One very important factor in scanner testing which hasn't even been mentioned, let
alone taken into account in these tests is TEMPERATURE.
One manufacturer's data sheet states that CCD dark current doubles for every 9 degrees
Celsius rise in temperature, and this seems entirely in line with the
known behaviour of semiconductor devices in general.
I'd be willing to bet that this means we can pretty much expect a doubling of scanner
noise for every 9 degree rise as well. (Astronomers go to great lengths
to cool their CCD detectors, and keep them at a steady temperature BTW.)
Since the tests were done spread over considerable time and distance, it's unlikely
that the scanners used were at anything like the same temperature.
Therefore no valid conclusions about noise can be drawn from these tests.
The 'protocol' also makes no mention of allowing any warm-up time for the scanner's
light source, and again, light output varies with temperature, for LEDs as
well. Even with the same scanner, there may be a noticeable difference in a scan taken
immediately after switch on, and another done a half hour later.
Really, unless all these tiny details are taken into consideration and controlled,
there's very little objective conclusion that can be arrived at from this
group scan test.
Regards, Photoscientia.
====================================================================
The filmscanners mailing list is hosted by http://www.halftone.co.uk
To resign, <mailto:[EMAIL PROTECTED]> with UNSUBSCRIBE FILMSCANNERS in the
title, or UNSUBSCRIBE FILMSCANNERS_DIGEST if you are reading the Digest.
