Someone wrote:
Now for colour. It is a quirk of the uman eye that
you can get the
visual effect of any paritcular colour by seeing the appropriate
amounts of red, green, and blue light.
Mouse wrote:
This is only approximately true. There are two reasons
it's only
approximate.
One is the likely existence of human tetrachromats.
There are two
viable alleles known for one of the colour receptor pigments;
There are *many* alleles for the green pigment, and at least a few for
red. The genes are long and it is common for them to get transcribed
wrong, i.e., to have a gene for the green pigment that partway through
turns into a copy of the tail of the gene for the red pigment. This
usually results in a pigment with a frequency response peak close to the
linear interpolation of the fraction of the green and red gene components.
That still means that any color perceptible by that person could be
produced as the sum of three primaries, though the necessary primaries
may be different than those necessary for another person.
a woman
(it's on the X chromosome) may have both of them, presumably leading to
a richer experience of colour than men or other women.
No one has been able to demonstrate that this has any effect; it is
possible that the brain is unable to distinguish red1 from red2, which
would then effectively be just a slightly wider peak for red.
The other is that it's (almost) certain there are
light spectra that
produce retinal excitation patterns not recreatable by RGB triples.
This is because the retinal pigments' sensitivities overlap somewhat;
for example, green light stimulates the red and blue receptors to some
minor extent. The sensitivity curves are not very complex, each being
a fairly broad single hump from what I can tell, but, especially at the
ends of the spectrum, there can be colour experiences that cannot be
reproduced with three-spike RGB spectra. (Example, using made-up
numbers: consider light in the high violet which produces stimulation
ratios R=1 G=10 B=100, while the light produced by blue phosphor - or
blue LCD filter component - cannot reach B=100 with G less than 25.
It's certainly possible to create B=100 with G=0. It's not surprising
that CRTs and LCDs do not have sharp enough spectral lines to do this.
There are laser-based display systems that do. With the right three
primaries for an individual (possibly four for a tetrachromat), with
sufficiently sharp spectral lines, it is quite possible to produce the
equivalent to any other perceptual color.