[sane-devel] Digital ICE support

Major A andras@users.sourceforge.net
Mon, 1 Dec 2003 19:01:14 +0000

> The alternative might be to try UV cleaning, though I suspect that the
> idea of illuminating precious negatives with UV might not be welcome.

I guess that's related to the fact that the UV light would be absorbed
by the emulsion -- which means UV cleaning would probably not work
very well.

> See the URL below for more information. Personally I would just have
> adjusted the relative position of the IR CCD line to account for it's
> moved position due to chromatic aberration.
>    http://www.canon.com/technology/detail/flatbedscanner/fare/index.html
> In fact the more I think about that idea the more I like it. Why stop at
> the IR line, have the positioning for all four lines set so that you can

Goodness me. This thing scans with white light, but has an "optical
path adjustment glass" which gets move in and out of the light
path. That means your IR channel gets scanned separately, so you have
registration problems anyway. I doubt the chromatic aberration makes
that much of a difference in this case.

After all it's only a transparency scanning solution, I thought Canon
had added an IR channel to their flatbeds in reflective mode...

> use cheaper optics and not have to worry about the chromatic aberration.
> Well you are worrying about it, but rather than spending money on
> reducing it, you just use it to your advantage, as the RGBI elements
> of the CCD in a line scanner cannot all be in the same position in
> the first place. That is assuming you are not using Foveon(tm) style CCD
> sensors of course.

Good idea, it would require very narrow sensitivity bands for the RGB
channels, though. Which means you would probably need a LED light
source, like the ones in the Coolscan scanners.


Major Andras
    e-mail: andras@users.sourceforge.net
    www:    http://andras.webhop.org/