--- Original Message -----
From: "Robert Meier" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Wednesday, February 14, 2001 10:17 PM
Subject: RE: EOS OT - New Canon Film scanner
>
> --- Paul Secinaro <[EMAIL PROTECTED]> wrote:
> > There has been some discussion of this on
> > the comp.periphs.scanners newsgroup on Usenet, and
> > my
> > understanding is that the 4.2 dmax claim is based on
> > the maximum theoretical dynamic range you could ever
> > get out of a 42-bit scan, neglecting noise and other
> > factors. In practice I suspect the real dmax will
> > be
> > somewhat less, but nonetheless I think the new Nikon
> > scanner will be unmatched in its price bracket,
>
> It seems like you are correct because, if I got the
> definitions right, log10(2^(42/3))=4.21. If you use
> multiple scans for one final scan you can reduce noise
> a lot. I don't think you can reduce it to 0, though.
> If you lose just two bits you are down to 3.6.
While your assumption my be correct, it might also not be correct because
if, for example, you had a positive transparent original with a measured
highlight density of say 0.4 and a measured shadow density of say 2.40,
giving a dynamic range of 2.0 or if you prefer 100-1 contrast range, you
could still have a 42 bit grey level range which would offer you a more
continuous tone like distribution of the various original density changes.
or in other words, a very nice bit of subjective reproduction.
You could, on the other hand, have a range of say 0.40-3.40 or if you like a
1000-1 contrast range, and still use a 42 bit interpolator. What you would
see is a more aggressive series of tone jumps across a reproduced image.
Now which of the two options would you prefer. I know what I would chose and
that would be numerous undo and accept the limitations placed on the
contrast option.
I would guess that what really matters is Nikons interpretation of the term
42bit.
What really matters would be their device's sensitivity to the lamp source
because again if you wished to reduce high density noise all you need to do
is increase the intensity of light passing through the original and onto the
focal plane. At that point you would wish to know how the device would
respond to this light level at the low density end of the scale.
My guess would be that the response plot against light intensity levels
would show some kind of a sensitive curve with a most decided roll off at
each end of the response scale. In other words the CCD device would have a
non linear response to the light source.
If that is the case then tone distortion would take place at both ends of
the repro density scale and it would be down to the software to correct this
state of affairs.
I know nothing about these consumer film scanners, coming as I do from a
high end drum scanner background where the photo multiplier rules OK and
density response curves were very well understood and accounted for.
If there is anyone on here who knows the technical details of these tiny
wonders currently available at a price around 100th of the price of an
original Crosfield. Hell or Dainippon drum scanner then I would be most
interested in hearing their views.
Richard Corbett
*
****
*******
***********************************************************
* For list instructions, including unsubscribe, see:
* http://www.a1.nl/phomepag/markerink/eos_list.htm
***********************************************************
