Archived posting to the Leica Users Group, 2000/03/11

[Author Prev] [Author Next] [Thread Prev] [Thread Next] [Author Index] [Topic Index] [Home] [Search]

Subject: Re: [Leica] opinion on film scanner
From: "Henning J. Wulff" <henningw@archiphoto.com>
Date: Sat, 11 Mar 2000 22:55:15 -0800

About 8 months ago after the Polaroid 4000 came out, I took it and a Nikon
Coolscan 2000 home for a couple of days.

The Polaroid did produce slightly higher resolution. This was noticeable,
but not a huge difference. The Nikon, even without the multiscanning,
produced less shadow noise. Again, noticeable but not hugely. With
multiscanning, the difference increased in favour of the Nikon. To test the
machines, rather than the software at this level, I did only raw 48 bit
scans and did the post processing in Photoshop.

The Nikon software was a lot better than the software that came with the
Polaroid. Third party sofware would probably even things out a bit more.

Finally, I was able to buy a refurbished LS-2000 with a 90 day Nikon
warranty for under $1000US. To that I added the batch scanner that Polaroid
doesn't offer, and now have a machine that cost less than $1500 that can do
50 slides without my interference. If I do batch scans, I generally set the
machine up to do a 4x sampling and save them as 48bit files. About 2 hours
for 50 slides. For Velvia I use 16x sampling. I only use the ICE for older,
dirty slides as there is a definite loss of resolution; more so than the
difference between the 4000dpi Polaroid and the 2700dpi Nikon. Then I go
into Photoshop to fix them up. A very useful setup. If I wanted anything
better, I would get a refurbished Howtek at about the same price as a new
Imacon, but a lot higher quality.

I had an LS-20 before. Passable for negs, but the density range is marginal
for low density slides and useless for Velvia or Kodachrome.

Density range quoted by manufacturer's is largely imaginary. 3.4, 3.6 or so
is often quoted, but no CCD machine produces more than 3.0, and we're
talking $50,000+ here. The Nikon has a true dynamic range of less than 2.5.
Photo multiplier machines can achieve around 4.0, which in the terms of the
consumer market would be called 5.5 or some such.

The point is that the quoted dynamic range and Dmax capability of the
machines are figments of someones imagination, and have no real common
reference standards on which they are based, so you must go and test them
yourself or trust someones test results. A good way to test is to get a
high quality drum scan of a very long tonal range Velvia, or better yet
Kodachrome II or 25 (it has a more demanding toe with more information) and
compare that to the best scan you can get from the scanners you are
interested in. If you think you have gotten everything out of a slide that
a LS-2000 or Artix 4000 or Imacon can give you, this will be an eye opener!

   *            Henning J. Wulff
  /|\      Wulff Photography & Design
 /###\   mailto:henningw@archiphoto.com
 |[ ]|     http://www.archiphoto.com