Extraordinary claims. Ordinary investigations.

Magenta and all the colors of the grey matter


Above you can see the visible light spectrum, which can be understood as electromagnetic waves with different frequencies, going from the longer ones on the right to the shorter ones to the left. Go even lower and you enter the infrared, microwave and radio waves, go higher and you advance over the ultraviolet, X and gamma rays.

We see the visible light spectrum everywhere, and in a specially beautiful form in the rainbow. It’s common to think that the rainbow contains all the colors, so much so that the expression "all the colors of the rainbow" has 134,000 hits on Google, almost all of them referring to all the colors in the world.


Now, find magenta in the spectrum. It’s simply not there. All the colors of the rainbow do not contain the magenta. How can we see it then? Is it a “magical color frequency” that don’t actually exist?

An article from Liz Elliot suggested last month by the folks of BoingBoing provoked a lot of discussion by claiming quite simply that “magenta ain’t a color”. It sounds unbelievable and yet…

And yet, well, it’s not actually true. Keep reading because the amazing thing is, the truth is even stranger.


In Yes, Virginia, there is a magenta, Chris Foreman corrects many of the errors in Elliot’s article, and right from the beginning, shows a CIE color space diagram:cie_color_bag

The neat thing is that magenta is right there. And what the CIE color space represents is exactly the human color perception. Created through a series of experiments in 1931 by the International Commission on Illumination (CIE), the diagram takes into account how we experience colors, and here is the simple correction to Elliot’s mistake. She simply referred to a wrong definition and representation of what colors are. Because they are what we experience.

Colors are not determined by their frequency in the electromagnetic spectrum. This may sound absurd, but keep with us. We simply don’t have an optical sensor capable of measuring the exact frequency of the light that enters our eyes, but rather have three different types of cells sensitive to three different ranges of frequencies. We literally interpolate the measure of a broad continuous spectrum through three sensors with very limited capabilities.

In the color space represented above, the visible light spectrum makes up the curved contour line. The light wavelengths are there in the continuous line, going from 380 to around 700 nanometers. But the light spectrum is only the the curved contour line: all the infinite colors inside the diagram, including the lower straight line with our beloved magenta, “do not exist”, or at least, do not exist as defined electromagnetic frequencies. You won’t find any of them in the rainbow.

Brown, pink and countless other shades of color inside the diagram go along with the magenta and are composed of different light frequencies stimulating different cells in our eyes simultaneously. And those are not the only colors that are not in the rainbow, as the CIE color space represented above is not complete.

We need one more axis to represent color intensity, going from black to white, and all the shades of grey in between. Imagine the drawing extending three dimensionally, with different brightness values. Because, if you didn’t notice, neither black, nor white nor any shade of grey can be seen in the visible light spectrum. Are they not real?


As it must be clear by now, the visible light spectrum that makes up the colors of the rainbow are just a very small part of the full human color perception. And thus defining colors as determined electromagnetic frequencies is simply wrong; it’s impossible to define colors without making reference to how we actually perceive them.



Understanding how we see colors and presenting the color space derived from this knowledge does illustrate something interesting. Magenta may not be a magical color frequency, but it is special in a way.

Remember that the curved contours of the CIE diagram are made up of the visible light spectrum, those are indeed defined electromagnetic wavelengths, they are spectral colors.

That curve is quite different from the straight bottom line between violet and red. A straight line that includes our magenta, and is known as the line of purples. They make up part of the edge of the diagram, and yet, are not spectral colors. And it’s easy to understand why.


The visible light spectrum is linear, it starts at some point and ends at another. But the color perception based on three chromatic cells in our eyes ends up producing a circular color space, where you can start at one point and travel through all the others in the same direction until you return to where you started. The straight line between red and violet is the “imaginary” line that completes this circle, making the linear spectrum into something circular.

And magenta is right there in the middle, a mix of red and violet. It is indeed special, as it is a completely saturated color that can’t be produced by a single light frequency. If we lived in an Universe with different physical laws and mathematical axioms, where the visible light spectrum was somehow circular, perhaps those purples, including the magenta, would be then finally part of the rainbow. Circular rainbows, I guess.

There are circular halos, in fact, but guess in what direction the spectrum is scattered? No, not like the color wheel above. Damned Nature, it keeps being coherent.

[From Calovi on Flickr]

Of course, even in a Bizarro Universe where the visible light spectrum is  circular and rainbows look like color wheels, all the infinite colors inside the CIE diagram would still not be there in the rainbow, and yet, would still be colors.

The truth is, quite simply, that despite all the poetry, colors are not that much related to rainbows. They have much more to do with our retinas, and as we will see, with our grey matter.



Perhaps you may have found cumbersome that the CIE color space diagram is not a regular triangle, but rather something resembling a horseshoe. Too bad, because that’s reflecting how our color perception is not that neatly geometrical: it would be a neat triangle if our chromatic cells reacted the same way to the wavelengths they are intended to react to, and if they were distributed evenly in the retina and the visible light spectrum. They aren’t, and the ideal triangle turns into that horseshoe.


In fact, our sensibility to different wavelengths varies a lot, and the cells sensitive to what we usually call green and red are heavily overlapping in their response. When you see a perfectly red, a spectral red sign for instance, the cells sensible to “green” are also being stimulated. Just slightly less so. How can we see the difference between green and red so clearly?

Many people can’t, as they are color blind. Just a small mutation, and the overlapping between those already so similar chromatic cells is so much that you are no longer able to differentiate between them. But to the rest of us that have a slightly more varied set of cells, we must thank our brains.

It’s amazing how we can experience such a vast and almost triangular color space despite all of our limitations. The complete set of the visual system is the result of millions of years of evolution of all of its components working together, and the brain is a very big part of that. Some see perfection in the human eye, but when you actually see it closely, you may find a lot of not so intelligent designs. One could say that what haven’t been implemented through our eye hardware is corrected through our brain software (in fact our brain is hard-wired to deal with these, but anyway). In the end, it simply works. Most of the time.

And if you look really closely to the subject of color perception, you may find something quite extraordinary. Like the hamburger below. Can you see the green lettuce, the baked bread, the yellow cheese, the brown beef?


Congratulations, you have just seen all these colors with your brain, not your eyes. The image is actually composed solely of one “color”, in shades of red and grey. See more details by the author, Chris Taylor.

hamb_upHow did you see colors that are not really there? If you read this far, you already know colors are what we experience, and what we experience depends heavily on our eyes perception and how our brain interprets that. And there’s one little detail that can change your color perception: color constancy, “which ensures that the perceived color of objects remains relatively constant under varying illumination conditions. A green apple for instance looks green to us at midday, when the main illumination is white sunlight, and also at sunset, when the main illumination is red. This helps us identify objects.”

Amazing as it is, this effect that can make one “color” be perceived as other has only been studied starting in the early 1970s by Edwin Land, inventor of the Polaroid camera, who called it the retinex theory. Some famous visual illusions circulating in the internet recently are playing with color constancy:


A and B are the same shade of grey. Which is not part of the rainbow, and is along with magenta all experienced in the brain. The truth is, the grey (and white) matter allows us to feel colors well beyond the rainbow. All the colors, from the spectral ones to the line of purples, are a product of the grey matter.

Blog Widget by LinkWithin

Popularity: 4% [?]

Posted in Science | 9 comments

9 Comments so far

  1. henteaser March 24th, 2009 6:04 am

    Magenta and all the colors of the grey matter | forgetomori…

  2. henteaser March 24th, 2009 6:26 am

    “As it must be clear by now, the visible light spectrum that makes up the colo…”…

    As it must be clear by now, the visible light spectrum that makes up the colors of the rainbow are just a very small part of the full human color perception. And thus defining colors as determined electromagnetic frequencies is simply wrong; it’s imp…

  3. Me March 25th, 2009 11:53 am

    I’m sick of seeing the A/B thing with the cylinder. The squares of A & B are NOT the same. Open two windows, or print them out, put them side by side and they are different. I don’t know why this continues. I guess it’s the same reason religous books aren’t in the fiction area of the book store.

    Before someone says the cylinder is doing it. I mean just the square. None of the background, or other things to fool your brain.

  4. Saoki March 25th, 2009 4:33 pm

    People forget how much of the world is actually a simulation made by our own minds, like movement, color or truth.
    So much easier to deal with a world that’s exactly what you’re seeing…

    You did forgot to define Qualia. Unless you wanted people to look it up by themselves, that is.

  5. saoki (saoki) March 25th, 2009 7:28 pm

    Uma ótima explicação de teoria da cor e como tudo depende de percepção. Leiam, artistas! http://tinyurl.com/cjvlsn

  6. Optimus March 27th, 2009 3:08 pm

    The squares A and B are really the same. Putting them together may not show. Open the image in any painting programm (even ms paint) and with the select/copy tool take small fraction of the A and B squares and put them together. If you moved the whole square it wouldn’t easilly show sometimes. They really are the same. I think the trick here is the shadow on the B square that makes the color matching that of A, but it’s not visible because the alternating patterns of checkers confuses the brain to perceive the color as the one of the brighter squares. It’s really an amazing illusion!

  7. Peter April 21st, 2009 4:51 pm

    At first the magenta problem seems perplexing, until you understand the physics of how color (and sound for that matter) work. I will augment your very good article with the following comment:

    No one is particularly surprised that music is not composed of pure tones. A pure tone (think: sound before an “emergency broadcast system” message) certainly does not comprise a “song”. It is extremely complex MIXTURES of these tones that make songs and just about any other sound on earth.

    Colors mix just like tones do. The visible spectrum represents the range of pure “tones” of color. As these colors “mix” they create new colors just as sound waves “mix” to form new sounds. You should be no more surprised that magenta is not in the color spectrum than you should be that “Stairway to Heaven” isn’t a pure tone. Methinks Led Zeppelin wouldn’t have sold many records if that were the case. =/

  8. […] be TinEye. And this is surely the best illusion in many years. Want to know what’s going on? Perhaps this will help.Related PostsClassical illusions revisitedThe image above is not a spiral staircase. Look closely […]

  9. Justin July 4th, 2009 4:02 pm

    this earlier mentioned that the suns light is white. it’s my understanding that the suns light is blue. that is why daylight film is color balanced on the expectation that the light source, presuming the sun, is strong in blue.

Leave a reply

Live Comment Preview