# Re: [Gimp-user] [Q] Primary colours

```On Sun, Mar 30, 2003 at 02:30:36AM -0600, Scott Lamb wrote:

> >If you mix red and green light your retina perceives all colours except
> >blue, which is equal to yellow (see above).
>
> That's not quite the way I would have put it. Your eye has cones that
> have different frequency responses roughly forming normal curves
> centered on red, green, and blue.```
```
Yes, I know.  I made a first-degree approximation to keep the explanation
simpler.  As it was, it took about a couple pages.  I also ignored the
complication that different frequencies are interpreted with different
intensities.

The model of a vector space with 3 independent vectors (red, green and
blue) is sufficient to describe the behaviour that we see.

The fact that the frequencies seen by the three kinds of cones form normal
curves means that you can add and subtract colours the way I did.  It is
precisely this normality that makes the vector-space model of the spectrum
work so well (though not perfectly).

> I guess you'd sort of see on three axii rather than a spectrum of
> color, if that makes sense. But it doesn't work that way, so a certain
> wavelength of yellow can be faked with certain levels of the peak
> sensitivies. It's indistinguishable by eye.

I know.  But this is an example of why the vector-space model works.
Because of the normality of the curves, the yellow colour is picked up by
both the 'red' and 'green' cones, but not with less intensity each, thus
sort-of justifying the equation:

yellow = red + green

Besides, I'm a mathematecian.  Vector spaces are my idea of "intuitive".
:)

> (Unless you're a tetrachromat.
> <http://www.redherring.com/mag/issue86/mag-mutant-86.html>)

That's really cool!
I can't believe that terachromats are all women!  (now *that* is weird).

Cheers,
--
Daniel Carrera