VueScan 8.1.1 Minolta Scan Dual IV problems

[Trustin Lee]

Hi,

I think what I'm mentioning are already known problems, but I just
wanted to know when they will be fixed.

1. Multisampling does not work; vivid stripes!

2. Underexposed slide shows vertical (or horizontal, not sure) noise
pattern (like lines)

If the problem is not fixed yet and more helps are appreciated, what
can I provide to Mr. Hamrick? (Debug logs, Sample images, ...)

I only use Linux and so I cannot use DiMAGE Scan Utility; VueScan is
the only alternative for me. I also wish I could know if there are
any alternatives for now.

Regards,
Trustin Lee

 

[Bart van der Wolf]

SNIP

If the problem is not fixed yet and more helps are appreciated, what
can I provide to Mr. Hamrick? (Debug logs, Sample images, ...)

Just make sure you report it to Ed Hamrick via his Support web page,
he doesn't bite. If he gets multiple reports, he will be better able
to prioritize.

For me, it seems clear that there is something in the Minolta
instruction set that isn't obvious enough for him to control/use. In
fact Minolta issued some upgrades for the DSE-5400 I have, to reduce
striping artifacts. I hope Ed has spoken with the producers of the
Minolta hardware during his visit to Germany, so we can expect a
solution.

Just follow the web page instructions
(http://www.hamrick.com/sup.html) as to what info you need to supply,
or simply ask if he needs more log-files to get to the bottom of this.

Bart

 

[Alex Stols]

Hi,

I think what I'm mentioning are already known problems, but I just
wanted to know when they will be fixed.

1. Multisampling does not work; vivid stripes!

2. Underexposed slide shows vertical (or horizontal, not sure) noise
pattern (like lines)

If the problem is not fixed yet and more helps are appreciated, what
can I provide to Mr. Hamrick? (Debug logs, Sample images, ...)

I only use Linux and so I cannot use DiMAGE Scan Utility; VueScan is
the only alternative for me. I also wish I could know if there are
any alternatives for now.

Regards,

Hello Trustin Lee,

My experiences so far with 8.1.1 and SD4 are:

1. Yes, multisampling is wrong again - it was fine in 8.0.14, but I
haven't used it yet.

2. I can't see these, though I am not sure the faint colored bands
that used to be present have gone altogether. That was a sever
problem in 8.0.14.

3. In my hands autofocus doesn't work, a pity since it would speed
things up quite a bit.

I have given up VS since version 8.0.14, but I tries 8.1.1 and found
that the 2.) problem is nearly gone, so one thing is better, the
other is worse again.
I still would love to use VS as it seems to give you the opportunity
to do things just as you want, but the image quality must be really
good, or else it is useless.

I also have sent Ed Hamrick several "VueScan Problem Reports", some
of which drew response, but most didn't. For one Ed said the
autofocus worked at his end, while it clearly didn't at my end...

Regards,
Alex

 

[Don]

1. Yes, multisampling is wrong again - it was fine in 8.0.14 ....
2. I can't see these, though I am not sure the faint colored bands
that used to be present have gone altogether ....
I have given up VS since version 8.0.14, but I tries 8.1.1 and found
that the 2.) problem is nearly gone, so one thing is better, the
other is worse again.
I still would love to use VS as it seems to give you the opportunity
to do things just as you want, but the image quality must be really
good, or else it is useless.

I also have sent Ed Hamrick several "VueScan Problem Reports", some
of which drew response, but most didn't. For one Ed said the
autofocus worked at his end, while it clearly didn't at my end...

Re-appearing bugs seem to be the only reliable VueScan "feature". ;o)

Judging by the never-ending flood of such messages VueScan appears
riddled with these bugs.

I suppose the only two options are either to find a version of VueScan
with most acceptable combination of bugs (although that may not be
possible as you indicate) or throw VueScan away and use something
reliable.

I have a Nikon and use NikonScan but, just out of curiosity, which
aspects of Minolta software are not performing they way you like? As a
Minolta outsider, the only thing I'm vaguely aware of is something
about not being able to use the Grain Dissolver without ICE, is that
correct?

Don.

 

[Alex Stols]

Re-appearing bugs seem to be the only reliable VueScan "feature". ;o)

Judging by the never-ending flood of such messages VueScan appears
riddled with these bugs.

I suppose the only two options are either to find a version of
VueScan
with most acceptable combination of bugs (although that may not be
possible as you indicate) or throw VueScan away and use something
reliable.

I have a Nikon and use NikonScan but, just out of curiosity, which
aspects of Minolta software are not performing they way you like? As
a
Minolta outsider, the only thing I'm vaguely aware of is something
about not being able to use the Grain Dissolver without ICE, is that
correct?

Hello Don,
I have followed your comments on VS for some time here, and I know
you don't like it a bit. I wouldn't have bought VS if I had known how
it really is for my scanner...

It has, for me, aspects I like a lot. I have been playing around with
it for some time but I gave up because of insufficient color quality.
Now in 8.1.1 the color is *almost* right inmy opinion, but new bugs
have appeared. The focus has in no version behaved on my machine, and
now the cropping is and the processing of RAWs also very difficult to
get right. Yesterday I liked VS, today I hate the program
Still - I will watch the development and see what happens, try every
new version and see if I like it.

As far as the Konica-Minolta DiMAGE Scan Dual IV is concerned: there
is no IR, so no ICE, and I haven't used the Grain dissolver at all. I
only just started scanning really, and my color negatives are so
badly scratched and damaged that I fear even ICE and Grain D. will
not be able to make much of it. For cleaning I use some PS plugins,
noise reduction by NeatImage, and so on, on a modest scale.

Regards,
Alex

 

[Fernando]

Just make sure you report it to Ed Hamrick via his Support web page,
he doesn't bite. If he gets multiple reports, he will be better able
to prioritize.

For me, it seems clear that there is something in the Minolta
instruction set that isn't obvious enough for him to control/use. In
fact Minolta issued some upgrades for the DSE-5400 I have, to reduce
striping artifacts. I hope Ed has spoken with the producers of the
Minolta hardware during his visit to Germany, so we can expect a
solution.

Just for the records: still streaks and low DMax with DSE5400 and
Vuescan 8.1.1 :-(
And a thing I didn't notice before: deep shadows (well, realatively
deep, given the bad black point ;-) ) have a definite green cast...

Fernando

 

[Don]

Hello Don,
I have followed your comments on VS for some time here, and I know
you don't like it a bit. I wouldn't have bought VS if I had known how
it really is for my scanner...

Hoi Alex!

That's not a typo, but me practicing my Dutch. ;-)

No, I don't like VueScan for the reasons I outlined. The point of
mentioning those reasons is to balance things out and help people so
they don't buy wrong software. But, like I always say, there are
others who are perfectly happy with it, however, they seem to use it
superficially.

It has, for me, aspects I like a lot. I have been playing around with
it for some time but I gave up because of insufficient color quality.
Now in 8.1.1 the color is *almost* right inmy opinion, but new bugs
have appeared. The focus has in no version behaved on my machine, and
now the cropping is and the processing of RAWs also very difficult to
get right. Yesterday I liked VS, today I hate the program

We've all been there with whatever software we use...

I can't comment on color quality but one thing I've seen the author
repeat frequently is to delete VueScan.ini (i.e. remove all the
conflicting settings) and then simply set the gray point. Of course,
this can be done with any other software and I'm sure you must have
tried it already, but I'm mentioning it just in case.

I'm, sort of, neutral on the NikonScan, but what I do hate a lot is
the Nikon's so-called "support". I don't think other big firms are any
different but it's so frustrating when people supposing to help you
know even less than you do!

Still - I will watch the development and see what happens, try every
new version and see if I like it.

That's right, you paid for it already so there's nothing to lose.

As far as the Konica-Minolta DiMAGE Scan Dual IV is concerned: there
is no IR, so no ICE, and I haven't used the Grain dissolver at all. I
only just started scanning really, and my color negatives are so
badly scratched and damaged that I fear even ICE and Grain D. will
not be able to make much of it. For cleaning I use some PS plugins,
noise reduction by NeatImage, and so on, on a modest scale.

Sorry, my mistake. Wrong scanner! I must have been thinking about the
5400. Like I said I'm not a Minolta user.

I've been at it for about year and a half and still haven't really
started. I ended up buying a second scanner Nikon LS-50 because my old
one Nikon LS-30 just wasn't up to the job. My biggest problem is
dynamic range because I want to do the slides first.

As for ICE, I'm really amazed at how well it does. Especially the
version (ICE 4) in the LS-50. If you can, I would definitely recommend
trying a few negatives on a scanner with ICE. Even if it doesn't
remove every single scratch it makes the job so much easier. It also
seems to reduce grain a little but without the bad side effects i.e.
that "plastic" look which most of grain reduction software seems to
do.

If you can, the best would be to try ICE on both Nikon and Minolta
because of different light sources and technology (e.g. IR channel
alignment).

Don.

 

[Alex Stols]

Hoi Alex!

That's not a typo, but me practicing my Dutch. ;-)

No, I don't like VueScan for the reasons I outlined. The point of
mentioning those reasons is to balance things out and help people so
they don't buy wrong software. But, like I always say, there are
others who are perfectly happy with it, however, they seem to use it
superficially.


We've all been there with whatever software we use...

I can't comment on color quality but one thing I've seen the author
repeat frequently is to delete VueScan.ini (i.e. remove all the
conflicting settings) and then simply set the gray point. Of course,
this can be done with any other software and I'm sure you must have
tried it already, but I'm mentioning it just in case.

I'm, sort of, neutral on the NikonScan, but what I do hate a lot is
the Nikon's so-called "support". I don't think other big firms are
any
different but it's so frustrating when people supposing to help you
know even less than you do!


That's right, you paid for it already so there's nothing to lose.


Sorry, my mistake. Wrong scanner! I must have been thinking about the
5400. Like I said I'm not a Minolta user.

I've been at it for about year and a half and still haven't really
started. I ended up buying a second scanner Nikon LS-50 because my
old
one Nikon LS-30 just wasn't up to the job. My biggest problem is
dynamic range because I want to do the slides first.

As for ICE, I'm really amazed at how well it does. Especially the
version (ICE 4) in the LS-50. If you can, I would definitely
recommend
trying a few negatives on a scanner with ICE. Even if it doesn't
remove every single scratch it makes the job so much easier. It also
seems to reduce grain a little but without the bad side effects i.e.
that "plastic" look which most of grain reduction software seems to
do.

If you can, the best would be to try ICE on both Nikon and Minolta
because of different light sources and technology (e.g. IR channel
alignment).

Don.

Thanks for your advice, Don, and I can tell you that I *did* use ICE
a short time ago by proxy: I gave a few of my bad films to a
professional, who has a multi-thousand $ Fuji Frontier machine. The
results were amazing indeed. Only the really bad damaged spots could
be found not to be perfect, but then: if part of the image is gone,
who can restore it? But the overall effect was astonishingly good. He
made a 16x24" print of one bad negative and it is wonderful. Can one
do that with a Nikon LS 50 or a Minolta 5400? This man is prepared to
do my bad strips, but he produces 8 bit files, not 16, so I have to
be content with his color settings and the like. I hesitate... I can
have no influence that wat on the results.

The great question is - shall I get rid of my Minolta SD4 and get me
a scanner with ICE??

Regards,
Alex

 

[Bart van der Wolf]

SNIP

But, like I always say, there are others who are perfectly happy
with it, however, they seem to use it superficially.

Demagogy! How do you know how users of 300+ different types of
scanners use it on 3 different platforms? You would be insulting to
non-superficial users if they didn't see right through your pathetic,
unproductive, little games.

Bringing a little balance in opinions is okay (although it doesn't
help the OP with his specific issue), but since you don't use the
products all you do is vent an unfounded opinion, no substance.

All you appear to base your biased opinion on, is a small percentage
of tens of thousands of users that stumble onto something that
appears to not work (as they expect) in their particular workflow.
Often those issues can be traced to hardware or user (RTFM) error, and
in other occasions multiple users report the same issue which then
becomes a support issue for Ed Hamrick. Most the latter issues will be
solvable within one release or so, others are harder, due to lack of
documentation from the manufacturer.

SNIP

I can't comment on color quality but one thing I've seen the
author repeat frequently is to delete VueScan.ini (i.e. remove
all the conflicting settings) and then simply set the gray point.

Yes, bringing the application into a known state, makes trouble
shooting much easier. So that seems sensible advice. And indeed, it
usually takes just a right-mouse-button click on a neutral/gray to get
the color balance right, it's that simple, with VueScan anyway. Again,
perfectly reasonable advice.

Of course, this can be done with any other software and I'm
sure you must have tried it already, but I'm mentioning it just
in case.

Careful, other software (e.g. Photoshop's gray-point sampler in
Levels) may just change the color of the point clicked by adjusting
the gamma, which is the wrong way of color balancing. Color balancing
requires setting the correct blackpoint and whitepoint per channel
first. Be cautious when comparing postprocessing software, because
where one image will be (auto)-corrected to a pleasing balance,
another image may fail completely using the same procedure (that is an
indication of a wrong method).

SNIP

As for ICE, I'm really amazed at how well it does. Especially
the version (ICE 4) in the LS-50. If you can, I would
definitely recommend trying a few negatives on a scanner with
ICE.

Yes, I agree ICE is a life saver in those extremely scratched film
cases, as is a scanner with a more diffuse lightsource or
wet-mounting, but that won't help the OP unless he uses another
scanner.

Bart

 

[Bart van der Wolf]

SNIP

And a thing I didn't notice before: deep shadows (well, realatively
deep, given the bad black point ;-) ) have a definite green cast...

Was that with slides? Slides have poor color accuray in their highest
densities, another reason to avoid underexposure as much as
overexposure. So the blackpoint issue with the DSE-5400 may be
unrelated, or not.

Bart

 

[Fernando]

Was that with slides?

Not only: it was also with the blade of the slanted edge test, that
should be "quite" opaque.
I know about the poor color accuracy of the slides (I have all the
datasheets of the slides I use), but this is not the case.

Fernando

 

[Paul Simons]

No, I don't like VueScan for the reasons I outlined. The point of
mentioning those reasons is to balance things out and help people so
they don't buy wrong software. But, like I always say, there are
others who are perfectly happy with it, however, they seem to use it
superficially.

Don,

I really dislike the way you are behaving on this news group.

Your attitude towards Vuescan and towards Ed personally as well, seamingly,
is clear. Stop repeating yourself. Thanks to Google, your opinion is there
to see for everyone who wants to.

About helping people: you are not helping people by trying to persuade them
not to buy or even look at Vuescan. It is obvious you are trying to do just
that. Instead, advising them to download the free evaluation copy, to see
how it performs in their particular setting, would be a much greater help
(f.e. Alex should have done this).

Your remarks concerning other users of scan-software are pathetic.
Apparently you have way to much self-esteem, blaming the software and
changing hardware, where others are able to establish great results with
similar tools.

Vuescan is discussed on this newsgroup a lot. That is not a bad thing, it's
because a lot of Vuescan-users post here, and bugs are mentioned in the
group because we know reporting them will cause them to be solved, sooner or
later.

Paul

 

[Alex Stols]

Don,

I really dislike the way you are behaving on this news group.

Your attitude towards Vuescan and towards Ed personally as well,
seamingly,
is clear. Stop repeating yourself. Thanks to Google, your opinion is
there
to see for everyone who wants to.

About helping people: you are not helping people by trying to
persuade them
not to buy or even look at Vuescan. It is obvious you are trying to
do just
that. Instead, advising them to download the free evaluation copy,
to see
how it performs in their particular setting, would be a much greater
help
(f.e. Alex should have done this).

In fact that was the first thing I did! But: being a novice in the
trade I could not determine for myself that what was claimed: Vuescan
supports Minolta Scan Dual IV, wasn't true. I was taken in by the
nice way you have control over the whole process, once you get to
know the layout. That multi-sampling was only producing completely
useless results, that auto-focussing wasn't working, that a plain
color scan was producing faint bands of varying colors at a right
angle to the scan diredtion: how was I to see that in a relatively
short time? OK, I was maybe over-enthousiastic.

Your remarks concerning other users of scan-software are pathetic.
Apparently you have way to much self-esteem, blaming the software and
changing hardware, where others are able to establish great results
with
similar tools.

Vuescan is discussed on this newsgroup a lot. That is not a bad
thing, it's
because a lot of Vuescan-users post here, and bugs are mentioned in
the
group because we know reporting them will cause them to be solved,
sooner or
later.

Let's hope so.

Regards,
Alex

 

[Don]

SNIP

Demagogy!

I'm sorry, but the fact that you're forced to use Minolta software
because of VueScan bugs and shortcomings makes all your protestation
sound very hollow...

If VueScan is so wonderful why are you forced to use Minolta software
to actually perform the scan?

Yes, bringing the application into a known state, makes trouble
shooting much easier. So that seems sensible advice. And indeed, it
usually takes just a right-mouse-button click on a neutral/gray to get
the color balance right, it's that simple, with VueScan anyway. Again,
perfectly reasonable advice.

I appreciate that you do acknowledge it (I really do!) although that
does conflict with other statements like the one above.

Anyway, let's just drop all that and...

Careful, other software (e.g. Photoshop's gray-point sampler in
Levels) may just change the color of the point clicked by adjusting
the gamma, which is the wrong way of color balancing. Color balancing
requires setting the correct blackpoint and whitepoint per channel
first. Be cautious when comparing postprocessing software, because
where one image will be (auto)-corrected to a pleasing balance,
another image may fail completely using the same procedure (that is an
indication of a wrong method).

....focus on good stuff like this!

Yes, I totally agree. Auto is a very blunt instrument. All it does is
eliminate empty bins at both ends *hoping* that such histogram
"alignment" of individual RGB channels will result in a color balanced
image. In many cases this may (appear to?) work, but it's really just
a wild guess.

For my color correction (limited to Kodachromes at this point) I use
curves to set the gray point. Of course I have the added problem of
Nikon's inability to scan Kodachromes correctly, even with the
"Kodachrome" setting on. Not to mention insufficient dynamic range (14
bits) to get those dark Kodachrome shadows to come out - but that's
another story...

Yes, I agree ICE is a life saver in those extremely scratched film
cases, as is a scanner with a more diffuse lightsource or
wet-mounting, but that won't help the OP unless he uses another
scanner.

I haven't tried a scanner with a diffuse light source so I can't
comment. However, I read here that a diffuse source (in addition to
also reducing grain, which is a good thing) unfortunately tends to
"soften" the image somewhat.

On the other hand, sharp LED scans tend to have a narrow depth of
field - painfully demonstrated with my old, cardboard-mounted
Kodachromes which over time got quite warped... :-(

So, it's a case of: six on one, half a dozen of the other. Can't win.
The bottom line seems to be prioritizing and then choosing the lesser
evil... :-/

Don.

 

[Don]

Thanks for your advice, Don, and I can tell you that I *did* use ICE
a short time ago by proxy: I gave a few of my bad films to a
professional, who has a multi-thousand $ Fuji Frontier machine. The
results were amazing indeed. Only the really bad damaged spots could
be found not to be perfect, but then: if part of the image is gone,
who can restore it? But the overall effect was astonishingly good. He
made a 16x24" print of one bad negative and it is wonderful. Can one
do that with a Nikon LS 50 or a Minolta 5400?

I don't print so I can't really tell. I only scan to have my images in
a digital format for archiving and then viewing on a computer.

The general rule is that the higher the resolution the larger the
print without artifacts. You can upsample (increase image size) but
since that simply "invents" pixels there is no new data so it's of
limited use.

But there are others here who are better suited to give advice on
printing and I hope they jump in an provide the information.

This man is prepared to
do my bad strips, but he produces 8 bit files, not 16, so I have to
be content with his color settings and the like. I hesitate... I can
have no influence that wat on the results.

Yes, that's a problem. Can't you ask him to supply 16 bit files? I
mean, in theory, it shouldn't make any difference to him.

I prefer to do the work myself because these images are very personal
to me and I don't think a "stranger" would be as careful and
discriminating as I am, but that's really my subjective opinion. If
you're happy with the results it's certainly an option. Much faster
too!

The great question is - shall I get rid of my Minolta SD4 and get me
a scanner with ICE??

If you can afford it I personally would indeed recommend it. There are
a few very important caveats though!

ICE doesn't work with Kodachromes - perhaps not such a big problem
except for people like me who have a large Kodachrome collection.

But more importantly, ICE doesn't work with B&W film! I seem to
remember reading there are now new B&W films that will work with ICE
but that's fairly new. The trouble is most of B&W film we're scanning
these days is old and ICE doesn't work with them!

Finally, do examine the differences between Nikon and Minolta closely
because they are significant. Most importantly: the light source!

Nikons use LED (sharper image, but less of depth of field, curved film
will be hard to focus on), Minolta uses a conventional light source
(less grain, no problem with Kodachromes, but this means using filters
and that's another thing to go wrong between the image and the scan).

Also, speaking of ICE, Nikons use an infrared LED simultaneously with
RGB LEDs so the alignment is perfect, while Minolta (as far as I
understand) has two light sources (conventional for exposure, and
infrared for ICE) so there may be alignment problems. Minolta 5400
users should know more about this than me so I let them comment...

On the other hand, Minolta does have 5400 resolution while Nikons are
"only" 4000. I read a heated discussion here over a year ago about
resolution and the conclusion appeared to be that to really get every
last bit of data from film you need a ~10,000 resolution.

On the third hand ;o) LEDs never fade or burn out, while a
conventional lamp is bound to fade with time and eventually burn out.

Etc, etc... ;o)

Don.

 

[Don]

Don,

I really dislike the way you are behaving on this news group.

I'm sorry you feel like that, Paul, but all I write are facts. If you
don't like these facts you always have the option of either ignoring
or filtering out my messages, and I urge you to do so.

If responses from VueScan users were more factual then there would be
no problem. Do note, also, that I never use obscenities in this
newsgroup, not even when the "other side" does!

Your attitude towards Vuescan and towards Ed personally as well, seamingly,
is clear. Stop repeating yourself. Thanks to Google, your opinion is there
to see for everyone who wants to.

Which begs the question, why don't you ask VueScan proponents to stop
repeating themselves?

Sure, there is Google, but following that logic we can close this
newsgroup down because *all* "the answers are out there", to
paraphrase X-files. ;-)

However, as long as this newsgroup exists I have just as much right as
anyone else to state the *facts* even if some people may not like
them. Especially, when I *respond* to *misleading* VueScan praises.

I have no attitude or opinion, only facts. Any of my responses you may
not like are just that: factual *responses*.

I have nothing against Ed personally, although he seems to have a lot
of personal animosity towards me. So much so, that he doesn't shy from
obscenities. That speaks more about him, than about me... Especially
since I refuse to stoop down to that level.

About helping people: you are not helping people by trying to persuade them
not to buy or even look at Vuescan. It is obvious you are trying to do just
that.

I'm sorry, but that's where you're totally wrong. Like I always write,
it's up to everyone to decided for themselves after examining *all*
pertinent facts. The key word here is "all", not just the praises!

Instead, advising them to download the free evaluation copy, to see
how it performs in their particular setting, would be a much greater help
(f.e. Alex should have done this).

And he has... And then he paid for it... And now he's sorry he did
because the software doesn't work as advertised, presumably because
key features are disabled in the review copy (e.g. raw scans).

And I'm sure he would have appreciated being told about VueScan's
problems and reoccurring bugs *before* he parted with his money!

Now, why don't you ask the VueScan author to give him his money back!?

Vuescan is discussed on this newsgroup a lot. That is not a bad thing, it's
because a lot of Vuescan-users post here, and bugs are mentioned in the
group because we know reporting them will cause them to be solved, sooner or
later.

I'm afraid that's where you're also wrong as many bugs seem to keep
coming back. They may be solved eventually, "one day...", but why do
you object to simply stating the *fact*, for example, that for over 6
months - and after repeated claims that it's been "fixed" - the
infamous and notorious VueScan stripes are still there when used with
Minolta?

Someone who just joined doesn't know that. Don't you think they
deserve to be told before wasting their hard earned cash?

However, existing - and very frustrated! - VueScan users with Minoltas
don't like being reminded of it so they misdirect their rage by
"shooting the messenger". That's not going to fix the VueScan bugs!

What they should be furious about is why are these *same* VueScan bugs
still there after all this time?

Don.

 

[Don]

I know about the poor color accuracy of the slides (I have all the
datasheets of the slides I use), but this is not the case.

Just out of curiosity, where did you get the datasheets?

I got a Kodachrome sheet link from Kennedy once, but any other links
would be appreciated!

Thanks!

Don.

 

[Xiaotian Sun]

Hi,

I think what I'm mentioning are already known problems, but I just
wanted to know when they will be fixed.

1. Multisampling does not work; vivid stripes!

2. Underexposed slide shows vertical (or horizontal, not sure) noise
pattern (like lines)

If the problem is not fixed yet and more helps are appreciated, what
can I provide to Mr. Hamrick? (Debug logs, Sample images, ...)

I only use Linux and so I cannot use DiMAGE Scan Utility; VueScan is
the only alternative for me. I also wish I could know if there are
any alternatives for now.

Hi Trustin:

I would like to know how you set up for SD IV under linux. I primaryly
use linux, but I always switch to windows for my scanning.

I haven't tried 8.1.1 yet. As Alex said, Ed hasn't been mentioning SD4
for a while; I missed 8.0.20 too. I'll give it a try tonight when I go
home.

Xiaotian

 

[Bart van der Wolf]

SNIP

Yes, I totally agree. Auto is a very blunt instrument. All it does
is eliminate empty bins at both ends *hoping* that such
histogram "alignment" of individual RGB channels will result in
a color balanced image. In many cases this may (appear to?)
work, but it's really just a wild guess.

Several settings, like for Auto Contrast, in Photoshop can be set to a
user preference. The location where that is done depends a bit on the
version, but with Version 8 (CS) it can be found e.g. in the
Levels/Curves dialog under Options. There the Clipping points can be
set, also to 0%, and they will apply to the Auto functions as well.

For my color correction (limited to Kodachromes at this point)
I use curves to set the gray point. Of course I have the added
problem of Nikon's inability to scan Kodachromes correctly,
even with the "Kodachrome" setting on. Not to mention
insufficient dynamic range (14 bits) to get those dark
Kodachrome shadows to come out - but that's another story...

14-bits can span a linear density range of 4.21. Should be enough for
any slide film. Even if we assume 1.5-bits of noise, there is enough
room for a D=3.76. If the slide is underexposed, there is significant
scene luminance compression in the shadows. That may be hard to
recover, unless extreme tonescaling is applied to that part of the
response curve. Just for that, the additional bits may allow a less
posterized result (but still with very poor color accuracy).

SNIP

I haven't tried a scanner with a diffuse light source so I can't
comment. However, I read here that a diffuse source (in
addition to also reducing grain, which is a good thing)
unfortunately tends to "soften" the image somewhat.

http://www.xs4all.nl/~bvdwolf/main/foto/scan/se5400/se5400-5.htm shows
an example of how it can look (on color negative film). I've also done
MTF measurements, but they show no significant reduction of
resolution, it is usually marginal and the choice is really a
no-brainer since the reduction in graininess also allows more
sharpening. But I do understand that only few scanners offer it as an
option.

On the other hand, sharp LED scans tend to have a narrow
depth of field - painfully demonstrated with my old, cardboard-
mounted Kodachromes which over time got quite warped... :-(

I think the output of LEDs has improved enough over the years to allow
an other compromise; smaller aperture versus faster scan. I'm not sure
what Nikon did in the latest models.

So, it's a case of: six on one, half a dozen of the other. Can't win.
The bottom line seems to be prioritizing and then choosing the
lesser evil... :-/

Such is life...

Bart

 

[Bart van der Wolf]

SNIP

Also, speaking of ICE, Nikons use an infrared LED simultaneously
with RGB LEDs so the alignment is perfect, while Minolta (as far
as I understand) has two light sources (conventional for exposure,
and infrared for ICE) so there may be alignment problems. Minolta
5400 users should know more about this than me so I let them
comment...

AFAIK, for sure it applies to the 5400 model, there are two exposures.
One is RGB and one is RGBI, both are aligned in a single stepper motor
position so there is no positioning/angleing issue. In fact it is, by
Minolta scan utility default, only possible to apply ICE together with
the diffusor plate in position, so RGBI is well mixed as it reaches
the film.

This differs from the Nikons which can alternate between R, G, B and
I.

On the other hand, Minolta does have 5400 resolution while
Nikons are "only" 4000. I read a heated discussion here over a
year ago about resolution and the conclusion appeared to be that
to really get every last bit of data from film you need a ~10,000
resolution.

That does assume that the image on film has enough resolution to begin
with. That will only be the case with good lenses and a steady hand
(tripod is better) on low ISO film. But even then, the higher scan
resolution helps in reducing grain aliasing.

Bart