Re: Grain aliasing: do some neg films suffer less than others?

["Alan Rew" <arew@xxxxxxxxxxxxxxxxxx>]

----- Original Message -----
From: Ralf Schmode <rschmode@gmx.net>
To: <scan@leben.com>
Sent: Sunday, March 31, 2002 5:11 PM
Subject: Re: Grain aliasing: do some neg films suffer less than others?


> Alan Rew schrieb:
>
> > Having acquired my new Nikon Coolscan IV ED (i.e. LS40), which I use
for
> > scanning negs, I am now seeing grain aliasing artifacts more clearly
> > than with my old Canoscan FS 2710.

<snip>

> Nikon Scan 3.x, in my opinion, has degraded a lot in comparison to
2.5.
> If you intend to keep on using Nikon's software, try to change the ICE
> rather than the GEM settings. ICE affects grain as well but does so in
a
> more subtle way than GEM.

Ralf,

Thanks for your detailed reply. It's good to know there are fellow Nikon
LS40 sufferers, and that there <might> be an answer that doesn't involve
throwing the scanner out of the window ;-)

BTW I'm using the latest Nikon Scan download from the Nikon Europe web
site. It calls itself version 3.1.2.

Following your advice, I've now tried Nikon Scan with the ICE setting at
'fine', with no GEM, and it does indeed reduce the granularity of the
grain shown on a slide scan (Fujichrome Sensia II, ASA 100) compared to
'normal' ICE, albeit at the cost of a softer image. So for some
applications, such as portraits, this seems to be the better setting to
use. A pity that most of my pictures are landscapes and architecture,
where sharp edge detail is beneficial ...

> BTW, carefully check the histogram of your
> negative scans - Nikon Scan 3.x has a mean tendency of clipping
shadows
> which usually shows itself in a rather abrupt slope at the histogram's
> left side.

All my histograms so far have had free space at either end, so I've seen
no clipping (yet) with either negs or slides. As I scan at 12 bit
resolution, massaging the tone in PS isn't a problem in terms of
posterisation etc.

> Earlier versions (not working with the LS-40) had a "prescan
> mode" control that would completely avoid the clipped shadows if set
to
> "lo-cont neutral" rather than "auto" - unfortunately, Nikon removed
that
> control in later versions, making Nikon Scan almost useless for my
> negative scans.

Thanks for the warning. I wonder why your neg scans are clipped and mine
aren't?

> > So... I was wondering if switching film brands would make a
beneficial
> > difference.
> >
> > Previously I've used Fujicolor Superia, ASA 100 or 200.
>
> If you're using these two, I strongly doubt switching to another film
> would solve your problem. Superia 100 certainly is some of the finest
> grained stuff around, and the improvement when using Reala will most
> likely be marginal.

Well, I've just got some Fujicolor Superia 'Reala' ASA 100 back from
development, so I'll know some time this week whether it scans any
better than the 'standard' Superia.

<snip>
> It is optics rather than resolution - the Nikon LS-40 is sharper than
> anything I have used so far, including Minolta Scan Elite II and Nikon
> LS-30. If the ICE/GEM juggling as mentioned above does not satisfy
your
> needs, there is only one final solution which actually is the one I am
> using right now:
>
> Go to http://www.hamrick.com and get yourself a copy of "Vuescan", Ed
> Hamrick's well known scanning interface for different types of film
> scanners (one software fits all scanners :-)).

Thanks, I bought VueScan a couple of years ago for my Canon. I was never
happy with the colours it produced, and so gave up with it. It probably
didn't deal with Fuji film very well.

I've now downloaded the latest version to try with the Nikon.

> As to your problem, two
> settings (apart from that negative has to be chosen as film type) are
> essential: "Infrared clean" -> "medium", "grain reduction" -> "light".
> Scan one of your negatives and look if the overall appearance of that
> scan is closer to what you have in mind, at least as to the grain
issue.

Well I've tried VueScan with slide film so far (I'm desperate to scan a
particular slide at present) and it does indeed produce better scans (as
far as grain is concerned) than Nikon Scan does with either of its ICE
settings.

With Nikon Scan, the 'normal' ICE isn't quite enough in terms of grain
hiding and 'fine' is a bit too much, whereas with VueScan the 'medium'
setting is a nice compromise. Well done, Ed ;-)

> If yes, you'll have to acquaint with the numerous controls of Vuescan
to
> get the colors about right (the main goal of Vuescan is to produce
"raw"
> scans with a maximum of image information in them; those usually need
a
> bit more of post-scan color correction).

Yes, I can confirm that the colour accuracy from VueScan with the Nikon
is just as woeful as it was with the Canon.

I've tried telling VS to use various profiles for the destination, such
as Adobe RGB and Wide Gamut RGB, but with no apparent benefits. To be
fair, I should start a separate thread on the Colour Management issues.
My main concern at present is the <grain> issue.

It's worth pointing out BTW that the one thing that Nikon Scan <has>
done remarkably well for me is produce accurate colours straight off the
bat for negs and slides, with virtually no changes needed in Photoshop
except tone tweaking. Amazing! If only VS was as good as this. BTW I
specify the 'Wide Gamut RGB (compensated)' in Nikon Scan's colour
management settings for best results.

<snip>

> The link in my signature shows where my LS-40 scans can be found (only
> the newer ones are LS-40 stuff; check each individual page for what
the
> photo was scanned with).

Nice photos! I'm now encouraged that the LS40 can deliver the goods if
treated with patience (and driven differently).

Thanks for your help,

Alan Rew


-
Turn off HTML mail features. Keep quoted material short. Use accurate
subject lines. http://www.leben.com/lists for list instructions.