Apache-Talk @lexa.ru 

Inet-Admins @info.east.ru 

Filmscanners @halftone.co.uk 

Security-alerts @yandex-team.ru 

nginx-ru @sysoev.ru 




      :: Filmscanners
Filmscanners mailing list archive (filmscanners@halftone.co.uk)

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[filmscanners] Re: SS120 - getting dust out

"Winsor Crosby" <wincros@earthlink.net> asked: "Is there a set of
instructions anywhere for using the Polaroid Dust & Scratch Removal
filter somewhere? I tried it a couple of times and it did not seem to do
very much and took a very long time to do it. The dust and scratch tool
in Photoshop Elements works pretty well for me."

Not that I know of, other than the cryptic Help in the Stand-alone
version (I think it should be "Help!"). After trying the Stand-alone
version, I've settled on using the Photoshop filter version, which I
find easier and quicker--at least you can see the results easier. I
usually scan negatives (Minolta ScanDual II, new WinXP). I set the
filter to do Light Dust, set the Tile Size to between the first two
marks (no higher than the second mark), set Defect Level to the same
general setting, set Mask Size to the left-most mark, leave Adaptive
Filtering off, and set Feathering to the middle mark. I then waltz
through the preview checking what spots are being "corrected" and which
highlights are being "filled". (The effect for each click of the preview
window is practically instantaneous.) I try to set the Tile Size and
Defect Level to minimize the number of "false positive" highlights while
fixing most of the spots and defects. After Correcting the image, I do
the left-hand-waltz through the image at 100% view to Clone (S) and Heal
(J) the remaining spots and scratches while History-Brushing (Y) any
false-positive highlights that need to be reset to pre-cleaning state.
(If I use more aggressive settings on the filter, there are too many
false positive corrections to deal with.)

I do the D&SR filter as the first step in scan processing. At first I
color-corrected the image first, believing that it was better to spot a
corrected version of the image that a raw one. However, I'd sometimes
forget to set the History Brush state to the after-color-correction
state from its default opened-file state, so when I tried to
History-brush out false positives, I'd get the pre-corrected color as a

I have generally been scanning negatives saved as jpeg files, and this
system has worked pretty well. It still takes a couple of minutes to do
a 2500x3800-pixel image, but the D&SR does get most of smallest defects
that are the biggest pain to eliminate.

I recently started a project to scan some family slides taken 50-60
years ago. These were in pretty sad shape, with dust, mildew, etc.,
making them useless in their native state. I have been carefully and
thoroughly cleaning with PEC12 before scanning (and then sometimes
cleaning and scanning again after the first scan still showed lots of
grunge). The D&SR filter has made this scanning possible--without it,
getting all the remaining dust, spots, mildew, etc. would have taken
many times longer than the 5-10 minutes per scan it does--so long I
doubt I'd have the patience to continue through 250 images.

I initially scanned the slides as jpeg images. I'd scan as
~2500x3800-pixel images, spot, retouch, color-correct in Photoshop, and
save as "archival" hi-res images. Then I'd crop to 1024x768-pixels and
sharpen for use as "slide-show" images. This is when I noticed funky
things happening. Any false-positive areas (in this case shadow areas
lightened) were accentuated in the sharpened versions, plus there were
some unusual patterns showing up in the resized images that weren't
noticeable in the larger images. I took these to be an interaction
between the jpeg artifacts, the D&SR filter, and the USM. At that point
I went back and started saving as tiff images from the scanner and have
been much happier with the results.

The D&SR isn't perfect by any means. I don't know how it compares to the
dICE hardware-cleaning approach, since that isn't available on my
scanner. I don't know how it might compare with the Scitex cleaning
software available for real money for their Eversmart scanners. I did
have one slide that was so badly "damaged" with surface cracks that my
scanner gave a useless image. I had that slide scanned on a DS-618 drum
scanner using oil mounting, and virtually all the imperfections
disappeared. But drum scanning the entire 250-300-slide library isn't an

Preston Earle

Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe 
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or 


Copyright © Lexa Software, 1996-2009.