Apache-Talk @lexa.ru 

Inet-Admins @info.east.ru 

Filmscanners @halftone.co.uk 

Security-alerts @yandex-team.ru 

nginx-ru @sysoev.ru 




      :: Filmscanners
Filmscanners mailing list archive (filmscanners@halftone.co.uk)

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

filmscanners: Trying Nikonscan 3.1

I've just tried Nikonscan 3.1 on my LS30 with a slide and a frame of Fuji
Superia 100.  Please bear in mind that's just two frames of film, so it's
hardly exhaustive.  I scanned the same frames using Vuescan 7.1 to produce
48bit colour and 16bit IR files of the same frames.  Here's what I found...

o Good news!  No jaggies so far.  But I have yet to try any of the frames
which were always a problem before.
o Nikonscan is waaaaay slower in the actual scanning process than Vuescan.
The only reason it was faster at getting an image from itself to Photoshop
is that it's TWAIN and dumped the image straight in.  Even so it's probably
slower overall.  I have no idea why focussing and exposure adjustment are so
much slower in Nikonscan than Vuescan.
o GEM and ROC are disabled.  Possibly they work better with 14 bits from the
newer scanners, but mostly I expect it's to encourage people to upgrade. :)
o Dust and scratch removal in Nikonscan and Vuescan was interesting to
compare.  Small scratches in some cases were minimised better by Nikonscan,
but generally the two programs were similar here.  Interestingly, Vuescan
was better at removing some large spots - they disappeared completely but
were still around in Nikonscan.  Both programs seemed equally effective at
dust removal.
o Colour - bizarre results!  The slide scan looked nice in Nikonscan but
closer investigation showed more colour and shadow detail in the Vuescan
output.  This isn't necessarily surprising since on the LS30 Vuescan
delivers 10bits per channel instead of only 8 from Nikonscan.  With the
slide, both programs produced nice results easily.  Nikonscan was a bit
quicker in terms of setting up the scan, but the end result from Vuescan was
better.  With Superia 100, the difference was staggering.  The output from
Nikonscan was oversaturated (and this coming from me - I like saturation
generally, but here the colours were really un-natural) and slightly
posterised due to the bit depth loss.  Tonal shading in the Vuescan image
was much smoother, and although the image looked flat initially, autolevels
and autocontrast in PS quickly fixed it.  So the Vuescan image required more
tweaking, but the end result was far better than the Nikonscan output.
o Interface - Nikonscan has changed quite a bit in the way it is laid out,
but the tools are actually quicker to get to.  I am still at a loss as to
why it defaults to 1350dpi instead of 2700dpi.

Fingers crossed that the jaggies have gone!  I'm impressed with the
improvements in Nikonscan 3.1.  However, for the LS30, Vuescan is still my
program of choice.  With other scanners that output more bits like the
LS2000 or more recent models, the differences may be more subtle.


PS I notice mention in the readme of two additional film carriers - one for
16mm film and one for medical slides.  Has anyone seen/tried these?  The
latter one could be handy for staff at the hospital where I work


Copyright © Lexa Software, 1996-2009.