subjective experience of red is the same as mine - presumably it is similar, but the existence of colour blind people shows that there is room for variation - their experience of colour must be quite different. And we need not speak to aliens to find out if they think us crazy. If we could just find a way to communicate with the other animal species on our own planet we would find many other perpectives since different species have different ‘visible’ ranges, and so must have different receptors and different subjective experiences.
But if I understand the original post, the question (paraphrased) was this: Let’s take the shortest wavelength of light humans can see, 380 nm or whatever it is. Let’s assume we have a light source such as a laser that is completely monochomatic, and emits light at that frequency. If we show this light to someone, they will describe it’s color as “violet” (or less accurately, “purple”). Now let’s take two other light sources, one that is blue and another that is red. If we mix together the blue and red light in the proper proportions, we end up with a light mixture that to human eyes is indistinguishable from the pure 380 nm light.
Since the human eye has only three color receptors, somehow those receptors must be responding identically to light of totally different spectral characteristics. I said as much that my resonance theory was a Wild-Assed Guess, but something in the 380 nm frequency is convincing the retina that it’s red receptors are being stimulated.
Let me start out by saying that calling the three kinds of cones “red”, “green”, and “blue” grossly misrepresents the response curves for the three kinds of photoreceptors in (most) folks’ eyes, but it’s good enough for describing basic color perception.
Once light hits your cones, their signals don’t go straight to the brain, but are cross-coupled with their neighbors in a neural network. The three color signals are transformed into three equivalent signals that wind up going to your brain: general brightness, red-versus-green strength, and yellow-versus-blue strength.
There is no red-violet frequency. Some color have a dominant frequency, but many don’t. The color wheel is a useful tool for visual perception, but neither maps to all visible frequencies, nor is made up entirely of single-wavelength colors.
No, you will not. Purple and violet are not the same color, physically or perceptually. Violet light does not stimulate your red cones. If you look at a CIE chromaticity diagram, violet has coordinates around (0.17, 0.0), while purple has coordinates around (0.3, 0.15).
Note, in the picture linked-to above, colors outside of the inner triangle are not correct. In fact, depending upon your monitor, the color inside the triangle might be somewhat off, too.
