Tim,
Thanks a lot for the info and links.
I found it all very interesting and enlightening.
--
Russ Butler (NJ USA)
Tim Hughes wrote:
> Ken, a couple of comments:
>
> People interested in the human perception issues of color, might enjoy
> reading the long popular article by Edwin Land (Polaroid fame), which
> appeared in Scientific American.(Edwin H. Land, "The Retinex Theory of Color
> Vision," Scientific American, Vol. 237, No. 6, pp. 108-128, December 1977.)
> His theory of color perception is termed "Retinex theory".
> here are a few related simplified links:
> http://people.msoe.edu/~taylor/eisl/land.htm
> http://www.wendycarlos.com/colorvis/color.html
> http://en.wikipedia.org/wiki/Color_constancy
> http://www.wendycarlos.com/colorvis/retinex.html
> http://color.psych.upenn.edu/brainard/papers/retinex.pdf
> http://www.rowland.harvard.edu/organization/land/theory.php
>
> One interesting thing is the brain can do a surprisingly good job of
> reconstructing a full set of colors from a display screen, even if only 2
> types of carefully chosen color pixels are used, in place of 3.
>
> There are so many variations on the more than tri-color camera sensor matrix,
> which have been suggested and patented. A very interesting variation but for
> color screen display use, is using RGBW. (W = white)
>
> Some of the Fuji and other >3 color sensors land up with non-rectangular
> pixel layout. This can mean the resolution is highest on a diagonal not
> horizontal and vertical axis. This is a slight disadvantage apparently, as
> we percieve better sharpness, if the resolution is highest on the horizontal.
>
> A dificulty with the Foveon sensor, is not just their execution and having to
> have a different dematrixing software/hardware, which is more expensive
> because of lower volumes,but the RGB absorbtion curves have very broad
> overlap with one another, making it much harder,to seperate the colors
> accurately. This tends to magnify the noise despite the inherent advantage of
> not throwing away more than 2/3 of photons, as in the normal 2D filter
> sensor. Using another type of vertical filter in a stackerd sensor, as you
> suggest, could theoretically retain this photon (noise) advantage, if it were
> practical to make.
>
> The human eye is similar to the Foveon sensor in that the absorbtion
> spectrum of the different Rhodpsins in the eye are very similar, with huge
> overlap making the brains job much tougher, in sorting out the real colors.
> This is an example of why the eye is a rather imperfect, but highly optimised
> design, despite what old creationists used to claim about godly design
> perfection in the eye!
>
> see curves here: http://www.wendycarlos.com/colorvis/conecurv.jpg
> and discussion of opsins here : http://en.wikipedia.org/wiki/Rhodopsin
>
> Tim Hughes
--
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/
|
