So in the end (Minolta 5400)...

[Fernando]

Unless you have very unusual requirements I don't think you really
need to oversample. Also cleaning is not needed unless you have
dirty originals.
Just make sure you use 16 bit depth. Try a couple of scans with
1x sampling, no gd and no cleaning and compare them to what you
are doing. I doubt you'll see much difference.
Lastly unless you are planning to make 18x enlargements or cropping
a lot you can scan at a lower resolution.
Downsampling in your image editor or the scanner produce about the
same result and the editing goes a lot faster on smaller files.
If you do sample at less than 5400 try to use a submultiple like
2700 or the scanning software will have to do lots of calculations.

Excuse me Robert, why you continue to assume that I do what I do
without knowing much about it? It's not the first time...
No offense meant or taken, but I use film scanners since 1997 (Canon
FS-2700), owned 4 different filmscanners and 2 TPU-equipped flatbeds;
I purchased the 5400 for the very reason that it has multisampling,
5400dpi and a decent dynamic range. I need clean shadows from my
Velvia and Astia, and I need all the resolution I can pull from my
shots, for the prints I need to do.
So I need the scanner to work at the best of its specs, that is,
5400dpi and 4x multisampling and, of course, 16bits/channel.
You're a very knowledgeable guy, and I respect you very much; but
please, at least read through all the message before assuming that I'm
a newbie or some kind of plug'n'play kid... thank you for your
contribution anyway!

Fernando

 

[Wilfred]

Fernando wrote:

1) I can't find a reason for the manual focus knob. Any situation
where you found it useful?

I thought in some cases it might be more accurate than autofocus. But
now that there are more results on
http://www.jamesphotography.ca/bakeoff2004/scanner_test_results.html,
I'm not so sure about this anymore.

2) How can I check the firmware version of the scanner?

In the scan utility, move your mouse over the 'explanation' area in the
top right corner and the firmware version will be shown.

3) It becomes quite warm after some use. Much more than my SS120.
Yours, too?

I don't know how hot an SS120 can get, but I observed nothing
extraordinary with my DSE 5400

 

[Fernando]

I thought in some cases it might be more accurate than autofocus. But
now that there are more results on
http://www.jamesphotography.ca/bakeoff2004/scanner_test_results.html,
I'm not so sure about this anymore.

Thank you; I should have been more clear: I was not asking about
manual focus vs. auto focus, but focus knob vs. slider-driven manual
focus. Seems to me that manual focusing with the knob is less accurate
(a bit more "fluctuation" in the focus bar during the action).

In the scan utility, move your mouse over the 'explanation' area in the
top right corner and the firmware version will be shown.

Thank you very much! Says 1.10. AFAIK, installing the Scan Utility
takes automatically care of the firmware upgrading processing, right?
I've installed Scan Utility v.1.1.5, so I guess firmware v1.10 should
be the latest...

Fernando

 

[Dierk Haasis]

1) I can't find a reason for the manual focus knob. Any situation
where you found it useful?

Not yet; I'd rather shift the autofocus point and let the scanner
refocus.

2) How can I check the firmware version of the scanner?

Open Minolta's software and hover the mouse over the big grey field to
the right, just under the title/menu bar.

If you've downloaded the current version of the software/firmware you
find a folder within the installation directory containing the newer
firmware; it shows the version.

3) It becomes quite warm after some use. Much more than my SS120.
Yours, too?

Warm: yes. Warmer: don't know.

 

[Hecate]

Ok, here we are!
5400 up'n'running, tried Vuescan 8.0.6 and Minolta Scan Utility 1.1.5.
First impressions: the scanner is very, very slow with my typical
settings (5400 dpi, 4x multisampling, ICE), expecially with Scan
Utility for it engages the Grain Dissolver with ICE... not a big
problem for me, anyway.

Generally, I find a 2x scan is more than enough, so you can save some
time there.

Vuescan does not work very well with this scanner.

I find GD useful at 5400.

I'll download and try Silverfast 6 Demo: if it works, it's a must have
for me: we'll see.

Yes, I want to try that as well when I get time. Tell us if you like
it ;-)

1) I can't find a reason for the manual focus knob. Any situation
where you found it useful?

Auto doesn't always work. That may be why you've found that some of
the images with GD are softer than expected.

2) How can I check the firmware version of the scanner?

Good question ;-)

3) It becomes quite warm after some use. Much more than my SS120.
Yours, too?

Let's just say my heating bills have gone down... ;-)

 

[Bruce Graham]

Vuescan does not work very well with this scanner.

new version....
What's new in version 8.0.7
Fixed problem with black level on Scan Dual IV

I don't have a 5400, just reporting the Vuescan release notes.

 

[Wilfred]

(...) I should have been more clear: I was not asking about
manual focus vs. auto focus, but focus knob vs. slider-driven manual
focus. Seems to me that manual focusing with the knob is less accurate
(a bit more "fluctuation" in the focus bar during the action).

Yes, it has some hysteresis. On the other hand, the slider seems tolack
direct interaction with the hardware. It could be a marketing thing,I
don't know. Anyone? Bart?

Thank you very much! Says 1.10. AFAIK, installing the Scan Utility
takes automatically care of the firmware upgrading processing, right?

Yes, but it only works if your scanner is turned off during installation
of the DSU.

I've installed Scan Utility v.1.1.5, so I guess firmware v1.10 should
be the latest...

I think so, yes.

 

[Wilfred]

Hecate wrote:

Vuescan does not work very well with this scanner.

I tend to agree. I hope Ed will find some time to solve remaining
issues, such as capturing the full dynamic range, and to implement
correct interpretation of the infrared channel that my 5400 produces. I
need VueScan for negs, because the Minolta software tends to clip the
highlights and shadows.

 

[Olaf Ulrich]

You guys won me over: ordered a 5400,
should arrive tomorrow.
[...]

Ok, here we are! 5400 up'n'running, tried
Vuescan 8.0.6 and Minolta Scan Utility 1.1.5.
First impressions: The scanner is very, very
slow with my typical settings (5400 dpi, 4x
multisampling, ICE) ...

Why do you routinely use 4× multi-sampling? Try using
single-scan and see if the shadows lose detail. I guess
they won't, at least not to any noticable degree. I found
multi-sampling expensive but fruitless in most cases. I'll
use it only in really severe cases.

Vuescan just can't drive it to its maximum
capabilities: streaking in the shadows (it
even streaks with BW neg films, on dense
areas) ...

Do you also get streaks when using Scan Utility?
I don't.

By the way, I found g.d. to soften details
considerably, in a way I cannot fully recover
by sharpening.

Indeed? How strange! Maybe that's a focusing issue?
Have you tried manual focusing *after* engaging ICE
and Grain Reduction?

Moreover, Scan Utility does crash
occasionally, expecially when used
at 5400 dpi with 4x engaged.

Oh ... I used to use Scan Utility 1.1.2 and currently I am
using Scan Utility 1.1.3 -- it has NEVER crashed on my
machine so far (Windows XP, 512 MB RAM, connected
via FireWire).

1) I can't find a reason for the manual
focus knob. Any situation where you
found it useful?

I feel focusing is a tad more accurate at times when done
manually.

2) How can I check the firmware
version of the scanner?

You got the answer already (move the mouse pointer over
the upper right output area in the Scan Utility window).
Firmware v1.10 is the latest. It originally came with Scan
Utility 1.1.3 and hasn't been updated as of now; it also
comes with Scan Utility 1.1.4 and 1.1.5.

Scan Utility 1.1.2 came with firmware v1.9.

3) It becomes quite warm after
some use. Much more than my SS120.
Yours, too?

It becomes quite warm indeed but not what I'd call 'hot.'

Olaf

 

[nodeal]

I tend to agree. I hope Ed will find some time to solve remaining
issues, such as capturing the full dynamic range, and to implement
correct interpretation of the infrared channel that my 5400 produces. I
need VueScan for negs, because the Minolta software tends to clip the
highlights and shadows.

Are these the same problems Hecate is having? VueScan is a product with
no specs, and therefore no warranty.

 

[Wilfred]

Are these the same problems Hecate is having?

I don't know. Hecate said that he or she also has problems with infrared
cleaning but I don't know what his problems are exactly. Searching this
newsgroup with Google didn't uncover any more specifics either.
The problem of not capturing the full dynamic range has been reported by
Bart van der Wolf.
The problem that I'm having with IR cleaning is that VueScan isn't
always successful in distinguishing defected areas from intact areas,
and thus sometimes blurs intact areas as well. See my other post about
this today.
Perhaps Hecate can jump in here and tell us what his or her problems are.

VueScan is a product with
no specs, and therefore no warranty.

I don't expect warranty. Besides, I'm a long-time user (since version 5
or so) so if there would have been any warranty it would have been void
by now. It used to work well with my previous scanner, a Minolta Scan
Speed. But at that time, the Minolta software was worse than it is nowadays.

 

[Hecate]

I don't know. Hecate said that he or she also has problems with infrared
cleaning but I don't know what his problems are exactly. Searching this
newsgroup with Google didn't uncover any more specifics either.
The problem of not capturing the full dynamic range has been reported by
Bart van der Wolf.
The problem that I'm having with IR cleaning is that VueScan isn't
always successful in distinguishing defected areas from intact areas,
and thus sometimes blurs intact areas as well. See my other post about
this today.
Perhaps Hecate can jump in here and tell us what his or her problems are.

I's a she ;-)

Had IR problems. Also had stability problems. Tried it for a short
time (a week) and gave up. I really don't have the time to debug
someone's software for them when it impacts a crucial area of my work.
Especially when the same problems have appeared in more than one
version. (The only beta programs I ever take part it are for The Bat
email client (because if there's a problem I can always use Outlook
(blech!) till it's fixed and ThumbsPlus because they fix problems
promptly). Vuescan seems, at least to me, to be a rolling beta.

 

[Ed Hamrick]

Had IR problems. Also had stability problems. Tried it for a short
time (a week) and gave up. I really don't have the time to debug
someone's software for them when it impacts a crucial area of my work.
Especially when the same problems have appeared in more than one
version. (The only beta programs I ever take part it are for The Bat
email client (because if there's a problem I can always use Outlook
(blech!) till it's fixed and ThumbsPlus because they fix problems
promptly). Vuescan seems, at least to me, to be a rolling beta.

Yes, there's some kind of problem with the infrared channel
on the Minolta 5400. Lots of people complain about it, but
nobody appears to be interested enough to help me diagnose and
solve the problem. I'm probably just going to delete 5400
support entirely.

Regards,
Ed Hamrick

 

[Wilfred]

Yes, there's some kind of problem with the infrared channel on the
Minolta 5400. Lots of people complain about it, but nobody appears
to be interested enough to help me diagnose and solve the problem.
I'm probably just going to delete 5400 support entirely.

That's not fair, Ed. I sent you a few sample files (raw scans), I think
in January, on which the problem was obvious, but I never received a
reply in which you specified why they were not useful to you.
What I could offer as a 'diagnosis' was put in a message to this
newsgroup yesterday:

(quoting myself)

It seems VueScan's cleaning algorithm isn't very successful in
separating the defected areas from the clean areas because the IR
channel produced by some 5400s (including mine) is very low in
contrast. If you have Photoshop CS you may want to try my actions to
perform the separation manually by setting the sliders in a Levels
dialog: http://www.vandervegte.com/Actions.html. Apparently, Nikons
produce a more contrasty IR channel, though I never made the
comparison.

BTW I have the impression that the problem occurs with Fuji emulsions in
particular (both slides and negs).

If you could uncover how VueScan distinguishes the defects from the
clean areas in the IR channel I may be able to offer some more concrete
help. I've been wondering if VueScan just selects all pixels that are
darker than a certain value. If so, how is this darkness value defined?
Is it fixed or based on the distribution of greytones in the IR channel?
Or does VueScan perform a levels adjustment to create a mask for the
cleaning filter? If so, does the levels adjustment have a fixed setting
for the black and white points to make the defects black and the clean
areas white? And if so, is the levels ajustment perfored with fixed
values on a 0-255 (or equivalent) scale?

I offered help on this before and I would appreciate it if you'd accept
it this time.

 

[nodeal]

That's not fair, Ed. I sent you a few sample files (raw scans), I think
in January, on which the problem was obvious, but I never received a
reply in which you specified why they were not useful to you.
What I could offer as a 'diagnosis' was put in a message to this
newsgroup yesterday:

(quoting myself)


BTW I have the impression that the problem occurs with Fuji emulsions in
particular (both slides and negs).

If you could uncover how VueScan distinguishes the defects from the
clean areas in the IR channel I may be able to offer some more concrete
help. I've been wondering if VueScan just selects all pixels that are
darker than a certain value. If so, how is this darkness value defined?
Is it fixed or based on the distribution of greytones in the IR channel?
Or does VueScan perform a levels adjustment to create a mask for the
cleaning filter? If so, does the levels adjustment have a fixed setting
for the black and white points to make the defects black and the clean
areas white? And if so, is the levels ajustment perfored with fixed
values on a 0-255 (or equivalent) scale?

I offered help on this before and I would appreciate it if you'd accept
it this time.

As stated before, VueScan is a product with no specs, and therefore no
warranty. There is not even a product description on what VueScan can do
at the site. Many buyers find the user's guide so lacking that they have
to figure out the user interface by trial and error. VueScan's attitude
is to just buy it and see if you can figure out how to use it and what
it can do for you. If you can't, you must be an idiot. Even for a long
time supporter such as yourself, don't expect to get any objective
support and response. Never mind trying to find out how it works
internally.

 

[nikita]

"" Had IR problems. Also had stability problems. Tried it for a short
time (a week) and gave up. I really don't have the time to debug
someone's software for them when it impacts a crucial area of my work.
Especially when the same problems have appeared in more than one
version. (The only beta programs I ever take part it are for The Bat
email client (because if there's a problem I can always use Outlook
(blech!) till it's fixed and ThumbsPlus because they fix problems
promptly). Vuescan seems, at least to me, to be a rolling beta.""

I just coulden't agree more....

nikita

 

[Ed Hamrick]

As stated before, VueScan is a product with no specs, and therefore no
warranty.

This makes no sense - the VueScan User's Guide and Getting Started
Guide describe in great detail what VueScan does.

Many buyers find the user's guide so lacking that they have
to figure out the user interface by trial and error.

Read the Getting Started Guide.

Regards,
Ed Hamrick

 

[Wilfred]

As stated before, VueScan is a product with no specs, and therefore no
warranty. There is not even a product description on what VueScan can do
at the site. Many buyers find the user's guide so lacking that they have
to figure out the user interface by trial and error. VueScan's attitude
is to just buy it and see if you can figure out how to use it and what
it can do for you. If you can't, you must be an idiot. Even for a long
time supporter such as yourself, don't expect to get any objective
support and response. Never mind trying to find out how it works
internally.

Who said VueScan is a product? In my perception it's a project in which
Ed Hamrick does most of the work. He gets paid for it because other
participants in the project use his work. I can understand very well
that this project-like nature of VueScan makes it not-so-suitable for
those who make money from scanning, such as Hecate and nikita. They are
probably better off with a product such as Silverfast, which is more
expensive and cannot be transferred to another scanner, but which comes
with warranty (at least I guess it does ...). But for me the project
concept is OK. It's much like open-source projects like Mozilla,
especially when you're not a programmer anyway, like me: I don't care
that VueScan isn't open source and I use it while it's evolving, just
like Mozilla.
So I would still appreciate it if Ed would do something with the input I
provided, give some feedback, and eventually makes VueScan work well
with the DSE 5400.

 

[Robert Feinman]

Yes, there's some kind of problem with the infrared channel

on the Minolta 5400. Lots of people complain about it, but
nobody appears to be interested enough to help me diagnose and
solve the problem. I'm probably just going to delete 5400
support entirely.

Personally I find Vuescan the only practical program with the 5400.
Using the "real" ICE is so slow that it makes the whole process
almost unusable.
I'm not sure what the IR problem that most people are referring to
is exactly. Is there an instance online that we can see as an
example?

 

[Ed Hamrick]

Who said VueScan is a product? In my perception it's a project in which
Ed Hamrick does most of the work. He gets paid for it because other
participants in the project use his work.

It's definitely a product - I just don't get the hair-splitting. It's
as silly as the argument over converting from 16-bits to 8-bits (which
by the way missed the important part, which is that converting from
16-bit linear to 8-bits is best done with gamma compression, with
a linear segment at the dark end).

I can understand very well
that this project-like nature of VueScan makes it not-so-suitable for
those who make money from scanning, such as Hecate and nikita.

Some of my best customers are people who make money from scanning,
including professional photographers and scanning services. VueScan
actually has a lot more paying customers than Silverfast.

They are
probably better off with a product such as Silverfast, which is more
expensive and cannot be transferred to another scanner, but which comes
with warranty (at least I guess it does ...).

No, Silverfast doesn't have a warranty either. From the SF 6 Ai manual:

"SilverFast® is provided "as is". LaserSoft Imaging AG does not warrant,
neither expressed nor implied the usefulness of the software SilverFast® for
a
particular purpose or its merchantability or the fitness for licensee's
requirements. Although every effort has been made to eliminate errors,
LaserSoft Imaging AG does not warrant that SilverFast® is free of errors."

So I would still appreciate it if Ed would do something with the input I
provided, give some feedback, and eventually makes VueScan work well
with the DSE 5400.

I'm happy to work on this, but I need:

1) Low-resolution raw scan file that can be used to reproduce the problem
2) vuescan.se5 (calibration file)
3) vuescan.log from scan that produced the raw scan file

Every time I ask someone for the raw scan file, they've sent me a processed
crop file (with no infrared data in it), so there's nothing I can do. I
get a huge amount of e-mail, so I just can't waste a lot of time on
incomplete problem reports.

Regards,
Ed Hamrick