filmscanners: A solution to softening using Digital ICE on LS4000

[Barbara & Martin Greene <martbarb@earthlink.net>]

I've posted a number of queries regarding the problem I've had with the loss
of detail due to the softening effect in using Digital ICE with the Nikon
Super coolscan 4000.  I was so distressed by this that I was seriously
considering trading it in for the Canon Canonscan FS 4000, which uses FARE
for dust removal with no detectable loss of sharpness, albeit it does not
remove scratches as ICE does.  There is no doubt that the LS 4000
exaggerates any dust or scratches on a slide or negative, making it
essential to use ICE.  Many members of the forum have responded to my
queries with helpful comments and lots of information.  A couple of
respondents wondered if there were a way to correct for the softness using
Unsharp Mask in Photoshop.

After much pondering and testing a came up with a solution which I'm pleased
with, and, as a result, I am happy with my Nikon LS 4000 and definitely
keeping it.  I have always found it almost impossible, even after reading
everything written on the subject, to determine the ratio between Amount,
Radius, and Threshold, and to do all that with any sense of having gotten
the best ratio between these three parameters in order to arrive at an
optimaly sharp image and print.  My solution is to use Nik Sharpener Pro,
which does an excellent job almost automatically, providing a choice between
three levels of sharpening.  What I now do with my scans made on the LS 4000
using ICE is as follows.  I make one scan with ICE and another without.
Then, in Photoshop, I use the Nik-Sharpener on the image made without the
ICE, thus obtaining what I consider to be the most satisfactory image.  I
then go to Unsharp Mask and make note of the level of Amount, Radius, and
Threshold recorded.  Next, I open the image made with ICE and set Unsharp
Mask using the same levels obtained from the sharpened image made without
ICE.  I compare the two images and increase Unsharp Mask> Amount so that the
image with ICE matches the image made without it.  Doing this, I was not
able to tell the difference In the quality of detail between the two images.
This may sound like an extremely complicated way to go about this, and, I'm
sure that with practice I can simplify the process.  Of course, if one is
expert at sharpening, it could probably be done even more easily.  The main
point is that I now feel that there are ways to use ICE and still obtain
optimal images and prints.

Martin