Minolta Scan Dual III and vuescan still not working right

  • Thread starter Thread starter David C Miers
  • Start date Start date
D

David C Miers

I'm wondering about all the other ScanDual III owners out there using
vuescan? Are all your problems now resolved with crashing and USB 2.0?
Unfortuanately mine are not. After downloading the latest version it now
locked up the whole computer and it didn't recover at all. I'm running win
2k here. The previous versions have been hit and miss with some working to
an extent, but after a couple of scans problems with either not completing
the prescan or the final scan, and/or locking up the vuescan program
entirely, predominate with most of them. When it locks up either the
program or the computer it also fails to generate a log file to send to
hamrick software. The scanner functions flawlessly with the Minolta
software, thus I highly doubt I have a hardware problem, unless somehow my
hardware is specifically not compatible with vuescan. I've tried using
different USB ports on my system and sometimes that would help for a time,
or with the current version of vuescan at that time. I really like the
results of vuescan when I can get it too work. But this has gotten so
exasperating that I've given up for the most part. I keep checking at
hamrick.com hoping for a new release that addresses these issues, but
evidently as there is a lack of comment on this newgroup regarding these
problems the other users are now content. I guess I'll have to keep trying
to get a log file that I can send to hamrick software. Any help or
suggestions here greatly appreciated.

Frustrated
 
David C Miers said:
I'm wondering about all the other ScanDual III owners out there using
vuescan? Are all your problems now resolved with crashing and USB 2.0?
Unfortuanately mine are not. After downloading the latest version it now
locked up the whole computer and it didn't recover at all. I'm running win
2k here. The previous versions have been hit and miss with some working to
an extent, but after a couple of scans problems with either not completing
the prescan or the final scan, and/or locking up the vuescan program
entirely, predominate with most of them. When it locks up either the
program or the computer it also fails to generate a log file to send to
hamrick software. The scanner functions flawlessly with the Minolta
software, thus I highly doubt I have a hardware problem, unless somehow my
hardware is specifically not compatible with vuescan. I've tried using
different USB ports on my system and sometimes that would help for a time,
or with the current version of vuescan at that time. I really like the
results of vuescan when I can get it too work. But this has gotten so
exasperating that I've given up for the most part. I keep checking at
hamrick.com hoping for a new release that addresses these issues, but
evidently as there is a lack of comment on this newgroup regarding these
problems the other users are now content. I guess I'll have to keep trying
to get a log file that I can send to hamrick software. Any help or
suggestions here greatly appreciated.

Frustrated
I have had problems with lockup using ds3dual utility and still wait for a
response from minolta.
I also got lockups with vuescan but after increasing scan memory allocation
to 170mb it runs ok.
 
"David C Miers" asked: "I'm wondering about all the other ScanDual III
owners out there using vuescan? Are all your problems now resolved with
crashing and USB 2.0?"
--------------------------

I've been using a Scan Dual III for 6 weeks or so, since my Scan Dual II
went on strike. I've been using Vuescan exclusively (WinXP Home, 384Mb
mem, 16Gb free on 80Gb disk, USB2.0). I've had a couple of instances of
Vuescan going off to computer-heaven while scanning (Vuescan quits
responding, must be force-quit (Ctl-Alt-Del) and restarted). This has
happened in the middle of scanning/saving, with both slides and
negatives. I'm usually doing something else on the computer while
scanning, and I've attributed this to Vuescan not playing well with some
other operation I'm doing. I haven't been able to see a pattern to the
problem to identify a culprit program. It doesn't create a crash log in
Inspector||X||.

At first, I had to restart WinXP to get Vuescan to load, but I
discovered if I just turned the scanner off and back on before
restarting Vuescan, everything worked fine. This has happened maybe half
a dozen times in 20-30 scanning sessions, so it has been a minor
annoyance rather than a major problem. It has happened with version
7.6.75, which I'm currently using, and version 7.6.70, which I
previously used.

I also have a problem of not being able to do preview scans of slides at
353ppi resolution. When I try that resolution, the scanner makes a loud
hum, quickly scans something, and delivers a blank (black) image. It
works fine for negatives at 353ppi. My solution has been not to use
353ppi for slide previews; I use 705ppi and everything is fine.

Preston Earle
(e-mail address removed)
 
My symptoms match those of Preston Earle and I discovered the same
(very disruptive) recovery techniques he documented.

My setup is VueScan 7.6.72, WinXP Pro, Pentium 2.4Ghz, 768 MB Ram,
working in a 28 GB partition with 8+ GB free, USB 2.0.

I recently scanned over 400 slides and it hung about every 15-40
scans.




"David C Miers" asked: "I'm wondering about all the other ScanDual III
owners out there using vuescan? Are all your problems now resolved with
crashing and USB 2.0?"
--------------------------

I've been using a Scan Dual III for 6 weeks or so, since my Scan Dual II
went on strike. I've been using Vuescan exclusively (WinXP Home, 384Mb
mem, 16Gb free on 80Gb disk, USB2.0). I've had a couple of instances of
Vuescan going off to computer-heaven while scanning (Vuescan quits
responding, must be force-quit (Ctl-Alt-Del) and restarted). This has
happened in the middle of scanning/saving, with both slides and
negatives. I'm usually doing something else on the computer while
scanning, and I've attributed this to Vuescan not playing well with some
other operation I'm doing. I haven't been able to see a pattern to the
problem to identify a culprit program. It doesn't create a crash log in
Inspector||X||.

At first, I had to restart WinXP to get Vuescan to load, but I
discovered if I just turned the scanner off and back on before
restarting Vuescan, everything worked fine. This has happened maybe half
a dozen times in 20-30 scanning sessions, so it has been a minor
annoyance rather than a major problem. It has happened with version
7.6.75, which I'm currently using, and version 7.6.70, which I
previously used.

I also have a problem of not being able to do preview scans of slides at
353ppi resolution. When I try that resolution, the scanner makes a loud
hum, quickly scans something, and delivers a blank (black) image. It
works fine for negatives at 353ppi. My solution has been not to use
353ppi for slide previews; I use 705ppi and everything is fine.

Preston Earle
(e-mail address removed)

Email:
User: roy27614
ISP: mindspring.com
 
I have had problems with lockup using ds3dual utility and still wait for a
response from minolta.
I also got lockups with vuescan but after increasing scan memory allocation
to 170mb it runs ok.

Where is the setting for "scan memory allocation"?? Thanks!


Email:
User: roy27614
ISP: mindspring.com
 
David C Miers said:
I'm wondering about all the other ScanDual III owners out there using
vuescan? Are all your problems now resolved with crashing and USB 2.0?

Have occaisonal lockups on Mac OS X 10.3 usually because I've got the
computer doing too much at once.

My main problem is what I expect is a calibration problem. With
Kodachrome I get streaks in the dark areas, which I definitely don't get
with the Minolta software (so its not dust on the CCD).

I have put an example up of a comparison between Vuescan and DS Dual3
Utility scan, both at 2820 dpi. The DS Image scan is actually a 16 bit
linear file from the DS Dual3 utility software scanned into Vuescan.

http://homepage.mac.com/kjbeath/scan comparison.pdf

Vuescan is at 1% exposure clipping, which reduces the problem compared
to 0.01%.
 
Roy Bishop said:
Where is the setting for "scan memory allocation"?? Thanks!


Email:
User: roy27614
ISP: mindspring.com

At the bottom of prefs page, (select all options).
 
Roy Bishop said:
My symptoms match those of Preston Earle and I discovered the same
(very disruptive) recovery techniques he documented.

My setup is VueScan 7.6.72, WinXP Pro, Pentium 2.4Ghz, 768 MB Ram,
working in a 28 GB partition with 8+ GB free, USB 2.0.

I recently scanned over 400 slides and it hung about every 15-40
scans.

I notice nobody has added to my own thread
http://groups.google.com/[email protected]
(sorry - giving the Google link, but I'm accessing and posting with
Google!). But clearly this is not just a problem for Minolta Scan Dual
III. Those are the same symptoms I am getting with Nikon CoolScan V
(LS-50) and VueScan.

I am interested to see that Roy Bishop is getting the problem even on
a much more powerful PC configuration than I own. Mine is VueScan
7.6.75, Win98SE, Pentium 350MHz, 384 MB Ram,
working in an 18GB partition with 6+ GB free, USB 1.1.

Almost everything in Roy Bishop's configuration far exceeds that, yet
he still gets the crashes.

It is clearly a VueScan problem since NikonScan does not crash
(produces horrible color balance and useless output but manages to
finish!) and problems are occurring for Nikon and Minolta users. I
wonder what evidence we need to provide to enable Ed Hamrick to look
at the problem. It must be difficult for him to investigate
intermittent faults without clear evidence of a reproducible pattern.

By the way, in answer to Roy Bishop's earlier post in this thread
"Where is the setting for "scan memory allocation"??" - it is on the
Prefs Tab, near the bottom, called Scan mem (MB). I set mine to 200MB
and still get the crashes. I have yet to try 300MB or even settings
that exceed the RAM of my PC as I don't know what VueScan does with
the chosen setting. Other things to try might be running with no other
USB devices connected to the PC. I am already scanning with absolutely
no other programs running on the PC.

Any other comments would be most helpful.
 
Thanks! I had the all options turned on but since all other tabs
fit on my screen I was too dumb to notice the scroll bar appear when I
clicked on the Prefs tab.

At the bottom of prefs page, (select all options).

Email:
User: roy27614
ISP: mindspring.com
 
I have some good news, however I cannot say exactly for sure what has cured
my problems. Since I installed vuescan 7.6.76 things seem to be working
alright. However several other things have happened as well. I did update
my memory allocation to 100meg of ram for preview and 200 meg of ram for
scans. However I also recently updated audio and 4in1via drivers as well as
some office updates. I don't see how the office updates could be an issue,
but one never knows.

I fired it up with the new version of vuescan and made sure I had the output
log option checked so I had something to send to hamrick software hopefully.
I then tentatively ran a few scans, but to my surprise it kept working
great. I then even pushed the envelope running 2 photoeditors, adobe and
paintshop pro, as well as ACDSEE all at the same time. I pushed Paint Shop
Pro really hard processing an image at the same time that vuescan was
scanning, and all I got out of it was a slight pause when I obviously ran
out of memory, but it then picked right back up a second later and continued
the scan.

Anyways, thanks Leon, Ed, and all the others here that contributed an answer
to my original post.

Dave
 
David C Miers said:
I then tentatively ran a few scans, but to my surprise it kept working
great.

I wonder how many consecutive scans you did without it locking up and
at what resolution?

I have not yet managed more than 8 consecutive scans at 4000 dpi even
with Scan mem set at 350 or 400MB.

I did several tests. Those with no other USB devices connected
involved:

Disconnected USB ADSL modem so no USB devices apart from scanner
present.
Restarted computer.

All tests, with and without modem connected involved:

Restarted and used ctrl+alt+delete to kill all processes apart from
Explorer and Systray.
Switched on scanner.
Started VueScan.

So no other processes were running.

In summary, my tests gave the following results.

No modem, Scan mem=350: ---7 consecutive scans OK.
With modem, Scan mem=350: ---8 consecutive scans OK. 9th scan locks up
at 99%.
With modem, Scan mem=400: ---2 consecutive scans OK. 3rd scan locks up
at 75%.
No modem, Scan mem=400: ---2 consecutive scans OK. 3rd scan locks up
at 27%.
No modem, Scan mem=400:--- 5 consecutive scans OK.

I can post my detailed procedure with all the settings but it would be
a long post!

Previous tests with the modem connected, but Scan mem = 200, did not
ever achieve 2 consecutive scans at 4000 dpi but never locked up at
2000 dpi.

It would be helpful if someone who knows how VueScan uses the Scan mem
setting and how it depends on other USB devices could explain it!
 
I would say offhand about 30. I only have my negatives in strips of 4
though, and everytime I ejected the film holder it seemed to dump the
memory. I'd say that's a good thing though. The ScanDual III only has a
maximum resolution of 2820 and all scans were at that size in the 24 bit
format. I did notice though that the output tab for print size was set at
only 6 x 4 inches. I normally had this set at 8 x 10 inches. I'm not sure
I should change this or not. Hamrick software recommends changing as few of
the original settings as possible. One thing different as well that I did
do, was to leave the original oriantation of the image as it was scanned and
rotated later in an editor. This is recommended as well for memory issues.
I've tried that in the past and it didn't help though. In the past I've
also tried release memory on save option. Last night I did not have that
checked. Have you tried that yet? You can also manually release the memory
anytime under scanner menu. I had a ram viewer/optimizer running to watch
my available ram as I worked as well. It was not set to optimize anything
though, just view available resources from my system tray. I watched it go
down into the red to only about 5 meg of ram left and still vuescan did not
mess up in any way. I was pleased and shocked to say the least. Vuescan
did not use up all this ram on it's own, this was when I was testing by
running processes on an image editor at the same time as a scan was running.
Not the proper way to scan, but I thought a good test of the stability of my
system and vuescan as well. I was using sharpen and grain reduction
settings all the way up to maximum for this test as well. These settings
were alternated as part of testing my output. The ScanDual III puts out a
very noisy scan unfortuanately and at some point, whether in the editor or
vuescan, it has to be cleaned up. It would be nice if Vuescan supported a
few more options in regard to sharpening and grain reduction. My best image
was created with all these options off and cleaning up later however. This
took a lot of work though as well. For a quick image that is usable for
most things vuescans output is pretty good.

I should have tried taking all other USB devices off of my system in the
past I suppose, but I never did. Vuescan was functioning last night with
the following devices also connected and powered up (although not in use):
HP Deskjet 5650, Epson 2400 scanner, and of course the ScanDual III
actually doing the scanning. Drivers for the HP were also just updated
before this test as well though.

I would recommend updating your operating system for all latest drivers,
Service Packs, and security updates as mine is currently. At least for Win
2000. I was less then pleased at times with Win XP and it had seemed that
putting all available updates on it significantly slowed it down. To my
notion, antivirus programs can be at fault sometimes and turning real time
antivirus off temporarily for scanning might be a good idea.

I notice all settings are not changed by dumping vuescan.ini file and
restarting vuescan. Obviously there are entries in the registry that
remain. Does anyone happen to know exactly where and which entries are
changed, entered or altered by vuescan on installation and usage? Cleaning
the registry up from previous problems might solve things for a lot of folks
possibly. If anyone reading has any documentation on this I'd greatly
appreciate some information for access.

Good luck with your endeavors

Dave
 
David C Miers said:
I would say offhand about 30. The ScanDual III only has a
maximum resolution of 2820 and all scans were at that size in the 24 bit
format.

Thank you for the detailed information Dave. 30 consecutive successful
scans would be a dream at 4000 dpi. The fact that you are running at
2820 might help to explain the differences, as calculations show
below.

I have been using release memory on save and also clearing memory
explicitly under the scanner menu after every negative strip (unless
it crashes before that!). FWIW, here are the settings I have been
using:

Prefs|Scan mem set to 350 or 400 (Preview mem left at default 80)
Prefs|Refresh fast and Refresh each scan both unticked
Prefs|Refresh delay set to zero (which turns off refresh)
Prefs|Clear before scan and Release memory both ticked.
Filter|Infrared clean set to Light. No other filtering selected.
Input|Batch scan set to List
Input|Batch list set to 1-3 or 1-4 depending on number of frames on
strip.
Input|Preview resolution 500
Input|Scan resolution 4000
Input|Auto save set to None
Input|Rotation set to Left
Input|Autofocus set to Always
Output tab – only raw file selected, with 64 bit RGBI, size reduction
of 2, and output with Scan.

I haven't tried turning rotation off but rotation doesn't actually
affect the raw file, which is all I am outputting. Nor have I tried
turning Infrared cleaning off as again, that does not affect a raw
file output with scan (but it would affect a raw file output with
save). Might be worth trying anyway!

As I said, I have no virus checker running when I scan. At very least,
I turn off the auto checking and furthermore, in my diagnostic tests,
I abandoned all processes using ctrl+alt+delete apart from Explorer
(which is the OS itself) and Systray. So not even the antivirus task
that allows you to start up auto checking was running.

My OS is fully up to date with patches, security and otherwise. The
only driver I am not sure about is usbscan.sys. However, it is
difficult to determine what the latest version of this actually is as
there is conflicting information between the several websites from
which it can be downloaded. Some sites say the version I have is the
latest. Others have the date a couple of months later. In general, I
am uncomfortable about messing with basic system drivers unless there
is strong evidence of a clear need. It can sometimes result in
stopping anything from working at all. I'm glad I'm on Windows98SE.
From what I see, I suspect the XP in WindowsXP stands for Xtra
Problems!

I also had the system monitor running to see what was happening to
memory while scanning. During my run of consecutive successful scans,
unused physical memory often dropped to zero without failure. Yet,
when VueScan locked up, there was still unused memory available, even
up to 88MB. Processor usage never exceeded 55%.

I'm really struggling to find any logic in what is happening! What I
can say is that the problem occurs at 4000dpi but not at 2000dpi. So
it is hard to see how drivers could be implicated and memory
management seems more likely. But note that Roy Bishop had the problem
even with 768MB RAM. So is there a difficulty in the way VueScan
manages the memory rather than with how much RAM is present?

Ironically, the CoolScan V produces scans that are pretty well totally
noise free. On Slides, the noise in dark areas is barely perceptible.
On negatives, there is some grain aliasing, which is improved by
scanning at 4000 dpi and downsizing the raw file by 2 factor. This
gives a better result than scanning at 2000 dpi. But the difference is
not so great that it is worth the hassle of crashes. Grain cleaning in
editor is adequate on the few occasions that it is needed. Infrared
dust cleaning of the scan is brilliant, however! Moreover, the colour
balance achieved by VueScan is amazing.

It is interesting to compare the memory requirements of what you are
doing on ScanDual III with what I am doing on CoolScan V.

36x24mm at 2820dpi is 10.2 megapixels (megapixel = 1024x1024 pixels).
Scanning at 3x16 bit depth (even if outputting at 3x8 bit) is 6 bytes
per pixel.
So ScanDual III at 2820, generates 61.2 MB raw file per scan.

36x24mm at 4000dpi is 20.5 megapixels.
CoolScan V scanning at 14 bit depth (taking same space as 16) with RGB
and Infrared, is effectively at 4x16bit depth, i.e 8 bytes per pixel.
So CoolScan V at 4000 dpi generates 164 MB raw file per scan.

On the assumption that memory usage per scan is about twice the raw
file size, (to hold both the input raw file and the output file being
created and reformatted) the memory usage per scan would be 122MB for
the ScanDual III at 2820 dpi compared with 328MB for the CoolScan V at
4000 dpi. Hungry for RAM!

This might explain why VueScan can succeed with the ScanDual III at
maximum resolution whereas it has problems with the CoolScan V at
maximum resolution.

Using the CoolScan V at 2000 dpi would only need 328/4 = 82MB per scan
as it creates 2000 dpi files by only sampling a quarter as many pixels
from the CCD as it does for 4000 dpi. This could explain why VueScan
can succeed with CoolScan V at 2000 dpi.

I wonder if Ed will start to have more problems reported as more
people use film scanners generating these larger files. Although
flatbeds must already be generating huge files if used at full
resolution.

I also don't know if log files would help him since (a) I don't know
when it is going to crash and a log file of 8 consecutive good scans
followed by a lockup would be enormous (b) I don't know if it writes
to the log file when it locks up completely anyway.

Best wishes, Stephen
 
Comments in text
I haven't tried turning rotation off but rotation doesn't actually
affect the raw file, which is all I am outputting. Nor have I tried
turning Infrared cleaning off as again, that does not affect a raw
file output with scan (but it would affect a raw file output with
save). Might be worth trying anyway!

Even though your outputting only a raw file, I think the program is still
has to hold almost as much in memory for viewing. So I'm not at all sure
outputting a raw file saves you anything here or makes the rotation option
less important.
As I said, I have no virus checker running when I scan. At very least,
I turn off the auto checking and furthermore, in my diagnostic tests,
I abandoned all processes using ctrl+alt+delete apart from Explorer
(which is the OS itself) and Systray. So not even the antivirus task
that allows you to start up auto checking was running.

My OS is fully up to date with patches, security and otherwise. The
only driver I am not sure about is usbscan.sys. However, it is
difficult to determine what the latest version of this actually is as
there is conflicting information between the several websites from
which it can be downloaded. Some sites say the version I have is the
latest. Others have the date a couple of months later. In general, I
am uncomfortable about messing with basic system drivers unless there
is strong evidence of a clear need. It can sometimes result in
stopping anything from working at all. I'm glad I'm on Windows98SE.
From what I see, I suspect the XP in WindowsXP stands for Xtra
Problems!

Leaving win 98 in the dust was one of the best things I ever did. I greatly
prefer even xp to 98. However xp is very resource hungry. I never felt my
overall system was anywhere near as stable with 98 as with 2000 or xp. xp
is of course based on win 2000 NT. When I had an application error on 98 I
usually had to reboot the entire system, but with either 2000 or xp usually
closing the app is all you have to do and the overall system is still
stable. I truly believe any system running at least a 500mhz processor and
128meg of ram will benifit from updating to win 2000. To consider XP at
least a 800mhz processor and 256 meg of ram would be minimums IMHO. However
these are minumum operation system requirements to my notion (Microsoft may
well indicate differently), and image processing requires a good deal more.
I realize there are alot of diehard 98 users out there using their system
successfully and as long as it works for you more power to you, these are
just my opinions. Please, no flaming on this issue! One further note for
XP & ME users out there needing more power, you can get it simply by turning
off system restore functions, but of course you lose the ability to restore
your computer to an earlier date if you have problems.

I also had the system monitor running to see what was happening to
memory while scanning. During my run of consecutive successful scans,
unused physical memory often dropped to zero without failure. Yet,
when VueScan locked up, there was still unused memory available, even
up to 88MB. Processor usage never exceeded 55%.

I'm really struggling to find any logic in what is happening! What I
can say is that the problem occurs at 4000dpi but not at 2000dpi. So
it is hard to see how drivers could be implicated and memory
management seems more likely. But note that Roy Bishop had the problem
even with 768MB RAM. So is there a difficulty in the way VueScan
manages the memory rather than with how much RAM is present?

Ironically, the CoolScan V produces scans that are pretty well totally
noise free. On Slides, the noise in dark areas is barely perceptible.
On negatives, there is some grain aliasing, which is improved by
scanning at 4000 dpi and downsizing the raw file by 2 factor. This
gives a better result than scanning at 2000 dpi. But the difference is
not so great that it is worth the hassle of crashes. Grain cleaning in
editor is adequate on the few occasions that it is needed. Infrared
dust cleaning of the scan is brilliant, however! Moreover, the colour
balance achieved by VueScan is amazing.

It is interesting to compare the memory requirements of what you are
doing on ScanDual III with what I am doing on CoolScan V.

36x24mm at 2820dpi is 10.2 megapixels (megapixel = 1024x1024 pixels).
Scanning at 3x16 bit depth (even if outputting at 3x8 bit) is 6 bytes
per pixel.
So ScanDual III at 2820, generates 61.2 MB raw file per scan.

I've only been scanning in vuescan at 8bit since installing the latest
version due to image editors not accepting true 16 bit format. I've found
of the editors I possess right now, Paint Shop Pro 7 to be the best at
cleaning up the noise in my scans and then I switch to Adobe Photoshop 7 for
final color adjustments. Although I often use PSP to get close to the
proper skin coloration simply because it's so easy there. Vuescan will
clean up my noise quite handily, but loses more detail then PSP from my
experience so far. If I don't need the detail though, vuescan's grain
cleaning is a bit more pleasant to the eye I think. Dealing with the noise
in my scans is still a work in progress and I need more experience here
before I make absolute statements.
36x24mm at 4000dpi is 20.5 megapixels.
CoolScan V scanning at 14 bit depth (taking same space as 16) with RGB
and Infrared, is effectively at 4x16bit depth, i.e 8 bytes per pixel.
So CoolScan V at 4000 dpi generates 164 MB raw file per scan.

On the assumption that memory usage per scan is about twice the raw
file size, (to hold both the input raw file and the output file being
created and reformatted) the memory usage per scan would be 122MB for
the ScanDual III at 2820 dpi compared with 328MB for the CoolScan V at
4000 dpi. Hungry for RAM!

This might explain why VueScan can succeed with the ScanDual III at
maximum resolution whereas it has problems with the CoolScan V at
maximum resolution.

Using the CoolScan V at 2000 dpi would only need 328/4 = 82MB per scan
as it creates 2000 dpi files by only sampling a quarter as many pixels
from the CCD as it does for 4000 dpi. This could explain why VueScan
can succeed with CoolScan V at 2000 dpi.

I wonder if Ed will start to have more problems reported as more
people use film scanners generating these larger files. Although
flatbeds must already be generating huge files if used at full
resolution.

I also don't know if log files would help him since (a) I don't know
when it is going to crash and a log file of 8 consecutive good scans
followed by a lockup would be enormous (b) I don't know if it writes
to the log file when it locks up completely anyway.

Best wishes, Stephen


Vuescan is puzzling at times for sure and I really don't know for sure how I
resolved my problems as there were too many changes in my system without
retrying vuescan at each change, not to mention just installing the latest
version. I will think twice about updating vuescan though as long as it's
working. I had it working a couple of other times and updating to new
versions sometimes brought back old problems. Again I ask for info on the
registry modification by vuescan as I have to suspect that cleaning it out
and starting over might be of help to problematic vuescan installations.

Best of luck, Dave.
 
David C Miers said:
Vuescan is puzzling at times for sure and I really don't know for sure how I
resolved my problems as there were too many changes in my system without
retrying vuescan at each change, not to mention just installing the latest
version. I will think twice about updating vuescan though as long as it's
working. I had it working a couple of other times and updating to new
versions sometimes brought back old problems. Again I ask for info on the
registry modification by vuescan as I have to suspect that cleaning it out
and starting over might be of help to problematic vuescan installations.

Yes this would be useful. Also it would be really helpful to know how
VueScan actually uses the Scan mem setting or have some other guidance
on how to optimise its value. In particular, is there any reason not
to set it very high, even to considerably more than the amount of RAM
in the PC? Might it help to set it to 999MB for example? (For Windows
will use the swap file as virtual memory in any case). Does VueScan
simply avoid using more RAM than the figure chosen and collapse if
more memory is needed for a single scan than that chosen setting? Or
if the figure is set too high, is that amount reserved for scans and
therefore unavailable even to the VueScan program itself so that it
could collapse through having insufficient usable program memory?
Even though your outputting only a raw file, I think the program is still
has to hold almost as much in memory for viewing.

Absolutely. That is why I am assuming it needs at least twice as much
as the size of the raw file, regardless of whether one is outputting a
raw or an image file.
So I'm not at all sure outputting a raw file saves you anything here or makes > the rotation option less important

I am using the raw file output not to save memory but to enable me to
experiment easily later with different colour settings etc without
having to re-scan the film. But I agree about the rotation. Even
though rotation does not affect the raw file (and nor do any colour
settings) memory might still be used or allocated for rotation.
I've only been scanning in vuescan at 8bit since installing the latest
version due to image editors not accepting true 16 bit format.

If I understand rightly, the format that image editors can accept is
not relevant. The reason to scan at 16 bit depth is that this makes
more information available to the scanning software (VueScan here) to
enable it to generate the best output, even if that output is at 8
bit. For best quality, you should scan at the greatest available bit
depth (although I have not tested to see how noticeable the difference
is in practice).

I've found of the editors I possess right now, Paint Shop Pro 7 to be the >best at….

I too am a fan of PSP7!!

Regards
Stephen
 
Back
Top