Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!bu.edu!snorkelwacker.mit.edu!stanford.edu!medisg.stanford.edu!cmgm!vangeldr
From: vangeldr@cmgm.Stanford.EDU (Russ Van Gelder)
Newsgroups: sci.bio
Subject: Re: Primary colors in human color vision
Summary: A little more information
Message-ID: <1991Mar24.002117.24100@medisg.Stanford.EDU>
Date: 24 Mar 91 00:21:17 GMT
References: <00945FE5.1F9B5480@aclcb.purdue.edu> <1991Mar23.193006.22992@pinhead.pegasus.com>
Sender: news@medisg.Stanford.EDU
Organization: Stanford Center for Molecular and Genetic Medicine - CMGM
Lines: 40

A little more information on trichromacy.  The original experiments
which indicated that there are three photoreceptors came from
experiments where subjects were asked to match any given hue with
combinations of 2, 3, or 4 single colors; 3 turned out to be
sufficient when the colors were red, green, and blue (RGB).  Further
work showed that color vision is mediated by the cones of the retina,
which are concentrated in the region of highest visual acuity, the
fovea.  The surrounding areas are primarily populated by the more
sensitive rods, which are used in night vision.  This distribution
explains why astronomers often peer at the stars using their
peripheral vision; looking dead-on at a star dims its appearance
because of the lower efficiency of capture and signal transduction of
the cones in the fovea.  

The cone pigments themselves are fascinating.  Like the rods'
rhodopsin, the cone pigments are made up of a vitamin A-derived
molecule, retinal, and a specific protein, opsin.  There are three
opsins, one for each color.  One cone cell makes only one kind of
opsin; how this level of differentiation is established is a great
open question.  Equally intriguing is understanding how the cone
pigments "tune" the retinal; it is the same photoisomerization of
11-cis retinal to all trans which is the essential act of
photoreception in all three cone rhodopsins.  Something about the
particular side-chains in the vicinity of the retinal must influence
the absorption spectrum of the molecule; Jeremy Nathans (who worked
out much of the molecular biology and genetics of human color vision)
has been making site-directed mutants of the color opsins and thinks
that the relative localization of charge on the retinal is responsible
for its spectral absorbance.

Higher order effects occur at the levels of center-surround
interactions in bipolar and ganglion cells, and cortical processing; a
number of optical illusions exists which show that the same spectral
distribution is sensed as different color depending on its
surround-field.  No good algorithm yet exists for predicting how a
particular spectral field will appear, which is to say that even the
psychophysics of higher order visual perception are not yet well
understood.

Russ