Minolta 5400 ICE

[Chris Luneski]

I am trying to get a feel for how effective the ICE cleaning is with the
Minolta 5400, using either the Minolta software or Vuescan. What it does
with negatives is of no interest to me, as I only shoot & scan slides. Using
a reasonably clean slide as a standard, what percentage of the dust and
scratches will be removed by the Minolta ICE?

I am using a Canon FS4000 and am dissatisfied with the current versions of
Vuescan --- artifacts and poor dust removal. The Canon software, Film Get,
is not an appealing alternative, as it produces very noisy scans with its
own peculiar artifacts.

I am thinking of switching to the Minolta, but only if the dust removal is
significantly better than on the Canon.

thanks,

Chris Luneski

 

[Guest]

I am trying to get a feel for how effective the ICE cleaning is with the
Minolta 5400, using either the Minolta software or Vuescan. What it does
with negatives is of no interest to me, as I only shoot & scan slides. Using
a reasonably clean slide as a standard, what percentage of the dust and
scratches will be removed by the Minolta ICE?

I am using a Canon FS4000 and am dissatisfied with the current versions of
Vuescan --- artifacts and poor dust removal. The Canon software, Film Get,
is not an appealing alternative, as it produces very noisy scans with its
own peculiar artifacts.

I am thinking of switching to the Minolta, but only if the dust removal is
significantly better than on the Canon.

thanks,

Chris Luneski

--
(e-mail address removed)
Alamy: http://tinyurl.com/5alew
Stock Connection: http://tinyurl.com/5uqjm
http://www.pbase.com/petphotos

http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=00AMmg

 

[Fernando]

I have the Minolta 5400.
With both Minolta Scan Utility (v1.1.5) and Vuescan (v.8.1.20) I get
very good dust removal, with minimal artifacts.
The problem is that with M.S.U. you are forced to turn on Grain
Dissolver (a light diffuser) to use ICE, and this slows down scans a
lot.
On Vuescan, you can use the two features independently (and you get
effective action even with the "light" setting), but I have some
quality problem with scans from Vuescan (faint strikes across the frame
when I sharpen up). Many other users don't have this problem.

Fernando

 

[Bart van der Wolf]

SNIP

http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=00AMmg

Version 8.1.20 is supposed to have an "improved infrared cleaning on
all scanners".
Haven't tried it myself yet.

Bart

 

[Bart van der Wolf]

Using a reasonably clean slide as a standard, what percentage of the
dust and scratches will be removed by the Minolta ICE?

Depends on the type of dust/scratches, but the Minolta software is
very good at addressing these issues, be it at a significantly longer
scan time.
Here is an example on a deliberatly mistreated (sandpaper+cutter
blade) slide:
http://www.xs4all.nl/~bvdwolf/temp/MDSE_NoGD_NoICE.jpg
http://www.xs4all.nl/~bvdwolf/temp/MDSE_GD_NoICE.jpg
http://www.xs4all.nl/~bvdwolf/temp/MDSE_GD_ICE.jpg
As the filenames indicate, first one without any restoration, with
only the grain dissolver, and with the Grain Dissolver and ICE, all
processed with the Minolta Software that comes with the scanner.

I am using a Canon FS4000 and am dissatisfied with the current
versions of Vuescan --- artifacts and poor dust removal.

Try if the latest (as of this writing, VueScan 8.1.20) improves the
situation for you (although the Canon uses a type of IR exposure that
is apparently hard to utilize).

Bart

 

[Guest]

I too would be very interested to try the latest Vuescan. There are
several versions posted with indicated "better cleaning" and similar
language. The first I tried had a problem with raw files for me, I
believe that's fixed now. Just got back from XMas trip to see our NEW
GRANDSON so haven't done anything for a while, first things first

 

[Alex Stols]

SNIP

Version 8.1.20 is supposed to have an "improved infrared cleaning on
all scanners".
Haven't tried it myself yet.

Bart

I did a short test with 8.1.20 and I couldn't see *any IR cleaning at
all*. Maybe I did something wrong, and I should try again. But as you
said: the Minolta sw is very good though slow.

Regards, Alex
alexUNDERSCOREstolsATxs4allDOTnl

 

[Chris Luneski]

Bart,

That is a very impressive demonstration. Much better cleaning than I can get
with the Canon.

I've tried 8.1.20 without no great success. I have gone back to using
8.0.10 --- I can't get 8.1.15, which was OK, from Ed.

thanks for your help,

Chris Luneski

 

[Bart van der Wolf]

I did a short test with 8.1.20 and I couldn't see *any IR cleaning
at all*. Maybe I did something wrong, and I should try again.
But as you said: the Minolta sw is very good though slow.

I just checked 8.1.20 myself, the IR cleaning is there in the
resulting file, but very mild (I "scanned" from a file, to 48-bit
Tiff, with GD selected). VueScan's cleaning with this scanner is much
less effective, even at the heavy setting. On my LS-2000 the light
setting was usually enough.

Bart

 

[Alex Stols]

I just checked 8.1.20 myself, the IR cleaning is there in the
resulting file, but very mild (I "scanned" from a file, to 48-bit
Tiff, with GD selected). VueScan's cleaning with this scanner is
much
less effective, even at the heavy setting. On my LS-2000 the light
setting was usually enough.

Bart

So maybe I didn't do things the wrong way... I did it much the same
way as you described: just a little 'cleaning'. Using 8.1.16 gave
more 'cleaning'. Stranger and stranger.

Regards, Alex
alexUNDERSCOREstolsATxs4allDOTnl

 

[Guest]

I have the Minolta 5400.
With both Minolta Scan Utility (v1.1.5) and Vuescan (v.8.1.20) I get
very good dust removal, with minimal artifacts.
The problem is that with M.S.U. you are forced to turn on Grain
Dissolver (a light diffuser) to use ICE, and this slows down scans a
lot.
On Vuescan, you can use the two features independently (and you get
effective action even with the "light" setting), but I have some
quality problem with scans from Vuescan (faint strikes across the frame
when I sharpen up). Many other users don't have this problem.

Fernando

I finally got around to trying v.8.1.20, and am totally unimpressed
with it's cleaning. Contrary, to several "what's new" claims of
improved cleaning, I could see no difference, compared to older
version.

Sure, it's fine for fairly clean, recent film, but for older, trashed
stuff, it's virtually useless, compared to ICE. I'm using same scanner
as you, incidentally.

ICE manages to clean my problematic slides very well, doing a much more
complete job, and much more "seamlessly". Vuescan leaves so much, and
leaves obviously softened areas. I'm really getting tired of even
trying new releases, it's a time consuming waste of time.

 

[Don]

So maybe I didn't do things the wrong way... I did it much the same
way as you described: just a little 'cleaning'. Using 8.1.16 gave
more 'cleaning'. Stranger and stranger.

Hi Alex,

VueScan's so-called "cleaning" has been buggy for a long time.

If you, nevertheless, insist on using VueScan I'd suggest scanning
with ICE using native software and then postprocessing with VueScan
afterwards. Although, if you choose that route, you'd be much better
of using dedicate image editing software instead.

Talking about strange... When I made the mistake of trying out VueScan
over a year ago the medium and high "cleaning" produced sharper images
than low cleaning. Logically, it should be the other way around
indicating major image fudging (read: corruption) to try and mask the
bugs.

I got flamed for daring to point this out a while back but apparently
the bugs are still there.

Don.

 

[Don]

Sure, it's fine for fairly clean, recent film, but for older, trashed
stuff, it's virtually useless

That applies to all VueScan features. If the image is fine to start
with even VueScan will manage to scan it.

However, for anything even remotely difficult it falls flat. Long time
ago I posted some dark Kodachromes to illustrate the point. The result
was the reluctant addition of individual Analog Gain setting which the
author previously refused to implement because "you don't need it".
:-\

Vuescan leaves so much, and
leaves obviously softened areas. I'm really getting tired of even
trying new releases, it's a time consuming waste of time.

That's an understandable and expected reaction but what puzzles me is
when people continue to put up with such programming incompetence
irrationally hoping things will get better when obviously the software
is permanently riddled with bugs as these messages repeatedly show.

I mean, it may be OK for casual use, but not if one cares for any
semblance of quality.

Don.

 

[Fernando]

I finally got around to trying v.8.1.20, and am totally unimpressed
with it's cleaning. Contrary, to several "what's new" claims of
improved cleaning, I could see no difference, compared to older
version.

Sure, it's fine for fairly clean, recent film, but for older, trashed
stuff, it's virtually useless, compared to ICE. I'm using same scanner
as you, incidentally.

I don't know what to say: Vuescan IR removal works well on my slides,
at least since 8.0.9, without visible softening. I use "Light" setting
and 48 bit scanning. I can't see great difference vs. Minolta Scan
Utility, apart from the slight softening in M.S.U. due to the G.D.
Anyway, my slides are well kept: there is dust and the occasional small
scratch (not visible on the lightbox but well visible at 5400dpi), but
nothing serious.
Probably the method adopted by Vuescan is not effective on more
damaged/dirty films...

Fernando

 

[Alex Stols]

Hi Alex,

VueScan's so-called "cleaning" has been buggy for a long time.

If you, nevertheless, insist on using VueScan I'd suggest scanning
with ICE using native software and then postprocessing with VueScan
afterwards. Although, if you choose that route, you'd be much better
of using dedicate image editing software instead.

Talking about strange... When I made the mistake of trying out
VueScan
over a year ago the medium and high "cleaning" produced sharper
images
than low cleaning. Logically, it should be the other way around
indicating major image fudging (read: corruption) to try and mask the
bugs.

I got flamed for daring to point this out a while back but apparently
the bugs are still there.

Hello Don.

Yes, I stopped using VS ---
What I did was buying a SE 5400 and it works like a charm! Even the
scans of my 45-years old Kodacolor negatives, badly scratched
sometimes, dusty, drying artefacts, are sometimes so good that I can
relive the time they were made!!

Regards, Alex
alexUNDERSCOREstolsATxs4allDOTnl

 

[Bart van der Wolf]

(e-mail address removed) wrote: SNIP

I don't know what to say: Vuescan IR removal works
well on my slides, at least since 8.0.9, without visible
softening. I use "Light" setting and 48 bit scanning.

Yes, it works, but not as well as one would need for *badly* damaged
films.
Here's a VueScan (8.1.20) example to complement my earlier Minolta
Scan Utility results:
http://www.xs4all.nl/~bvdwolf/temp/VS8120_NoGD_NoIR.jpg
http://www.xs4all.nl/~bvdwolf/temp/VS8120_GD_NoIR.jpg
http://www.xs4all.nl/~bvdwolf/temp/VS8120_GD_IRlight.jpg
http://www.xs4all.nl/~bvdwolf/temp/VS8120_GD_IRmedium.jpg
http://www.xs4all.nl/~bvdwolf/temp/VS8120_GD_IRheavy.jpg

I can't see great difference vs. Minolta Scan Utility,
apart from the slight softening in M.S.U. due to the G.D.

There shouldn't be much softening of the image due to the GD, apart
from graininess reduction and some softening of debris.

Anyway, my slides are well kept: there is dust and the occasional
small
scratch (not visible on the lightbox but well visible at 5400dpi),
but
nothing serious.
Probably the method adopted by Vuescan is not effective on more
damaged/dirty films...

That's my impression as well. After all, filling in what's gone
already can be done in many ways, and I find that for general use the
Minolta ICE is a bit too strong, although for those pathetic cases it
is very effective.

Bart

 

[Fernando]

Yes, it works, but not as well as one would need for *badly* damaged
films.
Here's a VueScan (8.1.20) example to complement my earlier Minolta
Scan Utility results:
http://www.xs4all.nl/~bvdwolf/temp/VS8120_NoGD_NoIR.jpg
http://www.xs4all.nl/~bvdwolf/temp/VS8120_GD_NoIR.jpg
http://www.xs4all.nl/~bvdwolf/temp/VS8120_GD_IRlight.jpg
http://www.xs4all.nl/~bvdwolf/temp/VS8120_GD_IRmedium.jpg
http://www.xs4all.nl/~bvdwolf/temp/VS8120_GD_IRheavy.jpg

Eheh, quite a damage!
I wonder what could be gained by using a sealed fuild carrier for
damaged (and grainy) film.
In the Spring I should manage to get an LS-9000 (my MF shots outnumber
my 35mm shots by a large margin now), and will ask Aztek a quote for
their fluid-type carrier.

There shouldn't be much softening of the image due to the GD, apart
from graininess reduction and some softening of debris.

Yes, I know; still, I'm seeing edge softening... I've even run Imatest
with and without GD, and found a (altough small) MTF drop by turning
GD on (I posted the results here, a couple of months ago maybe).
The slightly lower sharpness did not come from misfocusing, either: I
performed multiple (20 or so!) Imatest runs with very small focus
adjustments until I maximized the MTF figures.
Maybe the light scattering with the diffuser on lowers the overall
contrast (but this should not affect the MTF), maybe the diffuser
lowers the SNR because the longer exposure builds up more shadow
noise, maybe the diffuse light source gives some more flare within the
optical system...?

That's my impression as well. After all, filling in what's gone
already can be done in many ways, and I find that for general use the
Minolta ICE is a bit too strong, although for those pathetic cases it
is very effective.

Unfortunately, for me is mostly academic talk, since I'm forced to
Minolta Scan Utility (still streaks and low DMax with Vuescan. Did
those problems go away for you with latest versions?).

Bye!

Fernando

 

[Wilfred]

Yes, it works, but not as well as one would need for *badly* damaged films.
Here's a VueScan (8.1.20) example to complement my earlier Minolta Scan
Utility results:
http://www.xs4all.nl/~bvdwolf/temp/VS8120_NoGD_NoIR.jpg
http://www.xs4all.nl/~bvdwolf/temp/VS8120_GD_NoIR.jpg
http://www.xs4all.nl/~bvdwolf/temp/VS8120_GD_IRlight.jpg
http://www.xs4all.nl/~bvdwolf/temp/VS8120_GD_IRmedium.jpg
http://www.xs4all.nl/~bvdwolf/temp/VS8120_GD_IRheavy.jpg

It looks as if VueScan doesn't deal with the white area that typically
appears in the middle of scratches (it probably doesn't show up as black
in the IR channel). Perhaps ICE has a smarter algorithm that isn't
fooled by thinking that a white area with black borders is a clean area
inside a damaged area - as VueScan seems to do.

 

[Don]

Hello Don.

Yes, I stopped using VS ---
What I did was buying a SE 5400 and it works like a charm! Even the
scans of my 45-years old Kodacolor negatives, badly scratched
sometimes, dusty, drying artefacts, are sometimes so good that I can
relive the time they were made!!

Regards, Alex
alexUNDERSCOREstolsATxs4allDOTnl

Hi Alex,

In the meantime I got a Nikon LS-50 but I'm still not happy with ICE
on my 1980s Kodachromes. Also, the dynamic range is still a problem.

Anyway, I recently got the Software Developer Kit from Nikon so - time
permitting - I'm hoping to write my own software one of these days.

Don.

 

[Jan_from_Maine]

There shouldn't be much softening of the image due to the GD, apart
from graininess reduction and some softening of debris.

What is the difference between using GD and using something like
NeatImage? Can't grain be considered a noise source? If I'm off in my
thinking, would appreciate some clarification.