[Physltest] [Phys-L] Re: Color (was LED mini-flashlight price break)

[John Clement <clement@HAL-PC.ORG>]

There are two reasons why accurate recreation of any color perception is
often difficult.
1. The receptors overlap in color response so that pure spectral colors are
very difficult or impossible to reproduce.   The peaks in the receptors are
actually fairly broad, and they overlap some at practically all frequencies.
In addition one of the responses has an extra peak.  Rainbows and spectra
are qualitatively different from good reproductions of them.
2.  The eye/brain renormalizes the perception across color boundaries so
that the perception tends to correspond to the color of the object rather
than the actual spectral distribution of the reflected light.  See the
Howard Hughes article.

One can demonstrate that 3 colors tend to give fairly good results over a
broad range of conditions, but again there are some dramatic demos which
involve darkened rooms and only two frequencies projected through
transparencies on a screen.  Subjectively the color reproduction can be very
accurate under these specifically controlled situations.  Indeed this is
done with frequencies that are fairly close to each other.  This was
exploited by the original Technicolor system which was only a 2 color
system.

The 3 color model is approximate and can only be demonstrated well under
specific controlled conditions.  Fortunately it can be done well enough for
us to enjoy movies, TV, and pictures, but the reproduction is never truly
accurate.  One way to make it more accurate would be to exploit the color
constancy by shifting the actual color balance to compensate for our
perceptions.  I have not seen any articles about this being done.

John M. Clement
Houston, TX

> As Devil's Advocate, let me argue that color is the perception of the
> interaction of light with three widely overlapping broadband photosensors.
> Stimulating the eye with a) a single frequency near the overlap of two
> sensors b) two separate frequencies near the peak responses of the two
> sensors, or c) a broad range within the response regions of the two
> sensors will all create roughly the same response in the eye.  All three
> of these are "the same color".
>
> That's why three colors (not four or five or two) are "primary".  By
> choosing three inks or three lights that each stimulate primarily one
> receptor, then a rather accurate recreation of any color perception can be
> produced.
>
>
> It's us physicists who messed it up by trying to equate a specific color
> with a specific wavelength ;-)
>
> Tim F
_______________________________________________
Phys-L mailing list
Phys-L@electron.physics.buffalo.edu
https://www.physics.buffalo.edu/mailman/listinfo/phys-l