Flatbed Scanners

I have been looking at the Epson 4490, and 4990 flat beds for 35mm slide and
negative scanning. Any Ideas on any others brands or results from the Epson
sacnners?

 

If you're stuck with a flat bed, who cares? They all suck for 35mm. I have
the 4990, and wouldn't use it for less than 4x5 (where it performs quite
nicely). You want to go cheap but good, check out the BenQ (Acer) 2720S
scanwit

 

I have the Epson V700 and like it a lot the dual lens system is very nice
and for once it comes with good Epson drivers.

R

 

JOOC,
Why do drum scanners work better than flatbeds? It seems like the more
moving parts the less stable.
Just out of curiosity

 

Also have a look at the Canon 8400F, it may meet your needs, I use one to
scan slides and it performs very nicely and has some quality software.

 

It depends on what resolution you're after, and what final quality you
need. I looked at film scanners myself earlier this year, and eventually
settled for the 4490 too. The reviews were very good.

What kind of resolution are you expecting? I actually did some test
scans of 110 images, and got much better quality images than the
original prints from some 25 years ago. It's certainly a lot better than
scanning from old and faded prints. You can get a 3400x2300 image from a
100 negative, which will give you a similar size printable image as a
6MP digital camera.

Of course you have to decide whether you *really* need the extra cost of
a dedicated film scanner, there seem to be very few affordable devices
about now. If you're just looking at digitising your collection of
personal photos for viewing on screen, then something like the 4490 will
do the job nicely at a reasonable cost. You can then take any individual
images to a professional printer if you really want a good print.

In any case, the quality of the source material may even affect the
results too. Negatives that have been laying around in their paper
wallets with the prints for years, may already be damaged and
deteriorated to a point that even the higher end scanners can't get a
good quality image.

Attention to detail with the condition of the negatives, and keeping
them dust free, along with getting the settings right, might be a better
start that throwing $$$'s at new equipment.

 

even a dozen years ago a professional high end scanner could attain
18000ppi, but don't try to put one in your bedroom (think eight feet in
length).

 

Well, there is a big difference.

http://www.pbase.com/davidjl/image/40078324/original
http://www.pbase.com/davidjl/image/40078325

The last I checked, the Nikons were still in production. But they're

Exactly, again.

David J. Littleboy
Tokyo, Japan

 

Yes there is, I wasn't arguing against that, just that everybody seems
to assume that another person has unlimited funds to buy the perfect
equipment (even if it's not what they really need).

Having a quick look at the 35mm scans I've done on the 4490, I'd say
they are somewhere between the Epson and the Nikon scans you have there.

Indeed. Approx 4-5 times more than the 4490.

Thanks. It seems to me that we often ask the wrong questions. With many
posts I see asking advice about what to buy, they hardly ever state the
purpose or the budget. Responses never ask those questions either, and
we end up with assumptions that the OP needs the highest quality scans
at whatever the cost.

 

But if you want to scan film, using a flatbed is just a kludge in terms of
image quality. B&H lists at least 4 film scanners that cost less than the
4990, and though I've never used them, I bet they work better than any
affordable flatbed. I used for many years the Acer Scanwit 2720, and then I
switched to Nikon, and the $300 Acer wasn't far behind the Nikon 8000ED in
terms of image quality. I'm not assuming anything about the OP's needs, let
alone "highest image quality", I just know I wouldn't use a flatbed for film
smaller than 4x5--it's a waste of time.

 

[Snipped Text]

In your opinion perhaps, but I've scanned a few hundred images from 35mm
negs on my 4490, and it's produced quite acceptable results. The quality
of the images is ideal for viewing and manipulating in iPhoto, and
viewing on screen, either locally, or on my web site. I also send these
out on CDs to the family, and for most of them they are acceptable too.
If I really wanted a full size reprint of any of my images I'd just send
that one image off for specialist printing.

As it is I've already scanned a test image from a very old 35mm image
(on the 4490), printed it to 10x8 (on a Canon IP4200) and framed it, and
it looks bloody fantastic.

Here in the UK, film scanners *start* at about £450 at street price, I
paid £160 from my Epson 4490. For that I got a very good flat bed
scanner that is as good as any flat bed, and a film scanner that also
produces good enough images. Another thing I can do is scan unusual
sizes of negatives (such as 110 and 126 formats).

 

Flatbed scanners are useless for negative and slide scanning. Go for a
Nikon or Minolta negative/slide scanner, on Ebay, if you want a bargain.

 

[Snipped Text]

No, the 4490 is a new high end pro-sumer model. It's about halfway
between a really cheap scanner and the 4990.

Indeed, but I paid only £160 (which would probably end up being the same
in dollars anyway) for my 4490. This is a lower model than the V700,
being less than half the price, but is higher than the 3490 at the low
end of their scale.

 

IMO, owning a 4990, an LS-8000, and having used quite
a few of the film scanners of yesteryear -- that the 4990 is
roughly on par with the 2700 dpi film scanners of about
five years ago, eg., the Polaroid SprintScan 35+ and the
Nikon LS-2000.

The tests I've seen, and my own gut feel, tell me it's
delivering about 2000 dpi. It's clearly soft compared
to a couple of very expensive (and old) 2500 dpi film
scanners, eg. LeafScan 45 and Microtek 2500.

The saving grace of the Epson(s) are that they are
fairly low noise and somewhat oversampled, so they
take fairly well to sharpening (unsharp-masking.)


rafe b
www.terrapinphoto.com

 

Roger, you're about the only one I know who's claiming
a "major improvement" from the 4990 to the new V700/750.

None of the reviewers are making that claim, and the
tests I've done don't bear it out.

Do you have any side-by-side scan samples you can post?

I mean, it wasn't all that long ago that you claimed that
the 4990 matched drum scans.... so the V700 must beat
drum scans, by your reckoning. I'd love to believe that
but I don't.

The tests I've done show the V700 still well behind a
five-year old LS-8000.


rafe b
www.terrapinphoto.com

 

A modern film scanner would make a better comparison. The university paper
I worked for 12 years ago had the LS-2000, as I recall.

 

Here is what turned my attention to the flat bed scanner.
I do not need very high quality but the film scanners do seem better but
it looks like the flat bed are coming up.
I just wonder if the lack of sharpness is from the film not
placed in just the right focus location.

http://www.photo-i.co.uk/Reviews/interactive/Epson 4990/Page 8.htm

 

(This is getting a tad long, so I'm going to snip. A lot.)

I've always found it very hard to make any sense of the examples that appear
in that site's scanner reviews; they always seem to show almost no
difference between the flatbeds and the real film scanners, yet in my
experience, there's a world of difference.

I find your result odd. Maybe the source image didn't have the sharpness,
maybe the drum scan was done badly. But, as I've hinted before, I find
pretty much all your scanning results strange and hard to see what's going
on.

Ah, you mean edge response isn't MTF? I thought that graph was rather
non-representative of reality under the assumption of it being MTF.

As you know, the Epson scanners all use a "displaced" CCD; for the 4800 ppi
scanners, that's two 2400 ppi CCDs displaced by 1/2 a pixel pitch. The
"scanning aperture" at the pixel is thus twice the pixel pitch. So if you
scan at the advertised resolution, noise reduce aggressively, and downsample
to, say, 2/3 of the "real native resolution", you can expect to get images
sharp at the pixel level. (The Nikon 8000 produces images sharp at the pixel
level only by downsampling to 2400 ppi or so, and, as I've show, it's _way_
better at the pixel level than the 4800 ppi Epsons.)

As before:

The 8000 and the 4000 are same generation technologies, and should be quite
similar.

David J. Littleboy
Tokyo, Japan

 

Ah yes, the ever-optimistic Vincent.

Trust me, Roger. I'd love to believe the V750 is worth the bother.

I coordinated an LS-8K versus V700 test a few months ago with a fellow
named Alan Bridgewater. Maybe I can get Alan to send you the same
negative we each scanned. It's on my snippets page.

www.terrapinphoto.com/jmdavis

....about 2/3 of the way down. No V750 samples submitted yet. Would
you like to volunteer? I'd be honored. Give it your best. The
sharpest image you've got (it need not be interesting, just SHARP)
scanned at 6400 dpi -> hence 1600 x 1600 pixels.

Save as lowest-compression JPG and send it my way.

rafe b

 

Those aren't my Epson scans.

Yep. Of course, holding medium format, let alone large format, film flat to
within +/- 0.02 inches is physically impossible without a glass carrier. And
scanning MF film at 6400 ppi and 16 bits, noise reducing, and downsampling
to something somewhat below 3200 ppi requires more memory and computational
capabilities than most of us have.

The Nikon 8000 is _almost_ as good as I need; it ought to produce 2700 ppi
images that print nicely at 300 ppi, but the whole system is a tad iffy at
that point. Some of that is the film; for a 300 ppi image, one wants strong
contrast at 4 lp/mm, and that's 36 lp/mm from the film. The contrast at 36
lp/mm on Provia simply isn't going to be as nice as the contrast at 12.5
lp/mm on Velvia (I'm referring to your f/64 shot; at f/64 MTF50% should be
about 12.5 lp/mm, and Velvia has MTF gain at that frequency). Between camera
lens, film, scanner lens, and scanner CCD, there are a lot of MTF terms
between the subject and the file. Sigh.

So the bottom line here is that 6x7 produces overkilled killer 13x19s, but
it's a strain from 645. So I just shoot the 5D.

Uh, I meant downsample to 2/3 of "1/2 the advertised pixel resolution".

Will do. With your parenthetical note that ERF is higher than MTF, this is
beginning to make some sense; I did find with the 2450 that it would make a
nice sharp image of a sharp edge, but that all it produced from actual film
was mush.

David J. Littleboy
Tokyo, Japan