Question re: Coolscan V ED scan times, etc.

[keyes04]

The advertised scan time for the Nikon Coolscan V ED is 38 seconds.
Is that for color film, or B&W? How much does it increase once
Digital ICE is added to the equation? Any other comments on the
scanner will be appreciated, as I'm fairly near a buying decision, but
don't want to be disappointed.

I also have Vuescan and wonder about experiences with that & the
Coolscan V. It appears from what I've read here recently that scan
times are considerably longer with Vuescan vs. Nikon's software. What
are the advantages and disadvantages of Vuescan over the Nikon
software?

Thanks.

C.R.

 

[Denis Freund]

The advertised scan time for the Nikon Coolscan V ED is 38 seconds.
Is that for color film, or B&W?

preview

 

[keyes04]

preview

From the Nikon site itself:

"In about 38 seconds, you’ll have a high-quality digital image at
4,000 dpi."
http://www.nikon-image.com/eng/news_release/2003/ls50ed.htm

4000 dpi previews??? I'm missing something here, if you're correct.
I've read in one review that the preview time is only 14 seconds, with
a full scan of 38 seconds. I just can't find any reviews which
discuss real world scan times with and without use of the supplied
software (such as Digital ICE), or Nikon software vs. Vuescan. I
fully expect that scan times will be longer than 38 seconds with
anything except a very basic, unenhanced scan. I just have no idea
how much.

C.R.

 

[Roger Halstead]

The advertised scan time for the Nikon Coolscan V ED is 38 seconds.
Is that for color film, or B&W? How much does it increase once

Normally Nikon used the basic scan (including the preview) for their
figures. With the V ED I'd "assume" it is using NikonScan.

In the above case it should make no difference whether the scan is in
B&W or color (slide or negative) One note though. Digital ICE does
basically nothing with B & W negatives nor is it effective on many
batches of Kodachrome slides. I probably should say most instead of
many.

Digital ICE is added to the equation? Any other comments on the

Digital ICE requires a "pre-scan" which is done in the IR region so it
will essentially double the *basic* scan time, or at least it does in
the LS5000 ED and I'd expect a similar performance in the V ED.
What adds time are the post processing applications such as Restoring
the original color for faded slides (ROC), Grain elimination (GEM),
and shadow enhancement to bring out detail (DEE)

scanner will be appreciated, as I'm fairly near a buying decision, but
don't want to be disappointed.

You will not be disappointed with Digital ICE. The current versions
work great. They are no substitute for cleaning, but they do their
work without the loss of resolution prevalent in post processing dust
and scratch removal. OTOH it does take time.

HOWEVER and I have to emphasize this is true for all scanners that I
know of, The times gives by the manufacturers are for basic scans, but
they are indicative of relative performance between scanners given
specific software.

If the V ED is relatively fast in the basic scan it will "most likely"
be relatively fast compared to other scanners when doing the same
processes.

I also have Vuescan and wonder about experiences with that & the
Coolscan V. It appears from what I've read here recently that scan
times are considerably longer with Vuescan vs. Nikon's software. What

It depends on how you compare them. I run Digital ice on all scans.
I also use both ViewScan and NikonScan. Each has its plusses and
minuses. When using the same settings I notice little difference
between the two. Having said that, I know the times to scan for
NikonScan down to the second, but I have not timed them on ViewScan.
For most applications I prefer ViewScan. I also know that depending
on the settings can increase scan times considerably. So the two are
difficult to compare in that respect. For basic scans with Digital
ICE I find ViewScan to be quite fast and it doesn't have the scanner
setting there waiting as does NikonScan (referring to the LS5000 ED)

My only complaint about ViewScan is it sometimes has difficulty
detecting the space between images, but so does NikonScan on occasion.
However NikonScan is much better in this respect. Until you learn
ViewScan you sometimes get some strange results. The help files and
author are very helpful and conscientious. Nikon it appears, could
care less.

Once you learn ViewScan and it's abilities, it does an outstanding
job. (most of the time). NikonScan tends to be a bit on the buggy side
as well as a CPU hog. With NikonScan being a CPU hog and a bit buggy
I've had it cause my computer to hang (Wouldn't even recognize the
three finger salute <CTRL><ALT><DEL>) more than any other application
and this computer runs a lot of apps, usually at the same time.

NikonScan running the CPU at 100% is not multi application friendly.
ViewScan typically runs at 30%. NikonScan also seems to have a real
problem with both thin and dense slides and negatives.

are the advantages and disadvantages of Vuescan over the Nikon
software?

I guess I already answered this above.
The only thing I'd add is that ViewScan has a lot more individual
features so it has a bit steeper learning curve to start. It's well
worth taking the time to read the documentation and experiment.

I also find the colors to be more true when using ViewScan set up
properly.

As an added note, if scanning images at 4000 dpi, I would want a fair
amount of horsepower in the computer. From my own experience I prefer
at least a 2 Gig CPU, one Gig of RAM and a BIG hard drive.

Don't forget a good CD or DVD burner for archiving and be sure to back
up anything you think is worth saving.

Backing up and archiving are two topics by themselves that can get
quite involved.

Good Luck

Roger Halstead (K8RI & ARRL life member)
(N833R, S# CD-2 Worlds oldest Debonair)
www.rogerhalstead.com

 

[Kennedy McEwen]

The advertised scan time for the Nikon Coolscan V ED is 38 seconds.
Is that for color film, or B&W? How much does it increase once
Digital ICE is added to the equation? Any other comments on the
scanner will be appreciated, as I'm fairly near a buying decision, but
don't want to be disappointed.

Forget Denis, his sense of humour is on vacation!

The CS-V is similar to the CS-4000, without the bulk options and a USB
interface, so the times are similar.

That time is for a colour slide 4000ppi full scan without ICE or any
other processing such as GEM or ROC, or even any prescan, autofocus or
whatever, which can be implemented as part of the preview in any case.

Adding ICE is an open feast, depending on how much ICE intervenes (how
dirty, scratched or damaged is the slide) and the speed of your
computer. On my 3GHz machine it adds around 20seconds on an 8-bit scan
of a very old dirty slide, and around 35seconds on a 14-bit scan. The
real time hog is GEM, which adds another 80seconds.

I doubt that the CS-V is significantly different from these times,
especially as it uses the same driver.

 

[Scott]

Is there a specific film holder for the V or is it direct loading?

My old Minolta had a film strip holder but I seem to recall something
about direct loading with the V. (There are none within miles of me
so I can't just go look and play.)

Thanks!
Scott

 

[Kennedy McEwen]

Is there a specific film holder for the V or is it direct loading?

My old Minolta had a film strip holder but I seem to recall something
about direct loading with the V. (There are none within miles of me
so I can't just go look and play.)

Both.

The CS-V comes with 3 film adapters:
SA-21: motorised direct feed for 2-6 frame strips of film
MA-21: single frame mounted slide adapter

There is also an optional FH-3 film strip holder that slides into the
MA-21 and permits scanning one frame at a time.

 

[Roger Halstead]

Forget Denis, his sense of humour is on vacation!

The CS-V is similar to the CS-4000, without the bulk options and a USB
interface, so the times are similar.

That time is for a colour slide 4000ppi full scan without ICE or any
other processing such as GEM or ROC, or even any prescan, autofocus or
whatever, which can be implemented as part of the preview in any case.

Adding ICE is an open feast, depending on how much ICE intervenes (how
dirty, scratched or damaged is the slide) and the speed of your
computer. On my 3GHz machine it adds around 20seconds on an 8-bit scan
of a very old dirty slide, and around 35seconds on a 14-bit scan. The
real time hog is GEM, which adds another 80seconds.

Using a 2.8 Gig XP Plus Athlon with 1 gig of RAM, the 5000 scans in
20 seconds and digital ICE seems to take about 20 seconds no matter
how bad the slide or negative. I've run about 20 film strips and
negative strips as the time always averages to 40 seconds. Even with
the SF210 the time per slide was about 45 seconds.

I don't think there is much difference between the V, 4000, and 5000
as far as the works except the feed system.

Like GEM I found both DEE and ROC to be real resource hogs, at least
on this machine. Running Digital ICE, GEM, ROC, and DEE can push the
time per image into the neighborhood of 3 minutes.

I doubt that the CS-V is significantly different from these times,
especially as it uses the same driver.

Roger Halstead (K8RI & ARRL life member)
(N833R, S# CD-2 Worlds oldest Debonair)
www.rogerhalstead.com

 

[Denis Freund]

4000 dpi previews??? I'm missing something here, if you're correct.

http://www.filmscanner.info/NikonCoolscan5ED.html

It´s a good german site, you don´t have to read through, but look at the
tables.
The word "Vorschau" means preview, "Dia" means slide:

Vorschau 1 Dia 0:36 min 0:47 min (ICE)
Vorschau 1 Negativ 0:38 min 0:46 min (ICE)

Single Scan Slide 4000 dpi (ICE) 1:56 min
Single Scan Negativ 4000 dpi (ICE) 3:07 min

At least the negativ Scan is a bit faster with my Coolscan V but all in
all the times are quite correct.

best regards

Denis

 

[Denis Freund]

4000 dpi previews??? I'm missing something here, if you're correct.

I must admit, I have only looked for the scan time and didn´t notice the
4000 dpi. Anyhow look at:
http://www.filmscanner.info/NikonCoolscan5ED.html

It´s a good german site, you don´t have to read through, but look at the
tables.
The word "Vorschau" means preview, "Dia" means slide:

Vorschau 1 Dia 0:36 min 0:47 min (ICE)
Vorschau 1 Negativ 0:38 min 0:46 min (ICE)

Single Scan Slide 4000 dpi (ICE) 1:56 min
Single Scan Negativ 4000 dpi (ICE) 3:07 min

The used computer wasn´t very fast, which should primarily affect
processing time of the Digitel ICE features.
Especially the negativ Scan is a bit faster with my Coolscan V but all
in all the times are a good indicator.

best regards

 

[Denis Freund]

4000 dpi previews??? I'm missing something here, if you're correct.

I must admit, I have only looked for the scan time and didn´t notice the
4000 dpi. Anyhow look at:
http://www.filmscanner.info/NikonCoolscan5ED.html

It´s a good german site, you don´t have to read through, but look at the
tables.
The word "Vorschau" means preview, "Dia" means slide:

Vorschau 1 Dia 0:36 min 0:47 min (ICE)
Vorschau 1 Negativ 0:38 min 0:46 min (ICE)

Single Scan Slide 4000 dpi (ICE) 1:56 min
Single Scan Negativ 4000 dpi (ICE) 3:07 min

Especially the negativ Scan is a bit faster with my Coolscan V but all
in all the times are a good indicator.

best regards

Denis

 

[Kennedy McEwen]

Using a 2.8 Gig XP Plus Athlon with 1 gig of RAM, the 5000 scans in
20 seconds and digital ICE seems to take about 20 seconds no matter
how bad the slide or negative. I've run about 20 film strips and
negative strips as the time always averages to 40 seconds. Even with
the SF210 the time per slide was about 45 seconds.

I don't think there is much difference between the V, 4000, and 5000
as far as the works except the feed system.

There is an enormous difference between the LS-5000 and the other two
scanners, Roger. The LS-5000 uses two rows of CCDs in a technique
called "Time Delay and Integrate" or TDI to halve the scan time yet
retain the same sensitivity. That is why the LS-5000 scan times are
about half as long as the LS-50 or the LS-4000, which are essentially
the same scan engine with different interfaces.

Judging from your other posts, there is also another difference which I
have not had an opportunity to test for myself. You seem to find that
ICE causes a second pass of the scan head, indicating that the infrared
channel is captured separately from the RGB channels. This is certainly
not the case with the LS-4000, which appears to capture all RGBI on a
single pass, although the scan itself is slower due to the need to
process the ICE data. Of you scan a slide with very dirty corners, you
can hear the scanner slow down at the start and end of the slide because
there is more ICE data to process. In fact, this single pass ICE has
been one of the major advantages of Nikon scanners over other devices,
since it ensures perfect alignment of the infrared data with the RGB
image, so I am a little surprised to have seen you report this has been
changed in the LS-5000. However if indeed this is the case then it
could explain why you do not actually notice any significant time
difference as a consequence of the cleanliness of the slide.

Like GEM I found both DEE and ROC to be real resource hogs, at least
on this machine. Running Digital ICE, GEM, ROC, and DEE can push the
time per image into the neighborhood of 3 minutes.

ROC doesn't seem to add too much of a delay compared to GEM, perhaps a
few seconds, however enabling autoexposure and autofocus before each
scan, together with all the processing options can get up to that sort
of time - and its 30sec quicker if I use NikonScan 3.1.2 instead of
NS4.01. All of which makes the 20 seconds improvement in scan speed of
the LS-5000 over the LS-4000 pretty insignificant really - at least for
the current generation of computer technology.

I wonder what the odds are on Nikon continuing to support either scanner
after Windoze and computer speeds have evolved enough to make the
difference between the LS-4000 and LS-5000 significant. If we continue
with the speed equivalent of Moore's Law, that means something like 10
years time for that 3 minute processing overhead to reduce to less than
the 20 seconds difference in scan time. It will take a major change in
practice, and a complete reversal of current trends, for Nikon to
continue to support any product after that sort of period. ;-)

 

[Roger Halstead]

There is an enormous difference between the LS-5000 and the other two
scanners, Roger. The LS-5000 uses two rows of CCDs in a technique
called "Time Delay and Integrate" or TDI to halve the scan time yet
retain the same sensitivity. That is why the LS-5000 scan times are
about half as long as the LS-50 or the LS-4000, which are essentially
the same scan engine with different interfaces.

Judging from your other posts, there is also another difference which I
have not had an opportunity to test for myself. You seem to find that
ICE causes a second pass of the scan head, indicating that the infrared
channel is captured separately from the RGB channels. This is certainly

I based my assumptions on the scanning times and the time the scanner
motor is running. I've run these tests with multiple images many
times, but your post got me to thinking so I just finished running
another series of tests. Perhaps you can add some interpetation.

The one thing that came out, when running a single scan was no
seperate scanning was heard or seen on the CPU. BUT for some reason
times are running consistently 25% longer today:

AT any rate here is the data:

Insert a strip of negatives.
Positioning runs with CPU at 100% for about 4 seconds. Then the
preview runs for each frame at roughly 4 seconds each and 100% CPU
which is consistent with past results using NikonScan.

Scan all 4 images, no crop and no ICE.
25 sec each. This is about 5 seconds or about 25% longer than
previous tests.

Scan all 4 images, crop to 35mm frame size.
23 seconds each. Again this is longer than previous tests which were
also about 20 seconds.

Scan all 4 images with digital ICE
Position motor runs, then scanning sound for 10 seconds and CPU at
100%. Motor repositions and scanning sound for 28 seconds. The second
scan takes 28 seconds, but the processing and scan together take 43
seconds (not counting the first 10 second scan) before the motor
positions for the next scan. This totals 53 seconds plus maybe 2 for
repositioning for 55 seconds compared to 40 seconds average or 25%
longer than in the past tests.

The extra processing time is present only when Digital ICE is used.
AT least the differences are consistent even if I'm not sure what is
causing them.

So, I started over and ran the test with no cropping, with cropping,
and finally with cropping and ICE. The results were identical to the
above figures within my ability to measure with a stop watch.

HOWEVER!

This time I ran an additional single scan with digital ice. Once the
motor positioned there was only one continuous scan of 28 seconds
with a total processing time of 43 seconds.

Prior to this I have run the scan with and without ICE on 4, 5 and 6
frame strips with very consistent results of 20 seconds average for
the plain (cropped to 35mm size) and 40 seconds when running digital
ICE including the processing time.

Today the times are considerably longer and I am seeing the much
longer CPU times which were not present before.

This leads me to the conclusion that there is indeed only one scan for
ICE and the image. I was mistakenly figuring the 10 second scan time
which is there only when scanning with digital ICE turned on. However,
as that 10 second segment is not there when doing a single image, I'm
not sure what it is. It apparently has something to do with ICE, but
does not show up when doing a single image. Nor does the single image
have that 10 seconds added to the scan time.

Any one have an idea as to what it might be?

As to the longer times "it appears" there may be some page file
swapping going on, although running PhotoShop elements and Word at the
same time made no difference in the scan, or processing times.

This is with a 2.8 gig XP Plus Athlon with 1 Gig of DDR RAM and a 7200
RPM 200 Gig Hard Drive with a second 7200 RPM HD serving for swap
files. CPU temp is 48 degrees which is about a cool as I see it get.



Roger Halstead (K8RI & ARRL life member)
(N833R, S# CD-2 Worlds oldest Debonair)
www.rogerhalstead.com

 

[Kennedy McEwen]

HOWEVER!

This time I ran an additional single scan with digital ice. Once the
motor positioned there was only one continuous scan of 28 seconds
with a total processing time of 43 seconds.

Prior to this I have run the scan with and without ICE on 4, 5 and 6
frame strips with very consistent results of 20 seconds average for
the plain (cropped to 35mm size) and 40 seconds when running digital
ICE including the processing time.

Today the times are considerably longer and I am seeing the much
longer CPU times which were not present before.

This leads me to the conclusion that there is indeed only one scan for
ICE and the image. I was mistakenly figuring the 10 second scan time
which is there only when scanning with digital ICE turned on. However,
as that 10 second segment is not there when doing a single image, I'm
not sure what it is. It apparently has something to do with ICE, but
does not show up when doing a single image. Nor does the single image
have that 10 seconds added to the scan time.

Any one have an idea as to what it might be?

What are your settings in Preferences for autoexposure and autofocus for
single frame, batch and preview scans? Are they different?

It might be an autoexposure of the IR channel, which would only require
to be done once, so subsequent scans of the same frame, even at
different resolutions or crops would not require a further prescan. It
would have to be some sort of sub-resolution scan, such as autoexposure
because it completes the pass in 10 seconds, which is half the normal
scan time for a full resolution pass.

As to the longer times "it appears" there may be some page file
swapping going on,

possibly, but you should see that on the disk access lamp.

 

[Roger Halstead]

<snip details of previous batch scan test>

What are your settings in Preferences for autoexposure and autofocus for
single frame, batch and preview scans? Are they different?

Autofocus and autoexposure are both set for every image.

It might be an autoexposure of the IR channel, which would only require
to be done once, so subsequent scans of the same frame, even at
different resolutions or crops would not require a further prescan. It
would have to be some sort of sub-resolution scan, such as autoexposure
because it completes the pass in 10 seconds, which is half the normal
scan time for a full resolution pass.

possibly, but you should see that on the disk access lamp.

I've been watching for that, but none seems to be happening according
to the drive light., yet the times have changed considerably compared
to the earlier tests I ran.

I've also defraged the drive which now has quite a bit more on it, but
as the processing is going on in memory I'd not expect this to affect
it one way or another.

The times are still reasonable, but I baffles me as to why the times
have changed as much as they have.

Roger Halstead (K8RI & ARRL life member)
(N833R, S# CD-2 Worlds oldest Debonair)
www.rogerhalstead.com

 

[Kennedy McEwen]

I've been watching for that, but none seems to be happening according
to the drive light., yet the times have changed considerably compared
to the earlier tests I ran.

I've also defraged the drive which now has quite a bit more on it, but
as the processing is going on in memory I'd not expect this to affect
it one way or another.

The times are still reasonable, but I baffles me as to why the times
have changed as much as they have.

Upgraded the software? I found that NS4 was considerably slower than
NS3 with the LS-4000

 

[Roger Halstead]

Upgraded the software? I found that NS4 was considerably slower than
NS3 with the LS-4000

No, I've been running NS 4 since day one, but I am trying to figure
out what may have changed.

Roger Halstead (K8RI & ARRL life member)
(N833R, S# CD-2 Worlds oldest Debonair)
www.rogerhalstead.com