Re: filmscanners: Scanner resolution (was: BWP seeks scanner)

[rafeb <rafeb@channel1.com>]

At 09:48 AM 6/19/01 -0400, Andy Adler wrote:

>I have to agree with Dan that the Leaf 45 scan is quite visibly different
>than the Nikon (I'm using a HItachi 19" shadow-mask monitor, BTW), and on
>first look does seem superior to Nikon's. The question is whether such
>differences are meaningful at these resolutions, and whether one scan can be
>made to look like the other through some not-too-hysterical manipulation vis
>Photoshop, NikonScan 3.0/3.1 or whatever. I'm also very, very close to
>buying the Coolscan 4000, and admit to being a bit concerned over reports by
>some users that its scans tend to be n the dark side, with limited shadow
>detail. <snip>


That's just flat-out untrue.  In fact, if you run histograms of the 
two images that were discussed this morning, you'll see that the 
scan from the Leaf has a lower median value and lower mean value.

On the Leaf image, codes 0-8 account for 8.5% of all pixels.
On the Nikon image, codes 0-8 account for 0.7% of all pixels.

The differences are 100% attributable to scanner settings, and 
entirely "fixable," with either scanner, at scan-time.

The Nikon image can be made to look like the Leaf image (in 
Photoshop, after the scan) but not vice-versa, since shadow 
detail has been lost in the Leaf image.


rafe b.