Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!sol.ctr.columbia.edu!emory!gatech!purdue!haven!decuac!shlump.nac.dec.com!decuk.uvo.dec.com!hollie.rdg.dec.com!psw.enet.dec.com!winalski
From: winalski@psw.enet.dec.com (Paul S. Winalski)
Newsgroups: sci.bio
Subject: Re: Primary colors in human color vision
Message-ID: <1991Mar23.063140.27779@hollie.rdg.dec.com>
Date: 23 Mar 91 06:31:40 GMT
References: <00945FE5.1F9B5480@aclcb.purdue.edu>
Sender: news@hollie.rdg.dec.com (Mr News)
Reply-To: winalski@psw.enet.dec.com (Paul S. Winalski)
Organization: Digital Equipment Corporation
Lines: 56

In article <00945FE5.1F9B5480@aclcb.purdue.edu>, miguel@aclcb.purdue.edu (Phillip) writes:
|>   The additive primary hues, according to an undergrad general psychology 
|>text book I have (Gleitmann (sp?)), are blue and yellow (which are 
|>complementary) and red and green (complementary).  Blue, yellow and green 
|>all have a "unique" wavelength at which the human eye/brain percieves them 
|>to be without tinges of any other color.  Red is "extra-spectral" in that 
|>it requires a combination of wavelengths to produce a "pure red" 
|>sensation.  The complementary hues when mixed in equal amounts produce 
|>grey (i.e. blue + yellow = grey).  The process is called "opponent-pair" or 
|>something similar.  Mixing (adding) non-complementary hues produces an 
|>intermediate color (i.e. red + yellow = orange).
|>    The subtractive primaries are different.  Two pigments mixed together 
|>only allow wavelengths neither absorbs to be reflected.  
|>    So how does a color TV work?  I understand that it uses only red green 
|>and blue and that red and green mixed together (and surely this would be an 
|>additive process, not a subtractive one) produce yellow.
|>    Does anyone know?

I'm not an expert in retinal physiology and biochemistry, but here's my
recollection of the mechanism.

There are several (three major ones, I think) photochemically active pigments
in the cone cells of the retina (the ones responsible for color vision).
Each pigment molecule is capable of absorbing light quanta, which cause it
to eject an electron and thereby change state from oxidized to reduced (or
maybe the other way around; absorbing a photon causes the pigment to
bleach, or lose its color), thus turning the visual signal into an
electrochemical one.  The electrochemical signal eventually becomes a nerve
impulse on the optic nerve that is transmitted to the brain, where the visual
cortex interprets it as colored light.

The different pigments respond to different degrees to photons of a
particular wavelength (that is, the equilibrium between bleached/unbleached
pigment molecules will be shifted to one side or the other to different
degrees for different pigments).  The relative degree of bleaching of each
pigment is what determines which color will be seen.  Suppose we have 3
pigments (A, B, C) and the degree of bleaching in yellow light (say, one
of a single wavelength is A:20%, B:40%, C:80%.  Now suppose that there's
a red wavelength that produces the bleaching pattern A:18%, B:30%, C:10%
and a green wavelength with the pattern A:2%, B:10%, C:70%.  If you expose
the pigments to both the red and green wavelengths at the same time, and
you get the intensities right, you can get the bleaching pattern A:20%,
B:40%, C:80%.  The eye/brain will "see" yellow even though there isn't a
yellow wavelength photon to be found.

This is a crude example, but it illustrates the principle involved in
mixing pigments to achieve colors.  Dr. Land (of Polaroid fame) discovered
that the choice of primary colors mixed to achieve the spectrum is
fairly arbitrary--they don't have to be red-green-blue or the blue/yellow
red/green complement pair.  Likewise, the number of colors is arbitrary,
although you need at least three for best effects.  What is important is
that the colors you choose bleach the various retinal pigments to sufficiently
different degrees that the visual cortex interprets it as different
colors.

--PSW