followup on the Nikon scanners - I just spent a weekend testing a friend's 4000 ED, which is very much like the V model with the ability to use a stack feeder. From what I gather online, the 5000 ED is about twice as fast, but otherwise very similar.
Well, I'm glad I got this loaner to see for myself, because I am suddenly not so sure about getting a Nikon scanner:
reasons:
a) Nikon software awkward compared to Polaroid Insight. Vuescan is better, but had the same problems with asquisition I had in the Nikon software.
b) image quality: 4000 dpi makes no difference in image detail. All you get is more grain, even on Velvia. In turn you need to run the slow GEM process to cut down the grain. Cropping at full resolution for detail information shows that the 2700dpi Polaroid retains just as much information as the Nikon, at half the file size.
c) highlight bleed on the Nikon - any high contrast area will bleed a halo of light into the dark areas. I cleaned the mirror of the Nikon (difficult procedure, search online for instructions) to make sure it wasn't dust on the mirror as some folks post online. It is not. Completely unacceptable image quality, and extremely difficult to fix in Photoshop.
d) speed - my Polaroid gets a scan done in less than 40 seconds at full rez, which is more than twice as fast as the Nikon even without any of the multisampling options (which made no difference in addressing the problems I had with light bleed)
c) occasional complete failure to scan anything like the original - the first few slides I scanned came out with zero red channel information. This went away after inserting the neg strip holder, but came back after power up. Could not get rid of it the second time and eventually packed the scanner in the box and put my Polaroid back on the desk.
The good stuff:
when the 4000ED actually worked, the colors of the sky especially seemed to come out better than on the Polaroid. The shadow detail is slightly better, but that only works well on slides that have no major highlights at the same time, since those will cast a halo over the entire shadow area.
ICE is nice - dust and scratch removal clearly work very well with this enabled. It's just that you are trading the cleanup time in photoshop for images that just don't look right, or in some cases are completely useless due to massive light cast across shadow areas.
was the 4000 ED that I tested broken? Again, I see a lot of people posting about the light bleed issue and it seems to be something all Nikon Scanners show. I recall seeing it on some older models as well. Focus issues seem to be another story, while the Sprintscan doesn't have that problem due to a very different imaging technology.
Bottom line for me - since I have a Sprintscan 35 Plus that works well for me, I am not going to risk $1000 for a 5000 ED to realize that it has the same problems I see with the 4000 ED loaner.
Some examples:
Nikon 4000 ED
http://didnt.doit.wisc.edu/outdoor/gall ... e_4000.jpg
Sprintscan 35 Plus
http://didnt.doit.wisc.edu/outdoor/gall ... lisade.jpg
Sky color more natural in the Nikon scan, but the halo cast of light over the entire foreground destroys the contrast there.
Low light performance
Nikon 4000 ED
http://didnt.doit.wisc.edu/outdoor/gall ... g_4000.jpg
Polaroid Sprintscan 35 Plus
http://didnt.doit.wisc.edu/outdoor/gall ... fading.jpg
The color bleed on the Nikon is quite pronounced - looking at the original slide, the dark foreground in the Polaroid scan is much closer to the original, although it loses quite some detail in the darks, too. The Nikon scan may look appealing and warm, but that's not what is on the original slide.
I was going to do more comparison scans, but the software balked again and I was not getting usable scans from the Nikon and decided to go back to what I know.
and here two crops at 1:1 resolution pulled off each scanner at thei max dpi to compare what they can pull out of the original:
http://didnt.doit.wisc.edu/outdoor/niko ... amples.jpg
the sky seems nicer in the Nikon scan, but that's been run through the GEM thing on lowest level to get the grain down. Color of the sky is also a bit closer to the slide on the Nikon scan, however, there really is no additional feature information in the scan that I cannot see in the Polaroid scan as well.
Still, if I had no scanner at all, I probably would give a 5000 ED a chance, since you can easily resell them on ebay. The Polaroid is difficult to make work these days, (SCSI, pre-XP drivers), so I can't recommend that for folks who don't feel like tinkering with software issues. If you have dirty slides and don't know how or don't want to clean up scans in Photoshop, the ICE feature on he Nikons may be a deciding factor.