Vuescan-Minolta 5400-Noise

[thomas]

Since september 2003 I have been struggling with using Vuescan with my
Minolta 5400 scanner. I like the Vuescan workflow very much and
Vuescan makes it possible to calibrate the scanner witl IT8.7 target
(I use Provia and Velvia).

A couple of months ago I thought Ed had solved the problems with
Vuescan for 5400 and in the last months I scanned a lot of pictures
but thought some of the pictures were too noisy especially in dark
areas.

Now I bought Neat Image and found out, when checking the profile
viewer, that the figures for Cr and Cb were twice as high for scans
from Vuescan compared withthe same scan from the original software
from Minolta.
This is not acceptable!

I use same settings for resolution, grain dissolver and no colour
management as well as othe default values for Vuescan. Only difference
is that I with Vuescan use light grain reduction.

Why is Ed not correcting this?

Regards
Thomas

 

[Christian Tsotras]

I scanned a lot of pictures but thought some of the pictures were too
noisy especially in dark areas. [...]
Now I bought Neat Image and found out, when checking the profile viewer,
that the figures for Cr and Cb were twice as high for scans from Vuescan
compared withthe same scan from the original software from Minolta. This
is not acceptable!

I had the same problems of noisy dark areas with my Minolta Elite scanner.
The Minolta software removed much more efficiently the noise in the border
of the CCD than Vuescan.

Vuescan seems to have some problems with calibrating Minolta Elite (first
model) scanners. Maybe you have the same problem.



I found a workaround with my scanner, hope it will work for you:

When switching my (old) scanner on, it makes like a self-test: looks like
a self-calibration. This operation is also launchable with Minolta's
software with Shift+Ctrl+I.

But this calibration, when done on a "cold" scanner (switched off since
many hours) is not very useful because the scanner will slowly warm up:
30 minutes later, the CCD noise it produces has evolved and the hardware
calibration process will only correct the noise produced by the "cold"
CCD.

So you have to force a new CCD reset when the CCD is "warm": either by
Minolta's software, or by an off/on of the scanner. Beware of not doing
this on a SCSI unit: you must shut down the SCSI card of the PC too !!

My method is to switch on the scanner several hours before switching on my
computer. And just before switching my PC on, I do a off/on of the scanner
and then, switch on the computer.


Some variations include "calibrating" Vuescan with a cold CCD (and
keeping the vuescan.dse file for all future scans), or a warm CCD.

For me, the best results were obtained with a vuescan.dse file created by
a "cold" calibration of Vuescan, then keeping it all the time while making
"warm" hardware recalibrations.

I think you will have to do tests, I can't tell what happens with an Elite
5400. I would be interested on the results of your tests.

Why is Ed not correcting this?

Don't know. I posted some questions some months ago on this newsgroup, but
maybe Ed has more to do with supporting new scanners that perfecting the
calibration process of my "old" scanner.


If you're *very* interested, I could put online some of my tests
showing that Minolta software is better at removing CCD noise that Vuescan
***with my Minolta Elite scanner***. Can't tell if is true with other
Minoltas or any other brand.

 

[Ricky]

I am also a DSE5400 owner and I know exactly what you mean! I have *A
LOT* of scanning to do, but I can't make a start on it because I can't
use the Minolta software (the scanner just will not focus properly -
every scan is blurred), I am not keen on shelling out for SilverFast
(anyone know how it performs with this scanner BTW?), and scans from
Vuescan are unacceptable. I wish Ed Hamrick would fix the problem so I
can get on with my life...

I've stood by Vuescan for a while now, using it with my old FS4000
(even though it didn't exactly perform brilliantly with that either),
but enough is enough. A lot of people have paid their money to use
Vuescan with the DSE5400 and they deserve better!

Ricky.

 

[degrub]

Download a full demo of Silverfast and try it out. Free.

http://www.silverfast.com/sf-download/demo/en.html

i use Vuescan and Silverfast.
Frank

 

[thomas]

I scanned a lot of pictures but thought some of the pictures were too
noisy especially in dark areas. [...]
Now I bought Neat Image and found out, when checking the profile viewer,
that the figures for Cr and Cb were twice as high for scans from Vuescan
compared withthe same scan from the original software from Minolta. This
is not acceptable!

I had the same problems of noisy dark areas with my Minolta Elite scanner.
The Minolta software removed much more efficiently the noise in the border
of the CCD than Vuescan.

Maybe because Minolta software is clipping in dark areas?

Vuescan seems to have some problems with calibrating Minolta Elite (first
model) scanners. Maybe you have the same problem.



I found a workaround with my scanner, hope it will work for you:

When switching my (old) scanner on, it makes like a self-test: looks like
a self-calibration. This operation is also launchable with Minolta's
software with Shift+Ctrl+I.

Calibration starts for me when I start the program and not the
sacanner.

But this calibration, when done on a "cold" scanner (switched off since
many hours) is not very useful because the scanner will slowly warm up:
30 minutes later, the CCD noise it produces has evolved and the hardware
calibration process will only correct the noise produced by the "cold"
CCD.

I have missed this. Followed your advice ant it seems to help a bit.

So you have to force a new CCD reset when the CCD is "warm": either by
Minolta's software, or by an off/on of the scanner. Beware of not doing
this on a SCSI unit: you must shut down the SCSI card of the PC too !!

Why not calibrate in the Vuescan meny under scanner? Why on/off?

My method is to switch on the scanner several hours before switching on my
computer. And just before switching my PC on, I do a off/on of the scanner
and then, switch on the computer.

Why on/off of scanner?

Some variations include "calibrating" Vuescan with a cold CCD (and
keeping the vuescan.dse file for all future scans), or a warm CCD.

For me, the best results were obtained with a vuescan.dse file created by
a "cold" calibration of Vuescan, then keeping it all the time while making
"warm" hardware recalibrations.

Do not understand. What do you mean with a hardware calibration? If I
calibrate Vuescan in the Vuescan meny when the scanner is warm will
this be a hardware calibration?

I think you will have to do tests, I can't tell what happens with an Elite
5400. I would be interested on the results of your tests.

Yes it helped to calibrate on a warm scanner.

Don't know. I posted some questions some months ago on this newsgroup, but
maybe Ed has more to do with supporting new scanners that perfecting the
calibration process of my "old" scanner.


If you're *very* interested, I could put online some of my tests
showing that Minolta software is better at removing CCD noise that Vuescan
***with my Minolta Elite scanner***. Can't tell if is true with other
Minoltas or any other brand.

Thanks for your help. I appreciate it but things still look a bit
weired with Vuescan. Further after investing in Vuescan I do not want
to also invest in Silverfast.
Regards
Thomas

 

[David C Miers]

Speaking from experience with the Scan Dual III Minolta software only vs.
Vuescan it would appear at first that Vuescan does produce more noise then
the Minolta software. However I doubt this is really so. What I suspect
you are experiencing is the increased shadow detail that the default Vuescan
settings will produce as the default setting for the black point is "0". If
you change the black and white points to duplicate an image that would be
produced by the Minolta software I highly suspect that you would find less
noise. This ability to bring out the shadow detail is one of Vuescans
strong points that I've not been able to duplicate in the Minolta software
no matter how much I go into the settings. I've heard people complain that
Vuescan has no variable exposure settings as the Minolta software has, but
even pumping up the exposure in the Minolta software does not seem to equal
this shadow detail that Vuescan has been able to achieve. I feel also that
Vuescan is designed to be a professional use software that assumes that you
will pass the program on to editing software such as Adobe PS to get an
image that will be satisfactory. Minolta tries a bit more to generate an
image that is more usuable right out of the scanner software. Not
satisfactory IMHO, but more usable.

 

[David C Miers]

I forgot to mention that at least Vuescan does have some included grain
removal software. However if you have other grain removal software such as
Neat Image or Grain Surgery, you would probably be better off not using the
Vuescan grain reduction option as the dedicated versions generally do a
better job and retain more detail. Hint: Use history brush to bring back
detail in specific areas after using grain removal software plugins in
Photoshop.

 

[Christian Tsotras]

Maybe because Minolta software is clipping in dark areas?

I don't think so.

Look at my tests here:
http://christian.tsotras.free.fr/scanner_reset/

I made this page months ago to show some problems of noise and defect
removal with Vuescan compared to Minolta software. The first pictures show
the border noise problem.

Why on/off of scanner?

To force a built-in reset.

I work on Linux, and I prefer doing an electrical reset than launching
Minolta software with Wine (a windows emulator on Linux).

Do not understand. What do you mean with a hardware calibration?

What I described above: warming the scanner some time, then off/on and
working.

If I calibrate Vuescan in the Vuescan meny when the scanner is warm will
this be a hardware calibration?

I don't think so. Vuescan seems to make a "quick" calibration compared
to Minolta's reset (on my Elite scanner). I call this "hardware
calibration", but scanner specialists may call it "hardware reset" or
something else.

 

[Mac McDougald]

I don't think so.

Look at my tests here:
http://christian.tsotras.free.fr/scanner_reset/

I made this page months ago to show some problems of noise and defect
removal with Vuescan compared to Minolta software. The first pictures show
the border noise problem.


To force a built-in reset.

I work on Linux, and I prefer doing an electrical reset than launching
Minolta software with Wine (a windows emulator on Linux).


What I described above: warming the scanner some time, then off/on and
working.


I don't think so. Vuescan seems to make a "quick" calibration compared
to Minolta's reset (on my Elite scanner). I call this "hardware
calibration", but scanner specialists may call it "hardware reset" or
something else.

Just a comment on the first example:
Why on earth are you including the slide mount in your scans to begin
with? Of course VS will try to "get all the info" it can from the opaque
area also, at default settings anyway.

Mac

 

[degrub]

">>But this calibration, when done on a "cold" scanner (switched off since"

This is probably the lamp warming up and stabilizing. Ultimately, you
will have to replace the lamp when it gets too dim. Warming up CCDs only
increases their "noise". Actually, you want CCDs as cold as possible to
minimize noise.

Frank

I scanned a lot of pictures but thought some of the pictures were too
noisy especially in dark areas.
[...]

Now I bought Neat Image and found out, when checking the profile viewer,
that the figures for Cr and Cb were twice as high for scans from Vuescan
compared withthe same scan from the original software from Minolta. This
is not acceptable!

I had the same problems of noisy dark areas with my Minolta Elite scanner.
The Minolta software removed much more efficiently the noise in the border
of the CCD than Vuescan.

Maybe because Minolta software is clipping in dark areas?

Vuescan seems to have some problems with calibrating Minolta Elite (first
model) scanners. Maybe you have the same problem.



I found a workaround with my scanner, hope it will work for you:

When switching my (old) scanner on, it makes like a self-test: looks like
a self-calibration. This operation is also launchable with Minolta's
software with Shift+Ctrl+I.

Calibration starts for me when I start the program and not the
sacanner.

But this calibration, when done on a "cold" scanner (switched off since
many hours) is not very useful because the scanner will slowly warm up:
30 minutes later, the CCD noise it produces has evolved and the hardware
calibration process will only correct the noise produced by the "cold"
CCD.

I have missed this. Followed your advice ant it seems to help a bit.

So you have to force a new CCD reset when the CCD is "warm": either by
Minolta's software, or by an off/on of the scanner. Beware of not doing
this on a SCSI unit: you must shut down the SCSI card of the PC too !!

Why not calibrate in the Vuescan meny under scanner? Why on/off?

My method is to switch on the scanner several hours before switching on my
computer. And just before switching my PC on, I do a off/on of the scanner
and then, switch on the computer.

Why on/off of scanner?

Some variations include "calibrating" Vuescan with a cold CCD (and
keeping the vuescan.dse file for all future scans), or a warm CCD.

For me, the best results were obtained with a vuescan.dse file created by
a "cold" calibration of Vuescan, then keeping it all the time while making
"warm" hardware recalibrations.

Do not understand. What do you mean with a hardware calibration? If I
calibrate Vuescan in the Vuescan meny when the scanner is warm will
this be a hardware calibration?

I think you will have to do tests, I can't tell what happens with an Elite
5400. I would be interested on the results of your tests.

Yes it helped to calibrate on a warm scanner.

Don't know. I posted some questions some months ago on this newsgroup, but
maybe Ed has more to do with supporting new scanners that perfecting the
calibration process of my "old" scanner.


If you're *very* interested, I could put online some of my tests
showing that Minolta software is better at removing CCD noise that Vuescan
***with my Minolta Elite scanner***. Can't tell if is true with other
Minoltas or any other brand.

Thanks for your help. I appreciate it but things still look a bit
weired with Vuescan. Further after investing in Vuescan I do not want
to also invest in Silverfast.
Regards
Thomas

 

[Ken]

Speaking from experience with the Scan Dual III Minolta software only vs.
Vuescan it would appear at first that Vuescan does produce more noise then
the Minolta software. However I doubt this is really so. What I suspect
you are experiencing is the increased shadow detail that the default Vuescan
settings will produce as the default setting for the black point is "0". If
you change the black and white points to duplicate an image that would be
produced by the Minolta software I highly suspect that you would find less
noise. This ability to bring out the shadow detail is one of Vuescans
strong points that I've not been able to duplicate in the Minolta software
no matter how much I go into the settings. I've heard people complain that
Vuescan has no variable exposure settings as the Minolta software has, but
even pumping up the exposure in the Minolta software does not seem to equal
this shadow detail that Vuescan has been able to achieve. I feel also that
Vuescan is designed to be a professional use software that assumes that you
will pass the program on to editing software such as Adobe PS to get an
image that will be satisfactory. Minolta tries a bit more to generate an
image that is more usuable right out of the scanner software. Not
satisfactory IMHO, but more usable.

While there are differences between the way the Minolta software and
Vuescan process images, there is a way to see the effect of
calibration, and its not quite as good with Vuescan as with the Minolta
software.

Use the Minolta software to save a RAW image (16 bit linear) of a slide
that has problems with Vuescan. Then read it in with Vuescan as a RAW.
Problems with lines will be much reduced or non-existent.

One additional point is that if you are scanning well exposed E6 slides
in a cool environment and are prepared to regularly recalibrate the
scanner, then you probably aren't having these problems. If you want to
really experience problems then try scanning Kodachromes in a hot
environment.