Hello Pieter,
One thing you are not factoring in to this issue is the output side.
When the output is digital, you have the same basic problem. Each
pixel is only one color. What you are really referring to is a
dithering pattern. All inkjet printers do this, monitors do this and
I believe digital mini-labs do this. So in fact, the color doesn't
have to be faked as much as it has to be patterned. The downside to
this is that certain patterns (especially man-made) could come out
looking wrong. The natural random nature of film grain tends to hide
this rather than accentuate it. I don't think the Foveon crowd has
quite as much advantage as you think. They still have to create a
dither pattern from the sensor data as each pixel can only store 1
color.
Using film as a beginning but moving it to digital output is not much
different than the Foveon, capturing all three colors at 1 pixel point
but then creating a dither pattern out of it. Either the scanner or
Foveon chip do this. I suspect that the layout pattern of the
CCD/CMOS chip pretty much regulate this.
In the end, it all comes out in the wash. The only real comparison
would be between a purely analog film process vs a digital
capture/output.
My local labs no longer do analog. That means that my film is at a
disadvantage. It is subject to their scanner/software limitations.
The only alternative is to scan and manipulate the images myself.
Food for thought.
--
Best regards,
Bruce
Wednesday, December 17, 2003, 9:19:10 AM, you wrote:
PN On Tue, Dec 16, 2003 at 06:27:38PM -0600, William Robb wrote:
From: Pieter Nagel
Oh, I wasn't hoping to get any more quality out of the tiny APS sensor
with 2/3 faked colour.
I presume you have an ist D?
If so, you know how wrong this statement is.
PN I did not mean that as a slur on the istD specifically, I was referring to
PN the difference between film current CCD's in general.
PN Specifically, I was referring to the fact that each pixel of the CCD has a
PN red, green or blue filter and can detect only one of the three, the other
PN two need to be interpolated, ergo. 2/3 faked colour.
PN I do not ascribe mystical properties to digital imaging algorithms just
PN because digital is supposedly always better than analog. Therefore,
PN even though I concede that the interpolation of the colours might in
PN practice work fine, I do not for any moment believe that the interpolation
PN algorithm can always end up with the colour value the pixel would have had
PN if it had been able to sense the other two colours.
PN Yes, I have an istD, and I am quite happy with the colour - although I
PN must add that I have only had it for a few days now.
PN But I can't help wondering (as a theoretical curiosity) how much better
PN the colour rendition would be if each pixel could sense red, green and
PN blue simultaneously. Users of a Foveon chip could comment.
