On Tue, 15 Jul 2003 14:07:17 GMT, "Mike Ruskai"
wrote:

I know of no articles on the topic, but since blue light has the shorter
wavelength, it's going to register a hit more often than red or green to
begin with without any special "boosting" effect.

I don't think you can make this general statement. There are 5 different
chromophores in the human eye, all tuned to different wavelengths. The pigment
found in blue-sensitive cones, rhodonine-9, has only 56% the QE of the red
pigment and 70% of the QE of the green pigment. Since the pigment concentration
found in all the cones is the same, this means that the blue cones are
intrinsically less sensitive than the red or green cones (in addition to their
much lower density.)

It is perfectly possible to make a good RGB image using a sensor with a weak
blue response. It merely requires placing a higher weight on the blue signal
when combining the individual colors. Neural networks are very good at precisely
this kind of weighting. I see no reason that we shouldn't have good color vision
even though our RGB imaging system doesn't have intrinsically equal sensitivity
across each band.

_________________________________________________

Chris L Peterson
Cloudbait Observatory
http://www.cloudbait.com