Epson 4490 Vibration & Noise with Vuescan-83

[rogerxx]

I'm noticing allot of noise with the Epson 4490 & using Vuescan-83.
(This is a newly purchased Epson 4490 & - no - I haven't put the weight
of a book on the scanner. ;-)

Although during preview scans, noise and only slight vibrations isn't
too much of a problem, the problems with vibrations may become a
problem during a hi-res scan (say, 4800dpi).

There is a loud audible "click" during very end of Pass #1 or the very
beginning of Pass #2. The motor speed implemented with Vuescan on Pass
#2 also seems to be providing strong vibrations. When one uses the
Epson WIndows software or the Linux Iscan software, I think a different
motor speed is used, providing smoother scans and no audible "clicks".

I'm guessing Vuescan is using a motor speed on Pass #2 that is causing
stronger vibrations? I'm also guessing this selected motor speed is
causing audible "clicks" when re-initializing or (start/end of
repositioning) of the (possibly lower?) scan unit. (Usually at the
very beginning of Pass #2 and the end of Pass #2.)


Possible Issues caused by vibrations?

Pass #1 being infrared and Pass #2 being RGB(?).

The reason this might be an issue, if the vibrations are strong enough,
they may cause the 35mm negative or other sizes to be jarred enough for
the images provided by Pass #1 and Pass #2 to not match exactly
providing a reduced image. Not only this, but Infrared Cleaning (or
Digital ICE for the patented version) maybe hindered because if the
images do not match, the smearing (or cleaning) of scratches on the
infrared image may not match the RGB image providing an image that was
just as good as without infrared cleaning ever being done. Of course,
any vibration or jar to the unit during these two passes will
definitely provide the results.

I've also seen a UK based report of using Vuescan with multiple scans
of negatives. The report stated the scans did not match exactly
providing a poorer image quality then when one just does one scan of
the image. My guess is, the reviewer might be experiencing similar
problems as I am as I have done a scan of a negative 2+ times and
experienced similar with poor image quality? Albeit, I think 1 or 2
passes should surffice as a negative can only provide so much visible
spectrum. From what I've seen, multiple scans of an image is usually
only done with astrophotography (ie. CCD/CMOS imaging devices mounted
on telescopes).

 

[David J. Littleboy]

I'm noticing allot of noise with the Epson 4490 & using Vuescan-83.
(This is a newly purchased Epson 4490 & - no - I haven't put the weight
of a book on the scanner. ;-)

Does the noise also occur with the Epson software?

I've also seen a UK based report of using Vuescan with multiple scans
of negatives. The report stated the scans did not match exactly
providing a poorer image quality then when one just does one scan of
the image.

I'd be _incredibly_ surprised if any consumer scanner could provide
pixel-perfect registration between scans. That's why I like the Nikon
design: since the light source has three colors, there's no need for three
separate CCDs and moving the film between R, G, and B exposures. Thus RGB
(and IR) are always exposed at exactly the same point on the film.

My guess is, the reviewer might be experiencing similar
problems as I am as I have done a scan of a negative 2+ times and
experienced similar with poor image quality? Albeit, I think 1 or 2
passes should surffice as a negative can only provide so much visible
spectrum. From what I've seen, multiple scans of an image is usually
only done with astrophotography (ie. CCD/CMOS imaging devices mounted
on telescopes).

Multiscanning is standard for scanning chromes to get lower noise in the
shadow areas. On the better Nikon scanners, it does the multiscanning in the
same pass (i.e. with hardware support), so pixel registration between
multiscan passes isn't an issue. There are people who think that
multiscanning is also useful for dense negatives, and there are people who
don't<g>. Come to think of it, I've had some XP2 frames in which the
highlights blocked up in a normal scan but which could be rescued (at the
cost of losing the shadows and mid-tones) with a longer exposure.

David J. Littleboy
Tokyo, Japan

 

[theo]

I'm noticing allot of noise with the Epson 4490 & using Vuescan-83.

I've used VueScan on Kodak 3570, Epson 2480, 3490, 4490, plus others also
with their bundled/native s/w. My wife via p/p network or direct has
used Plustek, Umax, current Epsons thru APSE2. Neither have ever found
the vibrations described above. I'll wait for others to determine
etiology and determinants, but shoot-from-the-hip guess is RMA to vendor.

 

[theo]

I'm noticing allot of noise with the Epson 4490 & using Vuescan-83.

I've used VueScan on Kodak 3570, Epson 2480, 3490, 4490, plus others also
with their bundled/native s/w. My wife via p/p network or direct has
used Plustek, Umax, current Epsons thru APSE2. Neither have ever found
the vibrations described above. I'll wait for others to determine
etiology and determinants, but shoot-from-the-hip guess is RMA to vendor.
Regards,
Theo
=======
Pessimists remain morose precisely because they are so right too often.

 

[Don]

I've also seen a UK based report of using Vuescan with multiple scans
of negatives. The report stated the scans did not match exactly
providing a poorer image quality then when one just does one scan of
the image.

Yes, that's when you use the so-called multi-pass multi-scanning. That
is to say, the scanner performs one full scan, then the CCD assembly
goes back and performs another, and then goes back... etc. These scans
will never match and all you do is manage to blur the image and waste
a lot of time in the process.

By contrast, single-pass multi-scanning advances to a scan line,
performs a scan, and another, and another... etc. Only when done, does
the assembly advance to the next line to scan it. And so on. So the
registration between all those multiple scans is perfect since the
assembly doesn't move between them.

The catch is that even when you use single-pass multi-scanning the
results are very modest because you are just averaging noisy input.
This is particularly the case for slides where a much better approach
is to scan twice, once for highlights and once for shadows and then
combine these two scans. For negatives, however, multi-scanning can be
useful because the compressed dynamic range of negatives is more prone
to amplifying noise.

Don.

 

[Henk de Jong]

Please ignore Don: he is a troll...

With kind regards,
Henk de Jong

 

[rogerxx]

Does the noise also occur with the Epson software?

Doubtful. I'm about to test the windows epson software as soon as I
finish a batch on my Linux box.

My guess is, the Windows Epson software uses an extremely more slower
motor speed when scanning at the higher dpi (ie. 4800 dpi). This is
the reason why using the windows epson software takes a long time for
scanning each negative -- guessing 5-10 minutes per negative & 20+
using Digital ICE.

Vuescan is faster when scanning. But the scan speed it uses causes
audible clicks at the beginning & ending of the scanner element during
Pass #2.