Interesting analysis. I think you are making some assumptions about the
performance of scanners, and thus where the data bottleneck is, that may
not be correct. Some partial answers:
1. The range of densities into which a film maps brightness varies
between films, so the comparison you make is potentially more complex.
2. Most film scanners scan at at least 16 bit depth, for 64k levels of
brightness input data for each color. That range is usually then
downsampled to 8 bit, 256 level for output, although if your graphics
software can handle it, the higher level of detail is available. In any
case, the hardware and driver software don't lose any highlight or
shadow detail unless you ask them to. I've tried 16 bit, as both my
scanner and video system can do it, but haven't seen a real difference.
Maybe that's another test to try.
3. Whatever the performance of the scanner at hand, there will be a
better one in the future, and all that detail will still be waiting on
the film.
4. Whatever the theory, it's like the Velvia vs. Provia site shows,
lower contrast carries more tonal detail.
Moose
Daniel J. Mitchell wrote:
It does mean that low contrast film is an
inherently better choice, all other things equal, for those who
digitally process and print their images.
Surely that's only true if you have an infinitely sensitive scanner or if
the higher-contrast film is actually clipping?
Say, for the sake of argument, you have a scanner which has 256 levels of
resolution between black and white. (possibly pessimistic, but I don't know
how well the rest of the steps along the line will preserve higher colour
resolution)
<snip>
So either way you lose something -- it's just a matter of if it's more
important to preserve highlights/shadows or smooth graduations in the
midrange.
The original point is still fair enough, though; if you have a more
sensitive scanner, then the loss of detail in the midrange will be smaller,
and eventually you won't be able to tell it's not there any more -- but are
scanners (and the rest of the chain of image processing) that good yet? I
don't know, to be honest.
-- dan
< This message was delivered via the Olympus Mailing List >
< For questions, mailto:owner-olympus@xxxxxxxxxxxxxxx >
< Web Page: http://Zuiko.sls.bc.ca/swright/olympuslist.html >
< This message was delivered via the Olympus Mailing List >
< For questions, mailto:owner-olympus@xxxxxxxxxxxxxxx >
< Web Page: http://Zuiko.sls.bc.ca/swright/olympuslist.html >
|