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: Vuescan 7.1.18, LS-30 and Fuji Superia negatives: What in hell am I doing wrong?

Hi everybody,

as you may know, I have written a couple of times about my problems in
getting correctly exposed and naturally colored negative scans out of my
Nikon Coolscan III LS-30 using Ed Hamrick's Vuescan. A couple of days
ago, Ed addressed me via e-mail about the new 7.1.18 version and its
improvements as to negative film profiles. I decided to give it a try
but the results are still... well... I'd better explain in detail.

I have put four files on my web page for everyone to evaluate: 

http://schmode.net/vuescan.jpg is a 337 dpi Vuescan scan of a tiger in
Hagenbeck Zoo, Hamburg (Germany) using the settings to be seen in
http://schmode.net/settings.jpg; most important amongst them are white
balance as color balance method, a gamma of 3 and Fuji Reala 100
(Superia 100 does not exist) as film settings.

http://schmode.net/nikonscan.jpg is the same picture scanned using Nikon
Scan, sRGB, ICE on, analog gain to -0,25 and, as I always do,
comprehensive use of the curve tool in Nikon Scan until the preview
looks right on the money (I cannot get the Vuescan preview even *close*
to this)

I can see that the Vuescan version is sharper but, believe me, this is
due to the 337 dpi (Ed's algorithm as to lo-dpi scanning seems to be
better) - at 2700 dpi, Nikon Scan is film grain sharp with 100 speed
color negative film even with ICE on. As to the colors, Vuescan seems to
*extremely* give away in the highlights part of the histogram and clip
the shadows at that, plus putting a blueish/magenta cast on everything.
I can get better use of the highlights by setting the white point to .1
(everything above would clip them as well) but the blueish cast persists
as well as the shadows clipping. I am aware that I could get some of
this straight with comprehensive post-scan processing but, believe me,
this ain't what I was looking for. Neither am I going to scan, check the
result, correct, then scan again, check again and so forth until the
result is as I can set it in the accurate Nikon Scan preview and
histogram before one single full scan is done.

I have uploaded the original scan http://schmode.net/scan0001.tif as
well, with a link to it on http://schmode.net/scan.htm (to right
click/save in case your browser does not display .tif files) should
anybody be willing and able to evaluate it (those files won't open in
Paint Shop Pro 7 :-(). Meanwhile, I stand disappointed again, waiting
for one of the next versions of Vuescan to (maybe) correct what still
seems wrong with the current one.

Greetings from Germany -


My animal photo page on the WWW: http://schmode.net
Find my PGP keys (RSA and DSS/DH) on PGP key servers
(use "TrustCenter" certified keys only)


Copyright © Lexa Software, 1996-2009.