-----Original Message-----
From: filmscanners_owner@halftone.co.uk
[mailto:filmscanners_owner@halftone.co.uk]On Behalf Of Anthony Atkielski
Sent: Monday, April 01, 2002 3:15 PM
To: snsok@swbell.net
Subject: [filmscanners] Re: Nikon Super Coolscan 4000ED and
sharpness/focus
Gary writes:
> Without having my scans in front of me
> to compare, I'd have to say my scans
> are equally as soft; MAYBE just a tad
> bit sharper.
Interesting. I don't really consider this sample scan soft--it is typical
of all three of the scanners I've used, in terms of overall contrast between
adjacent pixels. Examination of the original Velvia transparency under a
microscope reveals only a slightly higher microcontrast, but not any
additional detail. At 4000 dpi, you are potentially resolving just over 78
lp/mm; Velvia is rated at 80 lp/mm at 1:1.6 contrast, and twice that for
1:1000. This is probably 1:32 contrast and handheld, so the scan is (in my
estimation) showing over 90-95% of what is on the film (and more in areas of
lower contrast).
This is only a two-pass scan; contrast might have been better if I had made
more passes. Note also that this scan represents only 0.56% of the entire
image; that is, if you display this on a 640x480 monitor with a 15"
diagonal, the corresponding size of the entire image would be 13'9" on a
side--large enough to hide a locomotive behind. So it's not too shabby;
certainly if you were looking at the entire frame, you would be unable to
see details this small in the image without superhuman eyesight. The entire
image (greatly reduced) is visible at
; a red square marks the section I
posted at full size.
The slight difference between my scan and the original could be the result
of:
1. Digital ICE (used during this scan)
2. Film not perfectly flat (I don't have a glass holder yet)
3. The fact that the image was taken handheld (braced, but handheld--no
sign of camera movement in the image, though)
4. The fact that the image was rotated slightly (although Photoshop is very
good about preserving detail in 16-bit scans during rotation)
> BTW, all my slide film is Kodak Elite II
> 100.
I don't have a data sheet for it, but I think it is safe to assume that it
is no sharper than Velvia.
Now you know why I got myself a small microscope; it helps me figure out
whether a scan is _really_ getting all the data or not.
> Some, yes. When I noticed this problem with
> the Nikon I scanned in a couple of slides
> that I had scanned with my old HP Photosmart
> scanner and compared. I felt that the HP
> scans were sharper.
Maybe the HP scanner does an unsharp mask.
> I emailed a friend who has the LS2000, asking
> if he's seen the problem. He's seen it as well,
> but since this is the only scanner he's had he
> figured it was "normal" and would work on his
> scans in Photoshop.
I have an LS-30 and an LS-2000 as well, and they both produce scans that are
no better than this, in terms of contrast in adjacent pixels. In fact,
every raw scan I can recall seeing has been like this, no matter what the
scanner. I suspect that desktop scanners in particular are very challenged
by extremely high contrast at frequencies near their sampling frequencies;
they can resolve differences in adjacent pixels, but with a noticeable loss
of contrast at the upper limit, which can appear as diminished sharpness
compared to the original image on film.
Even so, at 4000 dpi on 35mm film, you are getting scans of 5600x3700, which
is plenty for all sorts of applications, as long as you don't crop a lot or
view the images from way up close.
> Did the 2000 has the same narrow depth of
> field that you say the 4000 has?
It's hard to tell. I never noticed any variation in focus on the LS-2000
for mounted slides, but I did notice it sometimes for negatives, when one
edge of the negative was not being held flat by the carrier.
----------------------------------------------------------------------------
------------
Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe
filmscanners'
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title
or body
----------------------------------------------------------------------------------------
Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe
filmscanners'
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or
body
