Color question

[kleinwolf]

How do we know that 3 base colors are needed to compose any color and not 4 for example ? Is there a physical proof from medical analysis of eye's cell ?

---

btw : if it were a palette we would need 2 to localize the color : distance right and up from a corner for example.

Edited September 30, 2009 by kleinwolf

[Sisyphus]

I just looked it up on Wikipedia (http://en.wikipedia.org/wiki/Color). Apparently there are three types of color receptors in the eye, with peak sensitivities at different wavelengths. Color is perceived based on the proportion at which each type is stimulated. However, apparently some people have four kinds:

 

As many as half of all women are retinal tetrachromats.[7] The phenomenon arises when an individual receives two slightly different copies of the gene for either the medium- or long-wavelength cones, which are carried on the x-chromosome. In order to have two different genes one must have two x-chromosomes, which is why the phenomenon only occurs in women.[7] For some of these retinal tetrachromats, color discriminations are enhanced, making them functional tetrachromats.[7]

[granpa]

check this out:

http://www.gmodules.com/ig/creator?synd=open&url=http%3A//colorchoosergadget.googlecode.com/svn/trunk/colorChooser.xml&pt=%26context%3Dd%26type%3Dgadgets%26synd%3Dig%26lang%3Den%26.lang%3Den%26country%3Dus%26.country%3Dus%26start%3D0%26num%3D1%26target%3Df7KG%26objs%3Dd%2Corj%2C8AI%2CYy%2Cjt0Q&sn=f7KG〈=en

 

if that link doesnt work then try this:

http://www.google.com/ig/directory?synd=open&q=Color+Chooser

[Mr Skeptic]

How do we know that 3 base colors are needed to compose any color and not 4 for example ? Is there a physical proof from medical analysis of eye's cell ?

 

Yes, it relates to how many different types of cone cells you have in your eye. Most people have 3, colorblind people have less, and a few people (mostly women) are tetrachromats and have 4. You need one color for each type of cone cell.

 

Normally, light of a certain wavelength is observed as exciting the various cone cells to a different extent each, and if you have the colors for the cone cells in the right proportions you can't tell the difference. However, you can still split this "fake" color with a prism.

 

If you put a bit of water on your screen you will be able to see the colors of your screen without blending, at least for the old fashioned ones.

[kleinwolf]

Hi Sisyphus,

 

If some people are 4-chromat, do they see in 4 dimension ?

 

(Has this something to do with the 4-color colorization theorem of maps w/o enclaves ?)

 

Weird numbers : three people with ternary logic (like colors), have a decisional power of 77.7% ?

[Fuzzwood]

Sigh, since when are colors dimensions?

[kleinwolf]

Because we could write : [math]\vec{color}=a*\vec{red}+b*\vec{green}+c*\vec{blue}[/math]

 

since those 3 colors are independent. Probably the proof it is a complete basis for color is experimental with a prism, like said above.

 

May be asked : how to get the temperature or frequency out of this vector ?

[Klaynos]

Colour as perceived by a human, is different from the pure colour of a photon.

 

A human mixes and interoperates the photons setting of receptors with different efficiency.

 

Your equation doesn't really help with finding what you want.

[kleinwolf]

As an electromagnetic waves ? The duality is not solved, but then need an infinite base..

---

Merged post follows:

Consecutive posts merged

Thank you swansont, I have memory problems with my mobile with what I'm writing this.

 

I thought inbetween about : given a wave, would a Fourier transform give a continuous spectrum, whilst expanding in a pure wave basis would be discrete but infinite one.

---

Merged post follows:

Consecutive posts merged

3 independent color were a complete generating set, does this mean that the electromagnetic wave picture is completely out, and hence the wave can only be the probability's amplitude of presence ?

---

Merged post follows:

Consecutive posts merged

Your equation doesn't really help with finding what you want.

 

...then the grey text ?? if u do that, u dont turn back but you lose all your properties and maybe even the next step you should make ??

---

Merged post follows:

Consecutive posts merged

Can we say that the eyes' data are 13 dimensional (2(eyes)*2(xy)*3(color)+1(time)) ?

Edited October 2, 2009 by swansont
duplicates removed