Digital BW, The Print

Hi Peter,

I don't know if this will make you feel any better, but your findings are
depressingly familiar to mine with medium speed B&W film. You hit the nail on
the head when you say that a darkroom print doesn't show the same problem --
there is something happening in the interaction between the scanner and the
film. Out of interest, if you compare 10x8 prints, both wet process and
digital, is the digital print appreciably sharper? If so, you could soften the
digital print slightly to average out the speckle and still end up with a
print about as sharp as the traditionally made one. Not ideal, I know.

Anyway, here's my 2p theory about what causes the problem we're both seeing.
Consider the grain structure in the developed film, on a microscopic level. It
consists of clouds of stringy-looking metalic silver arranged very randomly.
Kodak have a shortish article about all this, including some nice
photomicrograph illustrations:

http://www.kodak.com/US/en/motion/students/handbook/sensitometric6.jhtml

From that article:

"At magnifications where the eye cannot distinguish individual particles, it
resolves random groupings of these particles into denser and less dense areas.
As magnification decreases, the observer progressively associates larger
groups of spots as new units of graininess. The size of these compounded
groups gets larger as the magnification decreases, but the amplitude (the
difference in density between the darker and the lighter areas) decreases. At
still lower magnifications, the graininess disappears altogether because no
granular structure can be seen"

Clearly each sensor site in your scanner's CCD is looking at a fairly small
area. It can't resolve individual grains but it can certainly see small enough
clumps that the amplitude will vary quite widely. (Referring to the article,
each pixel of the sensor is behaving very much like the small aperture used to
scan films when testing for RMS granularity.)

I believe we are seeing this effect right across our images, but we are less
concerned about it in the high and mid tone areas because in traditional
processes we always see a bit of grain there. However, when we see it in the
shadows, our immediate reaction is to assume it's scanner noise. In fact, when
scanning negs, the signal to noise ratio should be at its best in the shadow
region of the image.

There are a couple of things that can help reduce the problem. We can soften
the image post-scan as I suggested above, but none of us want to lose any hard
won sharpness. Using a scanner with a more diffuse lightsource can make a big
difference. Scanners such as Epson flatbeds do tend to give slightly softer
images than dedicated film scanners, but more importantly their flat
lightsources don't throw the grain structure into such sharp relief, meaning
that the local variations in density are less dramatic.

Sorry if this doesn't help you get better results out of the equipment you
already have. I mention these issues in the hope that discussion of them will
eventually lead to better solutions being found. I have seen major exhibitions
of digital prints by "big name" artists where shadow speckle has been visible,
so it's not just us!

Best regards,

-= mike =-

-----Original Message-----
From: Peter A. Klein [mailto:pklein@...]
Sent: 15 August 2004 10:35
To: DigitalBlackandWhiteThePrint@yahoogroups.com
Subject: [Digital BW] Shadow noise in scans


Folks:  I frequently encounter shadow noise in my scans. I can 
usually work around it, but I'm wondering if there is something I can 
do--perhaps while scanning--to minimize it.

Have a look at this:
http://users.2alpha.com/~pklein/temp/noise.jpg

This picture was taken with an M6TTL and CV 50/1.5 Nokton, on Kodak 
T400CN rated 400.  It was taken in a dim reception hall, but the 
subjects were illuminated by a shaft of light from some nearby large 
windows, and I exposed for them, not the background.  The scanner is 
a Canon FS-4000, run at 4000 dpi and driven by VueScan.  I scan to 16 
bit TIFF, and use Picture Window Pro in 16 bit mode for my editing.  
The various versions of the picture are shown at 1/3 actual pixels.

The leftmost image is of the scan pretty much out of the scanner (I'd 
spotted and cropped, but that's all).  You can see the noise in the 
darkest areas.  This is the kind of noise I am often plagued with.  
It tends to happen more often in available light pictures than well-
lit outdoor shots.  It doesn't show much on a 5x7 quadtone print, but 
is quite prominent on an 8x10.  And if I get the same picture printed 
in a wet darkroom, the noise isn't there. 

The center image is what I ended up with after curves, contrast, and 
sharpening (the latter using a mask so only the people were 
sharpened, not the surrounding area.  I've dropped the background to 
nearly jet-black, but there are still lots of noise speckles.  And 
I've lost some gradation in the girl's hair.

The righthand image has been noise-filtered with NeatImage before 
curves, etc.  I used it sparingly so it wouldn't soften 
or "plasticize" the image too much.  It's helped, but the noise still 
shows up--more in a print that on the screen.  I've found that most 
of the time, when I use NeatImage or Picture Window's Despeckle 
feature, I can only get rid of some of the noise before the image 
gets messed up.

What might I try to get rid of this noise?  Is there anything can I 
do to minimize it during the scan?  I've tried the multiple exposure 
feature in VueScan, and it doesn't really help with this type of 
noise.  

BTW, here's the whole picture.  This one is a reduction of the Costco 
CD Jpeg I got when the film was developed.  They went way dark, so 
you see no shadow noise (or detail).  
http://users.2alpha.com/~pklein/JonAntionWedding/Show/JonAntionWedding
/slide_13.jpg

Thanks,
--Peter Klein
Seattle, WA


==============================================================================
This message is for the sole use of the intended recipient. If you received
this message in error please delete it and notify us. If this message was
misdirected, CSFB does not waive any confidentiality or privilege. CSFB
retains and monitors electronic communications sent through its network.
Instructions transmitted over this system are not binding on CSFB until they
are confirmed by us. Message transmission is not guaranteed to be secure.
==============================================================================