Nikon Coolscan 5000 - different results with NikonScan and Vuescan

[Ed Hasler]

Hi,

I've recently purchased a Nikon Coolscan 5000 scanner to replace my
Canon FS4000 (which I thought was faulty, but it turns out is not -
doh!), and having compared the two have decided to keep the Nikon and
sell the Canon.

I've also been comparing the results I get when using NikonScan 4 and
Vuescan, and have noticed the following:

1) Colour matching

I've created a scanner profile for Vuescan using Wolf Faust's IT8
targets, and using this, scans from Vuescan are very slightly closer
to the original slide than NikonScan. I can get almost exactly the
same results from NikonScan, but this seems to take a bit more
tweaking whereas Vuescan is pretty much spot straight away.

I've set NikonScan to use my OptiCal created monitor profile, and
output in Adobe 1998 (which is my preferred Photoshop workspace)

I'm guessing I just need to play around with NikonScan a bit more, and
work out how to use the histogram, so am not too worried about this
but is it normal to have to make minor tweaks to colours, or should it
be possible to get very close with default settings?

Overall, I prefer Vuescan in this respect, but it's very close.

2) Grain/pepper spots

The Vuescan scan is slightly less grainy than the NikonScan output,
whereas the NS scan is less smooth and also small black "pepper" spots
dotted around. Running an unsharp mask filter on both scans with the
same settings makes the difference even greater, i.e., the Vuescan
image looks even better compared to the NikonScan image. Vuescan is
run with "Grain reduction" set to Medium, and I just used the Default
settings for NikonScan.

Overall, Vuescan appears to be better.

3) Sharpness

With sharpening turned off in both Vuescan and NikonScan, the
NikonScan images seem very slightly sharper, however this may relate
to the following point:

4) Jagged lines/big pixels

When I look at each scan (saved as a Tiff file, then opened in
Photoshop 7), there is a noticeable difference between the two images.

Images scanned in Vuescan appear to be made up of bigger pixels, and
lines appear to be more jagged. This is obvious when you compare two
images side-by-side at 100% scale on screen, however I've yet to print
both images to see if it makes a difference on paper.

In this respect, NikonScan is noticeably better (on screen at least).

Overall, I prefer the colour matching abilities of Vuescan, along with
the lack of grain, however NikonScan appears to produce a sharper
image with smaller pixels and smoother lines.

Does anyone have any ideas/comments as to why this might be?

For information, I'm running Windows XP on a P4 PC with 1GB RAM, and
my monitor is calibrated using OptiCal and a Colorvision Spyder.

Thanks in advance for any comments or advice.

Ed

 

[Luca Amateis]

(e-mail address removed) (Ed Hasler) wrote in

2) Grain/pepper spots

The Vuescan scan is slightly less grainy than the NikonScan output,
whereas the NS scan is less smooth and also small black "pepper" spots
dotted around. Running an unsharp mask filter on both scans with the
same settings makes the difference even greater, i.e., the Vuescan
image looks even better compared to the NikonScan image. Vuescan is
run with "Grain reduction" set to Medium, and I just used the Default
settings for NikonScan.

Overall, Vuescan appears to be better.

I own a Canoscan FS4000 and using grain reduction set to medium in Vuescan
made a visible difference in sharpness, at least on screen. Does the
default setting in Nikon Scan activate ICE? Perhaps you could try a scan
without grain reduction in both programs to better evaluate grain and
sharpness.

3) Sharpness

With sharpening turned off in both Vuescan and NikonScan, the
NikonScan images seem very slightly sharper, however this may relate
to the following point:

4) Jagged lines/big pixels

When I look at each scan (saved as a Tiff file, then opened in
Photoshop 7), there is a noticeable difference between the two images.

Images scanned in Vuescan appear to be made up of bigger pixels, and
lines appear to be more jagged. This is obvious when you compare two
images side-by-side at 100% scale on screen, however I've yet to print
both images to see if it makes a difference on paper.

This seems strange, is the output resolution the same? Again you could try
to scan without postprocessing with both programs.


Have you seen significal differences in image quality between the FS4000
and the Nikon? As I have told, I own a FS4000 and I was thinking about an
upgrade (I am considering the Coolscan V, not the 5000), but I don't know
how much this could improve my scans. Any information about quality (dmax,
sharpness, smoothness, etc) is appreciated.

Thanks in advance for any comments or advice.

Ed

Best regards,

Luca

 

[Ed Hasler]

Have you seen significal differences in image quality between the FS4000
and the Nikon? As I have told, I own a FS4000 and I was thinking about an
upgrade (I am considering the Coolscan V, not the 5000), but I don't know
how much this could improve my scans. Any information about quality (dmax,
sharpness, smoothness, etc) is appreciated.

To be honest, there wasn't a great deal of difference between the two
scanners.

I spent several hours scanning various different slides with both
scanners (using Vuescan with the same settings), and found the
following:

Grain

If anything, the Canon was slightly better at this - it produced a
smoother image when both were set to use no grain reduction, and the
Nikon only beat the Canon when set to Medium grain reduction (with
Canon still set to off)

Sharpness

There was virtually nothing in it, even at high zoom levels. I was
actually quite disappointed, as I'd hoped that a £1000 Nikon scanner
should be noticeably better than a £500 one, but they were very
similar.

Noise

This is one area where the Nikon really does beat the Canon. Dark
parts of the slide suffer from a fair bit of noise on the Canon even
with multiple passes. The same slide scanned on the Nikon has no
noticeable noise in the dark areas - even where it's totally black.
This was the main reason I decided to keep the Nikon and sell the
Canon.

Speed

This is the other main advantage that the Nikon has over the Canon -
it's a lot quicker. I've not done any timings, but it is certainly
quicker doing single scans, and even more so when you take into
account having to do multi-pass scans to reduce noise on the Canon
when scanning dark slides. The one advantage that the Canon has is the
4-slide feeder compared to the single slide capacity of the 5000.

Overall, the Nikon is better, but not by much at all, and I don't
think it's enough of an improvement to justify spending the extra
money.

If you want any more information, then please let me know.

Regards,

Ed

 

[Luca Amateis]

To be honest, there wasn't a great deal of difference between the two
scanners.

I spent several hours scanning various different slides with both
scanners (using Vuescan with the same settings), and found the
following:

Grain

If anything, the Canon was slightly better at this - it produced a
smoother image when both were set to use no grain reduction, and the
Nikon only beat the Canon when set to Medium grain reduction (with
Canon still set to off)

Sharpness

There was virtually nothing in it, even at high zoom levels. I was
actually quite disappointed, as I'd hoped that a £1000 Nikon scanner
should be noticeably better than a £500 one, but they were very
similar.

Noise

This is one area where the Nikon really does beat the Canon. Dark
parts of the slide suffer from a fair bit of noise on the Canon even
with multiple passes. The same slide scanned on the Nikon has no
noticeable noise in the dark areas - even where it's totally black.
This was the main reason I decided to keep the Nikon and sell the
Canon.

Speed

This is the other main advantage that the Nikon has over the Canon -
it's a lot quicker. I've not done any timings, but it is certainly
quicker doing single scans, and even more so when you take into
account having to do multi-pass scans to reduce noise on the Canon
when scanning dark slides. The one advantage that the Canon has is the
4-slide feeder compared to the single slide capacity of the 5000.

Overall, the Nikon is better, but not by much at all, and I don't
think it's enough of an improvement to justify spending the extra
money.

If you want any more information, then please let me know.

Regards,

Ed

Yes, noise in dark areas is the main reason I was thinking to upgrade. But
if there are no other real improvements (speed is not a big problem for me,
at least after switching to SCSI interface) I will wait something really
better. If I have a difficult slide I use Vuescan with long exposure pass
or multipass and I usually get decent results.
Thank you for your help. Now that the scanner is eliminated from my upgrade
list I can start thinking about a new photo printer.
Regards,

Luca

 

[Ed Lusby]

I find a HUGE difference between the quality of Nikon Scan profiles
for the LS5000 compared to a profile generated with Vuescan.
The Vuescan it8 generated profiles were extremely close to the
original slide, needing but minor cooling down. However, Nikonscan
reds were dull, the contrast of some slides much flatter than the
original, greens were dull, etc. Vuescan is far superior in my
opinion. And I don't see any difference in sharpness using the 2
programs.

Nikon Scan IMO has a fatal flaw in that it doesn't allow you to use
any profiles except their own ones, which appear not to be ICC
compatible, and are not good profiles to begin with. Why they do this
on such a professional instrument is beyond me.

However, I am not sure that the vuescan profiles are ICC compatible,
either. When I scan a file in raw mode in vuescan, and convert to
profile (generated by Vuescan) in Photoshop, the result is poor color
matching. I would have expected the color to match that obtained by
scanning using the Vuescan profile directly.

Of course, the profile issue is moot if you want to start with raw
scans, which is the best choice if you want to Photoshop the image
anyway. Raw scans look the same to me using either program.


Ed Lusby

 

[Aaron Queenan]

However, I am not sure that the vuescan profiles are ICC compatible,
either. When I scan a file in raw mode in vuescan, and convert to
profile (generated by Vuescan) in Photoshop, the result is poor color
matching. I would have expected the color to match that obtained by
scanning using the Vuescan profile directly.

That's because you shouldn't "convert" to the generated profile, you need to
"assign" the profile to the raw image (and then possibly "convert" to
AdobeRGB).

Regards
Aaron Queenan.