Chronology Current Month Current Thread Current Date
[Year List] [Month List (current year)] [Date Index] [Thread Index] [Thread Prev] [Thread Next] [Date Prev] [Date Next]

Re: a link to some interesting illusions

The subject of color perception is extremely complicated and actually goes
far beyond the simple models presented in most physics classes all of which
are only valid for special cases.

As an example the statement that you can stimulate your eye with a mixture
of colors to simulate pure yellow is not really accurate. The receptors for
each color have a broad response curve with large overlap. You can not
stimulate the red and green receptors with red and green light without also
stimulating the blue receptor. As a result you can not actually achieve the
same stimulation as you can with the pure spectral yellow sodium lines. The
approach that recognizes this is embodied in the CIE chromaticity diagram
which is a closed horseshoe shape with pure spectral colors on the curve and
white in the center. http://www.yorku.ca/eye/ciediag1.htm I personally
have noted that pure spectral colors look different from all reproductions
and that to my eyes they almost vibrate compared to other colors.

However the reality is that it is possible to fool the eye in various ways
so that you think you are viewing true colors. Indeed the original
Technicolor process used only 2 colors red and green. Then because the
pictures were a bit blue deficient they flooded the screen with blue light
between frames and you automatically added in the blue where needed. Many
of the schemes to reproduce full color with 2 primaries need darkened rooms.
I have not seen a good demo, but I understand it is possible to use 2
closely spaced frequencies of light and still achieve remarkably good color.

I saw the demonstration given by Edwin Land at an APS meeting many years
ago. In it he had a series of different color matted papers. Then he had 2
spotlights on each side, one with a blue filter and the other red. He then
adjusted them so that the red paper on one side reflected the same spectral
balance as the blue paper on the other. However we still saw each paper
with the correct color. He then proceeded to explain how one could
construct an electrical network which simulated this effect, the Retinex
Theory of color vision.

Essentially the simple 3 color theory only works when you adjust the
conditions correctly, and is difficult to demonstrate accurately in a
consistent manner. The eye-brain compares the differences in colors at
boundaries and then deduces what the color of an object is. This also works
for gray scale. Land demonstrated the same effect for different shades of
white/gray/dark papers with a white spotlight on one side and the angle
adjusted so dark paper near the spot reflected the same amount of light as
white paper on the other side.

Incidentally according to the CIE diagram it is possible to pick a variety
of primary colors and still cover just as much color space as the current
red, green, blue colors in monitors. Indeed the red, yellow, blue
subtractive primaries of artists may be sufficient for many applications
which do not involve photographic quality reproduction. Indeed red, cyan,
yellow can be used as primaries.

One of the best ways to demonstrate the additive color model is to use a
monitor and the demo at
http://www.yorku.ca/eye/ciediag1.htm the author kindly gave me a modified
version that starts out with the 3 ovals not overlapped. Students can mix
the colors. Then they need to look at the screen with a magnifying glass
and observe that the 3 colored dots are emitting light. You can also make
fairly good primary filters using an inkjet printer and transparency
material.

John M. Clement
Houston, TX




The 590 nm yellow sodium line is a pure color, a single frequency (ok,
it's a doublet). Your eyes clearly see this, right? It's just yellow
light. You could also stimulate your eyes by mixing the right combination
of red and green light and see the exact same color. Your eyes cannot
tell the difference.

Sam

Robert Cohen wrote:

Thank you for the responses. I did indeed get confused by
the statement "blue, cyan, green, yellow and magenta are
NOT seen in the rainbow" because, for some reason, I only
saw the answer, not the question.

I still have a question about "yellow is a mixture of
green and red light". I agree that the perception
of yellow CAN BE a mixture of green and red light but
couldn't yellow also be "just yellow" (i.e., a single
frequency that excites the red/green cones in just
the right way for us to perceive yellow)?

____________________________________________
Robert Cohen; rcohen@po-box.esu.edu; 570-422-3428; http://www.esu.edu=
/~bbq
Physics, East Stroudsburg Univ., E. Stroudsburg, PA 18301--

Sam Sampere
Syracuse University
Department of Physics
Syracuse, NY 13244
315-443-5999