Filmscanners mailing list archive (email@example.com)
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[filmscanners] RE: comments on using NikonScan/PC
- To: firstname.lastname@example.org
- Subject: [filmscanners] RE: comments on using NikonScan/PC
- From: "Jack Phipps" <JPhipps@asf.com>
- Date: Fri, 4 Apr 2003 11:30:20 -0600
- Unsubscribe: mailto:email@example.com
I've waited to weigh in on this since it has been talked about so much but I
guess there is still some debate on Digital ICE softening.
A Digital ICE enabled scanner scans for red, green, blue and infrared. The
infrared scan makes a defect map locating any surface defects (dust, dirt,
scratches, finger prints, film defects, etc.). With this defect information,
software adjusts for the defects, this is the patented Digital ICE process.
While other scanners may employ an infrared scan, only Digital ICE corrects
for the defects _without_ "making up" pixel hue information by cloning using
a nearest neighbor technique or other methods to fill in the identified
errors. In other words, if it is not a surface defect there shouldn't be any
changes to the image. If there is a change in the image it should be in
amplitude only. For example, if you scratch all the way through the
emulsion, Digital ICE will find the defect, but it will not be properly
repaired because there is insufficient information. Now of course there are
some exceptions. Kodachrome can _sometimes_ cause some errors in the defect
mapping process as well as poorly developed film with silver in it.
Now to the softening that appears to happen. Sometimes you eyes can play
tricks on you. When you are evaluating an image for sharpness, you look for
the finest detail. In many cases the finest details are the surface defects.
When you remove the surface defects the image _appears_ less sharp.
Actually, if you look for fine detail, in _most_ cases, it will be there.
If you look at the examples on our website you will find this to be the
case. Look for missing detail. If you want your image to _look_ sharper, one
way is to add noise. In effect that is what the surface defects are, noise.
My recommendation for improving the sharpness is to use the Digital GEM
plug-in. you can move the noise sliders to zero to not remove any grain type
noise and then use the clarity slider to give as much sharpening or softness
as necessary. You can download a trial version at www.asf.com.
The same thing holds true for grain (or random noise) removal. When the
noise is removed the image appears less sharp. With the Digital GEM plug-in
you can get _very_ aggressive with noise removal and actually cause
artifacts. But when the sliders are between the default (50%) and 0% you
should see noise removal without a loss of detail. However, you will see a
reduction in _apparent_ sharpness. That is why we added the clarity slider.
You can add sharpness back or, if you are working on a portrait, (many
people look better with less sharpness), you can reduce sharpness, an
inexpensive way to add a soft focus effect.
Remember, when you are photographing sand or snow or other totally random
patterns, it is difficult to differentiate between signal and noise. So be
careful out there! Like any tool, use it where you need it.
BTW, I'll be in Munich, Germany for several weeks working with our Digital
PIC machine (the one that scans unprocessed film). I may not be monitoring
the group much while I'm gone. Please don't hesitate to contact me directly
if you have any questions, because I will be checking my email. Also, if you
are in the area and would like to see one of our machines up close, email me
and I'll arrange a visit.
I hope this helps.
Applied Science Fiction
>As I understand ICE, it only softens exactly where the dust spot was,
>and not on any other part of the image, so I'd assume the total
>"softening" (if true) is only noticeable when a slide is very dirty or
>has unusually sharp, fine details?
Unsubscribe by mail to firstname.lastname@example.org, with 'unsubscribe
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or