Re: Colour perception

From: Stevan Harnad (harnad@cogsci.soton.ac.uk)
Date: Thu Feb 27 1997 - 18:22:55 GMT


> From: "Liz Lee" <EAL195@psy.soton.ac.uk>
> Date: Thu, 27 Feb 1997 09:02:36 GMT
>
> OK, I've come to terms with the fact that nothing in itself has colour
> (although I still have difficulty accepting this), what I really don't
> get is that if peak absorption for cones is at 419, 521 and 558 nm, how
> do we perceive yellow at 570-590. Plus, does this mean there are
> "colours" at much higher wavelengths that we're unaware of?

To appreciate that colour is in your mind and not on the rose, consider
this: When you put your hand in cold water, is the way it FEELS part of
the cold water, or is it part of you? If you think it's part of the
water rather than you, put your hand in still colder water for a bit,
then put it back into the first cold water: it now feels warmer than it
did the first time (because of adaptation to the colder water): Does
that mean its temperature has changed? Or are you beginning to see that
when the temperature sensitive nerve endings in your skin touch
something of a certain temperature, then that makes you FEEL it as
cold.

The same is true of colour. In a narrow range of the electromagnetic
wave length spectrum, between infrared (which is invisible) and ultraviolet
(which is also invisible) a narrow range of wavelengths look like the
colours in the rainbow.

Now to answer your second question first: Are there "colours" at
wave-lengths that are below and above the visible range?

First, the question is wrong, since the colours are not there in the
visible range either! The visible range simply has the effect of making
you SEE (= FEEL) colours (by activating cones in your retina), just
as in feeling something "cold" or "warm."

Are there wave-lengths above or below our visible range that are visible
to other creatures with minds (i.e., able to feel)? Yes, some insects
and some birds can see wavelengths we can't see, but please don't ask me
what they look like! That's like someone colour-blind (or completely
blind) asking you how colours look!

Now how does a combination of activity in three cones -- the Red-, Green-
and Blue-sensitive ones -- make you able to see yellow?

It turns out that all the visible colours can be perceived from
different combinations of activity of the R, G, and B receptors. Think
of those three as the axes of a 3-dimensional colour space. Each colour
we see is a weighted combination of those three: (A% Red) + (B%
Green) + (C% Blue). "Red" might be a high percentage of Red plus low
percentages of Green and Blue, and so on. Purple might be high on Red
and Blue but low on Green.

You can generate every point in colour space that way, just as you can
generate any point in a 3 dimensional space by weighted sums of
(1,0,0), (0,1,0) and (0,0,1).

[Exercise: how would you get the point (3,4,5) by adding weighted
combinations of the three "basic" points? Are the basic points the only
three points from which you could generate all the points in 3-D space?
No; there are an infinite number of such basic triplets. The only
requirement on any basic triplet like this is that none of the three
is a weighted combination of the other two. Any three independent points
can generate all of 3-D space. The same is true of colour space. The
"generator" colours that you combine (additively) need not be Red, Green
and Blue. Any three colours of which none is a combination of the other
two can generate all of colour space.]

This is not the whole story. There is an opponent relation between
Red & Green and between Blue & Yellow: If you stare at red for a
while, then look at an off-white surface, you'll see its opponent
colour, green. And we have only been talking about Hue. There's
Saturation and Brightness too. But that's the basic story.



This archive was generated by hypermail 2b30 : Tue Feb 13 2001 - 16:24:17 GMT