Ken and everyone,
Thank you - your answer has been very enlightening.
A synthesis of the reason of some difficulties I've been getting while
scanning.
Not too recent: I remember a post of yours where you stated than 8 bit
were enough. I checked it and found that, for those frames I checked,
they were not enough, in the shadows at least. Add to this, that I'm
just at the threshold of getting enough precision when adjusting curves
/ levels in NikonScan. I had never thought such a 'manual & sight'
precision would be necessary to render acceptable results. This is a
recent discovery ... and skill.
Underlining the differences between digital camera's files and scanned
files regarding curves, provided an answer for several apparent
contradictions I have been gathering while reading the posts at this
list and elsewhere.
Your sentence: " ... a film negative or slide has substantial random
noise (grain) which will create the effect of greater bit-depth than is
actually there." is almost the theory I was looking for. Have only once
read Moose warning me on this, but no theory underneath. This will lead
me to think about what is digitizing, random noise and 'true'
information measured in bytes. You know, in my profession we give equal
truth-value to random noise and to pieces of clear thinking. Audio will
be heard under this new light, too.
Might have been taken for granted, or said thousands of times, or being
such an obvious fact. But just now I'm realizing it's weight, the weight
of a simple theory, like Copernico's. Before, I was not prepared to
understand this.
Same runs for your sentence: " ...in Nikonscan, there is the scanning
bitdepth and then there is the file-save bitdepth." This one is what I
was looking for when discussing why or why not fed Photoshop with the
scanner output, no NikonScan or anything. I could not reach a
conclusion, not enough theory and not enough data. I made my trials, and
attached to NikonScan. The trial was that big white flower on which I
doubted how to chose proper depth of field ... pixels showed better when
processed in NikonScan first - but that was an empirical finding, no
theory underneath.
Random thoughts follow:
- measuring information in bytes, defines what random noise is. Am I too
deterministic in my philosophic cornerstones, that I don't believe in
random facts other than in maths.
- information / data is not equal to truth.
- so Irfanview in another sftwr for scanning - heard of it here at the
list, years ago perhaps. Certainly from Moose, I'm almost sure. Have it,
never tried it. Is it outfashioned now?
- profiling (discussed not long ago) - I should buy a target slide. I
remember having searched the internet for one, but stopped at the moment
when Mr.Ling advised me to profile the monitor first. I'm waiting for
someone to bring home a Spyder Pro ... an used Zuiko may pass through
Customs, but not a new Spyder bought at BH.
My daughter is calling me, must go out now.
Thank you.
Fernando.
Ken Norton wrote:
> Sorry, Fernando, this has been a recent discovery.
> AG
>
--
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/
|