Test your color vision with our Visual Mill

We are ALL colorblind

The fact is, every human being is colorblind. Yes, from an objective point of view, even people with normal color vision have profound colorblindness. We use colors to distinguish between objects that reflect the visible spectrum differently, but we confuse completely different spectra as the same color. However, the same "colors" may be distinguishable by animals, insects, and by some people with anomalous color-vision (and surely by many aliens).

Colorblindness of normal color vision
The six spectra below look the same purple to normal color-vision people =>


Spectra obtained using a simulation by Hughes, Bell and Doppelt (Brown University)

What's color, anyway?

Vision, Spectra, Colors and Metamers

We have three color detectors in the retina (the "cones"), each tuned to a certain frequency range of the visible spectrum. All the colors we see we derive by combining those three signals. For example, the six spectra above all excite the long, medium and short wavelength cones in the same way:

Obviously, three variables cannot define a spectrum that can be divided into much finer detail, say 100 frequency slots. To reasonably distinguish every "physical color" with that resolution, we would need 100 different types of cones, not just 3!

Colors that look identical but have different spectra are called metamers. However, what may be metamers to somebody, may not be metamers to somebody else.

Who says who is colorblind?

About 8% of males (in most populations) are traditionally defined as colorblind. They are commonly considered to have lower color discrimination than "normal" color-vision people. And in general terms that is true, specially for dichromats who only have two functional cones. But, even dichromats are able to distinguish between colors that to the average person are identical. Some of the purple metamers above will definitely look different to them.

Moreover, 3 out of 4 so-called color-blind people are anomalous trichromats: they have 3 types of cone photopigments as do non-colorblind people. Their color vision is different because at least one of their photopigments does not have the same absorption characteristics than those of the average person. These trichromats may have different metamers.

In fact, the standard clinical instrument for testing color vision is the anomaloscope, where the subject must match spectral (monochromatic) yellow with a combination of red and green light. The "normal" mixture ratios were found by Lord Rayleigh in 1881. Some colorblind people are happy with a match over a large number of ratios, showing lower color discrimination. But other colorblind people will not accept the Rayleigh match and they will make the match at another red/green ratio. So, one could say, they are not fooled by the Rayleigh mixture of red and green into thinking it is identical to spectral yellow.

Testing colorblindness using the common pseudo-isochromatic plates (e.g. Ishihara, Dvorine, HHR, AO) may be unfair to some anomalous trichromats. After all, these plates have been optimized to make him fail and to pass the average vision guy! Similar plates could be designed for him to pass and for the "normal" trichromat to fail!