Re: a link to some interesting illusions

[Larry Woolf <larry.woolf@GAT.COM>]

Stating that the eye-brain system perceives color independently of
illumination is in general false (although it is true in various
situations). This can be easily proved by anyone who has bought an article
of clothing in a store and then noticed that the color appears very
different when taken outside.

Generally, those involved in color use and measurement quantify color using
three factors:
1. The illuminant.  These illuminants are typically daylight (denoted
illuminant 65 or C), incandescent light (illuminant A) and fluorescent light
(illuminant F2 or F11)
2. The reflectance spectrum of the object being viewed.
3. The observer.  The CIE color systems generally use either the 2 degree
standard observer or the 10 degree standard observer.

There are color systems that then utilize these three factors to generate a
3-dimensional coordinate to specify a color.  Color differences are then
specified as the distance between these coordinates.

The most commonly used color systems are :
CIE xyY -the color system often used by color scientists and those
investigating color vision.  This is the system mentioned earlier by JC.  In
this color system, uniform differences in color distance do not correspond
to uniform differences in color perception
CIE u'v' - a color system in which uniform color distances correspond to
uniform color perception differences.  This system is most widely used to
compare color gamuts of visual display systems such as TVs and computer
monitors.
CIEL*a*b* - this is also a uniform color system and is the most widely used
in the color industry. Here L* is the lightness, positive a* corresponds to
redness, -a* corresponds to greenness, b* corresponds to yellowness, and -b*
corresponds to blueness. Note that this system is based on the opponent
nature of our color perception - we do not see reddish greens or yellowish
blues. So they (red/green, yellow/blue) are opposed to each other.  In this
system, the coordinates have a nice physical meaning.

To convert between color systems, including RGB and CMY, go to:
www.easyrgb.com

Color scientists use color coordinates to communicate in a quantitative way
with each other and to ensure reproducibility and quality control of their
products. The color difference in the L*a*b* system is calculated by "delta
E" = square root of the sum of delta L^2 + delta a^2 + delta b^2. A nice
example of a three dimensional coordinate system that is actually used - and
a nice example of the three dimensional use of the Pythagorean theorem. We
cannot perceive color differences that differ by a delta E of 1 or less.

Our visual perception of color also depends on the surrounding color, our
experiences (white walls are white, leaves are green in a wide variety of
lighting conditions - color constancy... which can also lead to colored
shadows) and time dependent effects (complementary afterimages for example.)
Also, our perception of color depends on the gloss of the surface.

Color and color vision are complex subjects.  My favorite references are:

Principles of Color Technology Third edition by Roy S. Berns.  Many pictures
and diagrams - clear explanations.
Human Color Vision, Second edition by Peter K. Kaiser and Robert M. Boynton.
A standard reference - and includes many do-it-yourself experiments
demonstrating the complexity of color vision in boxed in sections.

Nice references are also available from color instrument companies such as
BYK Gardner or Minolta.

Larry Woolf;General Atomics;San Diego CA
92121;Ph:858-526-8575;FAX:858-526-8568; www.ga.com; www.sci-ed-ga.org

> -----Original Message-----
> From:  John Clement
> Sent: Thursday, February 06, 2003 10:41 PM
> Subject: Re: a link to some interesting illusions

> http://www.hhmi.org/senses/b140.html has a good description of the way in
> which the eye-brain system perceives the "true" color of a pigment
> independently of illumination.