Olympus-OM
[Top] [All Lists]

Re: [OM] IMG: Faviana and Reymundo

Subject: Re: [OM] IMG: Faviana and Reymundo
From: Ken Norton <ken@xxxxxxxxxxx>
Date: Wed, 13 Jun 2012 20:56:03 -0500
Tina,

I'm struggling understanding what is going on with your Kodachrome 64
scans. I'm suspecting that we have a combination of issues you are
fighting and not realizing that you are fighting. In the Faviana and
Remundo photo, we have several issues, but they seem to be universal
to some of the other K64 shots too. Here's the observations behind my
stating this:

1. Kodachrome is an extremely fine-grained film. It is perfectly
capable of a double-truck spread without being much more than a touch
sandy. Sometimes, we'd see an interaction between the half-tone
screens and the grain, but we didn't get the globular mess that we saw
with Ektachrome.

2. The colors tend to have good tonal separation in Kodachrome 64.
What I'm seeing is specifically the oranges getting blocked up.

3. The blues are definitely off. If the boy's hair is blue in the
slide then I suspect the film suffered from heat or x-ray damage.
Meanwhile, the blues desaturated elsewhere. This is evident in the
greens which turned olive. (I'm usually used to greens turning bluish
gray in Kodachrome). Yet, what I think is a wood stump behind the boy
is gray. (Traditional Kodachrome rendering of wood, whereas
Fujichromes will turn wood into some shade of tan).

4. The slide does have a slight look that it's been projected and
suffered a little bit of damage from that.

I supect that your scan came pretty close to the visual examination of
the slide when viewed over a NON-color-balanced, full-spectrum viewing
table. Unfortunately, if the lightsource behind your viewing table is
not full-spectrum, you're going to suffer from misinterpreting the
color rendering in the slide. For some reason, from my own experience,
Kodachrome is easier to misread than other films. Not in overall
tonality, but in how some colors are rendered.

The color profile for the scan is off. It may match a color target or
even other slides, but this one may have other issues which have
caused the profile to be incorrect.

The scan has the look that it was underexposed during the scan process
and everything was boosted through the profile or level management.
You mention that you use Silverfast. I'm not very familiar with that
program, but usually use Vuescan for my scans. A side benefit of
Vuescan is that Vuescan's IR cleaning algorithm now works on
Kodachrome! Anyway, this looks like we're pulling up the exposure of
the scan in the computer instead of increasing the exposure time
during the scan. Unlike Ektachromes and Fujichromes, which pull up ok,
Kodachrome scans more like a B&W film and doesn't handle shadow
lifting. If fact, it is almost like trying to work with a digital
camera image that has been pulled up 2-3 stops.

Judging from the grain pattern and other artifacts, I'm guessing that
you are using a 4000dpi Nikon scanner. Mine is a V-ED. The Nikon
scanners tend to find grain where there is no grain. It's big on grain
aliasing. One way around this is to overscan (multi-pass). This will
reduce the scanner's ultimate resolution, but much of that loss can be
recovered with care.

Which leads me to the next point. I'm seeing some artifacts in the
bokeh of the Faviana and Reymundo photo which is a characteristic of
the 35/1.4, but is exaggerated by USM sharpening. Earlier, I was
mentioning an overscan and multi-pass sharpening method I use on some
of my slides. The big problem with the Nikon scanner in single-pass is
that we have a slight resolution loss thanks to the scanner's optics,
yet we also have the grain aliasing. It seems pretty natural to throw
a 2 or 3-pixel USM on the scanned image. What that does with some of
the tight bokeh traits of that particular lens is that it makes the
bokeh harsh and unnatural. Instead of a nice penumbra, the
transitional fade pattern will develop steps.

I'm not sure what advice I'd pass on yet. Most of your images in your
pbase account have the smoothness I'd expect to see from a K64 slide
(especially in the railroad gallery), but these Guatamala slides just
aren't quite there. If I were to recommend trying something I'd say to
scan "The SC40-2" (Image 964s) again (to achieve identical results of
the original scan) and apply those same exact settings to the "Faviana
and Reymundo" slide and see what happens.

My point in all this is that this particular slide in question did not
digitize in a manner that I would ever expect to see from K64. What I
did is idenitfy a different image which did digitize in an expected
manner and am recommending that you attempt a baseline adjustment with
that as a starting point for this slide. It might also turn out off,
but at least it would be worth a try as part of the problem
identification. I do notice that other images seem to suffer from
blues in the shadows, so we might definitely have a profile problem.

Ken

-- 
Ken Norton
ken@xxxxxxxxxxx
http://www.zone-10.com
-- 
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/

<Prev in Thread] Current Thread [Next in Thread>
Sponsored by Tako
Impressum | Datenschutz