[filmscanners] RE: over resolving scans

["Laurie Solomon" <laurie@advancenet.net>]

I agree with all that you have said and did not think I was suggesting
otherwise. What I think I was suggesting was that there were going to be
tradeoffs with whatever alternative one selected and that which tradeoff was
accepted as acceptible was in effect a matter of taste.

> My experience is that, if anything, resolution is more
> important for grainy B & W because,
> to me anyway, sharp, small "clean" grain is MUCH more
> attractive than somewhat larger mushy
> grain you get from lower resolution scanning.

That is also my viewpoint; but even so it is also a matter of preference. I
am not sure if I understand fully Anthony's comments on grain or alaising on
true silver halide films being emphasized and more noticable with lower
resolution scans than higher resolution scans.  I have heard numerous
complaints to the contrary.  But his experiences and the others complaints
may also revolve around matters of taste and preference.  He may prefer the
mushier grain appearace to the sharper more emphasized one; whereas others
do not.

Personally, I always scan at the maximum optical resolution that the scanner
is capable of if for no other reason than to give me future flexibility in
terms of manipulations, resizing, and resampling without having to resort to
interpolation most of the time to produce a satisfactory product at an
acceptible size.

> -----Original Message-----
> From: filmscanners_owner@halftone.co.uk
> [mailto:filmscanners_owner@halftone.co.uk]On Behalf Of focus@adnc.com
> Sent: Tuesday, October 08, 2002 1:19 PM
> To: laurie@advancenet.net
> Subject: [filmscanners] Re: over resolving scans
>
>
> ANY digital representation of an analog image, like film, is
> going to be an imperfect
> rendition of the original; the more resolution you have, the
> less you're going to lose, but
> you're never going to get to 100%; more resolution just gets
> you sucessively closer. 3000
> dpi can theoretically represent 60 lpmm, but 60 lpmm at what MTF??
>
> Also, if you scan at less than the max resolution of your
> scanner, you invite downsampling
> artifacts - if you do decide to downsample, you're much
> better off letting PhotoShop or a
> Fractal program do it than your scanner driver.
>
> My experience is that, if anything, resolution is more
> important for grainy B & W because,
> to me anyway, sharp, small "clean" grain is MUCH more
> attractive than somewhat larger mushy
> grain you get from lower resolution scanning.
>
> Is 2700dpi "good enough" in many, many cases? Sure. Do you
> lose anything from scanning at
> higher resolution? Nothing except time and hard drive space.
> (Yes I've read Dan Margulis
> contrary views on the subject - his arguement includes the
> idea that less micro-detail
> sometimes looks better - obviously, if you agree with that
> there IS a good case for lower
> resolution, although you can still do it yourself better in Photoshop.
>
>
>
> Anthony Atkielski wrote:
>
> > Frank writes:
> >
> > > i just read a post on the piezography newsgroup
> > > that stated that the resolution of Tri X is
> > > somewhere in the upper 2000's and that scanning
> > > at a higher resolution than this (i.e.: 4000)
> > > would be "over-resolving". if this is true, then
> > > can anyone comment on this statement as to what
> > > this actually means?
> >
> > There's no such thing as "over-resolving"; you can never
> scan with too much
> > resolution.  However, the higher the resolution, the
> smaller the increment
> > of additional, _useful_ information that you obtain with the scan.
> >
> > The resolution of Tri-X is actually quite good, and you can
> extract a small
> > amount of additional detail by going to higher resolutions
> (beyond the
> > 2000s).  However, I usually don't bother, because the grain
> in Tri-X is so
> > obtrusive that it can be difficult to appreciate the
> resolution, even at
> > 2700 dpi.
> >
> > The grain itself is hard to resolve at 2700 dpi.  Sometimes
> the grain looks
> > a bit more pleasing if you scan at higher resolutions.
> Apparently the grain
> > size is just about right to cause some aliasing in some
> cases at resolutions
> > in the 2000s, making it look larger than it really is.  But
> it's going to be
> > there in any case, so these are fairly trivial details.  I
> still scan at
> > 2700 dpi.
> >
> > Add to this the type of subjects usually shot with Tri-X.
> I use it for
> > handheld work, such as street photography.  Handheld work
> usually won't
> > produce even the resolution available in the 2000s, no
> matter what film is
> > used, so this is one more reason not to scan at very high
> resolution.
> >
> > But the real deciding factor is your scanner.  Just scan at
> the highest
> > resolution available--you can never have too much, but you
> can always have
> > too little.  If your scanner goes to 2700 dpi, scan at that
> resolution; if
> > it goes to 4000 dpi, then scan at 4000.  In my case, I have
> both a LS-2000
> > and a LS-8000, so I use the LS-2000 for grainy films like
> Tri-X and Portra
> > 800 (just to save wear and tear on the LS-8000), and the
> LS-8000 is for
> > fine-grained films and medium-format films (because the
> LS-2000 cannot
> > handle medium format).
> >
> > If I come across a really nice shot on Tri-X, sometimes I
> rescan at 4000 dpi
> > to extract as much as I can from the image.  The LS-8000
> also has better
> > range, which is handy sometimes because many B&W films,
> including Tri-X, can
> > handle an extremely broad range of light and shadow and can
> be as difficult
> > to scan as slides in consequence; in this case, the 4000
> dpi is just a
> > fringe benefit of scanning on the larger scanner.
> >
> > > what is the result of "over-resolving" ...
> >
> > The result is slightly better detail, or significantly
> better detail,
> > depending on the film.  It's much more obvious with Provia
> or Velvia than it
> > is with Tri-X or fast color negative films.  I've scanned
> Portra 800 at 4000
> > dpi, and all I get is better resolution of the obvious
> grain, so now I just
> > stick with 2700 dpi.  I always scan slide films at 4000 dpi, though.
> >
> > And, of course, if you are shooting Technical Pan, even
> 4000 dpi is way too
> > low; a 10,000-dpi scan would be much more reasonable for
> Tech Pan, and even
> > that may miss detail (if you had a really good lens and
> tripod, that is).
> >
> > > ... and if one had to scan a 35mm Tri X neg at a
> > > resolution in the upper 2000's, how would one make
> > > a larger print from that?
> >
> > I think that 2700 dpi is fine for 8x12s and can probably suffice for
> > 16x24-inch images of Tri-X if the print is of good quality.  Higher
> > resolutions can produce _slightly_ better results, but the
> grain is going to
> > make most of the difference irrelevant.
> >
> > > i would assume that up-sampling would be far
> > > worse than "over-resolving" ...
> >
> > Absolutely correct.  More resolution never hurts, but
> sometimes it doesn't
> > help very much (as in this case).
> >
> > But all of this is academic unless, as in my case, you have
> to choose
> > between two or more scanners with different maximum
> resolutions.  If you
> > have just one scanner, then just scan at the highest
> resolution the scanner
> > provides, no matter what kind of film you are using.  You
> may not get much
> > extra with Tri-X at 4000 or 6000 dpi, but if your scanner
> provides it, you
> > may as well use it and get whatever extra detail is there.
> >
> > > ... but i don't know this for a fact. comments anyone?
> >
> > It sounds like another urban legend, and it probably
> appeals to the same
> > photographers who insist on always exposing a film at some
> magic ISO number
> > _other_ than the number actually on the box.
> >
>
> --------------------------------------------------------------
> --------------------------
> 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