I've had my Nikon LS4000ED for about two months now the results have
o Many of my negative strips come from the lab with a fine surface
scratch running down the non-emulsion side of the negative. I have
an older filmscanner that entirely ignores these. Same with my new
flatbed scanner (ScanMaker 8700). But the Nikon not only "sees"
these scratches, it turns them into glaring white scars across the
length of the image. ICE does an incredible job of removing most or
all signs of these scars, but they shouldn't be there in the first
o The Nikon strongly emphasizes the grain in Kodak Gold 200 negatives
to the point where even a small amount of unsharp masking overwhelms
the image with graininess. GEM often, but not always, does an
excellent job of softening the grain to that the scans can be nicely
sharpened but increases processing time.
o You have to decide whether or not to use ICE and/or GEM BEFORE you
scan. You can't do a bunch of scans, examine them, then run ICE
and/or GEM on only the few scans you feel really need them. The
result is that you either have to commit ALL your scans to ICE and/or
GEM (at a huge increase of processing time!) or (potentially) scan
each image twice: once to see if you need ICE/GEM, and again to use
them. Also, each time you decide to set the ICE/GEM levels higher or
lower and compare the results you have to do a completely new scan.
o So far I haven't had a problem with scratches becoming scars when
scanning Kodachrome of Fujichrome slides.
o Scans of Tri-X negatives extremely grainy (but then this is a grainy film).
o If you use out of the box color management settings you will get
very nice color and tone from Kodachrome slides. However you don't
get the maximum dynamic range or shadow detail that the scanner is
capable of. So if you want to pull the most detail out of deep
shadows, and/or if you want excellent color and tone, you have to
turn color management off, do a "raw" scan (although these "raw"
scans have good color and tone, so obviously the data is NOT coming
straight off the CCD array), then apply a custom profile which you
previously made using an IT8 target and third-party profiling
software (I've been using IT8 Scan from profilecity.com).
o The stories you've read about poor depth of focus are true!!! I've
got some 20 year old Kodachromes that are perfectly clean and well
maintained but curved (which is normal for this film). The depth of
focus is so poor that the out of focus parts of the image (when
scanned at 4000 dpi) are blurrier than when scanned with my old 2400
dpi scanner. For some slides this is still true even if I set the
focus point of the scanner to be halfway between the curvature
extremes. One user has reported vast improvements by remounting all
his slides in Quickpoint mounts, and others have discussed using
glass mounts, but I have thousands of slides that I definitely DON'T
want to remount.
o In general I like the NikonScan 3.1 software on my Macintosh very
much, although the UI is very slow to respond whenever it has to draw
or update an image on the screen. In standalone mode it also is
extremely slow at saving files, loading files, refreshing image
windows when change which one is on top.
o The basic slide adapter is very nice, you just pop a slide into the
slot, and when you're done push a button and pop it out.
o The filmstrip holder is not so nice. It had four independently
moving parts, some plastic and some metal. Getting the filmstrip
inserted and the holder closed with all the moving parts aligned with
each other is a real pain, and it's impossible to do without getting
fingerprints somewhere on the film. Once the film is mounted you
have to slide the inner (plastic) frame back and forth within the
outer (metal) frame until the negative you want clicks into view in a
cutout in the metal frame, then you pop the holder into the same slot
on the scanner used for slides. This works pretty well once you've
gotten the film into the holder.
o It also comes with a motorized filmstrip adapter. The is a cool
idea in that you just feed the end of a filmstrip (of 2 to 6 frames)
into the slot and it grabs it and pulls it into the scanner. In
practice however it is cranky, inconvenient, and frustrating. (I'm
tired of writing so I won't go into detail.)
o This is NOT a production scanner. It takes so many trials to
decide how best to scan each individual image, and it takes so long
to actually perform each pass, that you couldn't possibly do
high-volume work. Think of each scan as being a hand crafted job and
you'll be in the right frame of mind.
o You need tons of RAM if you want decent speed. First you need RAM
just for doing the scan (120MB per full frame, 4,000 dpi, 48bits
max), then you need lots of RAM for the post-processing. After
getting the raw data NikonScan uses your computer's RAM and CPU to do
the post-processing required by color management, ICE, GEM and/or
ROC. I did one 120MB scan with ICE and GEM that took SEVEN HOURS to
complete. This was because I didn't have enough RAM and it was
accessing the hard disk continuously for those seven hours. Once I
got enough RAM the time went down to 20 minutes (?) or less. Also
note that Photoshop itself needs working RAM that is 3 times the size
of the image you're working on.
Well, that should be enough for now. Good luck,
At 10:27 AM +0100 1-11-01, Bernard COMOLET wrote:
> I am looking for any
>experiences, good or bad with 4000
> In a few days , I plan to
>buy one Nikon ls 4000 ; any comment
> There is the other brand,
>Canon FS4000, Polaroid, what is
> All the reply are welcome,
>thanks a lot.
>bonjour de ANGOULEME
Bill Fernandez * User Interface Architect * Bill Fernandez Design
(505) 346-3080 * firstname.lastname@example.org * http://billfernandez.com