Canon 4000 DPI Scanner

Where did you find this 4000DPI Canon scanner. On the US site they have a new scanner listed but it does 2,780DPI

It's on the European Canon site: (link)

It's also listed on B&H for just under $1,000. I'd also be interested in anyone's experience, esp. with B&W negs.

Canon USA's web site has information on this 4000 dpi scanner which was announced on Feb. 11, 2001 to be available second quarter 2001:

http://www.powershot.com/powershot2/fs4000/index.html

Does anyone know if it is out yet or is it vaporware?

Canon lists the 2400uf as "available now" and many online retailers are selling it (though it is unclear if it is in stock) Has anyone used one yet? How does it compare to the Epson 1640 or 1680?

Information about the FS4000US film scanner can also be found at Canon's USA website:

http://www.canon.com.sg/scanner/fs4000us/scanner_fs4000us_prod_info.html

The only negative as far as I can see is the Dmax is only 3.4. Personally, I've come to put a premium on scanners with high Dmax as I scan alot of E-6. Obviously, if you're only planning to scan C-41 it is probably an excellent choice.

You may place pre-orders for this scanner at www.cdw.com on June 1st. I have not tried one (as it is not shipping until June/July) but initial specs look good. If you order from CDW you can get a better deal than advertised price, as the account managers work on cost + %, just ask them for their best price.

Has anyone seen a scanner test that actually measured Dmax?

Does anyone know of an ANSI standard for the measurement of dynamic range of a scanner device?

Saw a demonstration of the new Canon Scanner at a photo exhibition in Singapore over the weekend. Also saw a lot of other stuff like Coolscan IV, Coolscan 4000, Coolscan 8000, D1x, D1h, Coolpix 995, Minolta Dimage 7, Contax N1 and N Digital, new Nikon super-telephotos, new inkjets from Canon and Epson, but I digress....

canon's offering is targetted towards a more consumer audience with most of the functions being easy to use. Example is FARE (similiar to Digital ICE3) which can only be turned on or off - no additional settings or tweaking available. Compare with Nikon's implementation where you can turn individual elements of ICE with variable settings for each. Nikon Scan 3 software interface, by contrast, allows you to change almost every aspect of the scanning process, with a lot of tweaks available to the user.

Performance was a bit slow - on a USB interface, scanning at 4000dpi took more than 5 minutes (altho the demo person explained that the PC was "underpowered")

Details were very sharp, thanks to the 4000dpi resolution. Could not really judge color fidelity as monitor was not calibrated.

Canon has not yet made this scanner available, but indicated that the price is around US$1000, which makes it cheaper than the Coolscan 4000. However, bearing in mind that the Coolscan 4000 is a higher end scanner (with 4.2 DMax, Multisample scanning, FireWire interface, batch/roll scan options), I think the Canon scanner should be more evenly matched by the Coolscan IV. By the way, I just purchased a Coolscan IV at the princely sum of US$655.

regards, Darren

A dynamic range of 4.2D is impossible in a single pass scan with CCD arrays. ALL CCDs top out around a 5000:1 brightness range, or 3.7D.
Greater dynamic ranges can only be acheived by multipass or multisample scanning, which combines two scans at different exposure settings, and there's no reason why this feature couldn't be incorporated in the new Canon filmscanner.
In fact, Canon have not yet released any figures for Dmax or dynamic range, and I'd be very surprised if they didn't take advantage of the multisample 'trick' to match Nikons claimed dynamic range. (Ed Hamrick's Vuescan already uses a multi exposure technique to get an extended dynamic range with some existing 'old' 12 bit scanners. See what Ed has to say about Dmax and dynamic range HERE)

However, bearing in mind that the Coolscan 4000 is a higher end scanner (with 4.2 DMax,

Keep in mind that there is no commonly-agreed-upon standard for measuring dmax, which of course invites manufacturers to measure it with any methodology they see fit. ("Hey, Aunt Martha, before you go to the store, can you tell me if you see anything when I shine my keyring light through this negative?") This is not to say that the Nikon doesn't have a higher dmax than the Canon, only that there's no reason to assume it does.

Armed with a densitometer you can come up with your own dmax but there's no reason to assume it has anything to do with what the manufacturer claims. Since almost no formal reviewers bother to actually test manufacturers claims WRT dmax, some individuals have taken to performing such tests on their own. The results are not encouraging. See, for instance:
http://mejac.palo-alto.ca.us/leica-users/v10/msg12058.html

If anyone is aware of major reviewers who do test dmax routinely, could they please identify them here?

I've run tests on my own scanner using pieces of ND filter gel. They're available from Lee filters in densities of 0.15, 0.3, 0.6, 0.9 and 1.2D. Stacking only 3 pieces of 1.2D gives you an accurate 3.6D, and this is usually enough to reach the true limit of dynamic range on most CCD desktop scanners.<br>Remember that dynamic range is Dmax minus Dmin, and Dmin should be determined also. A lot of scanners don't seem to register densities less than 0.15D, and transparency film has a Dmin closer to 0.1D<br>The issue is further muddled by the residual infrared sensitivity of the CCD modules used in some scanners, and the fact that most scanner light sources (LEDs included) put out a fair bit of IR. This can result in a slight IR 'fog' that masks the true Dmax of the scanner, and gives different results with different film materials. Kodachrome, and silver based B&W films might give different results from E6 films and ND filter gels, for example. <p>I suggest the following methodology for testing:<br>A 'step wedge' or filter mosaic is made up in a (glassless) slide frame, using known density filter material; going from 0.15D down to at least 3.75D in 0.15D steps. The frame should also include a totally clear and a totally opaque area, possibly using aluminium foil to occlude even IR.<br>If no difference is detected between the clear area and 0.15D, the conclusion must be drawn that the white point is set too low, and that the scanner may not scrape all the highlight detail off a slide.<br>If any discrepency is found between the truly opaque area and the maximum filter density where a change in level can be detected, then IR fogging should be suspected as the culprit.<br>I'm not saying this test is perfect, but it's better than sticking a wet finger into the wind and guessing.

Mr. Andrews -- What KIND of scanner?

I purchased the Canoscan FS4000 from Adorama. I was disappointed with the scanner and just packaged it up yesterday and returned it. The scanned images always had a lot of noise, particularly when scanning negatives. Results from slides seemed to be come out better. I compared images I scanned on the FS4000 to the same images scanned by Kodak Photo CD and found the Kodak Photo CD much better, even though the Kodak Photo CD scans where at a lower resolution then the FS4000. Yes, it produces a large file with the high resolution but with high noise level.

Today I visited a local store that had an FS-4000 and Nikon LS-4000 attached to the same machine -- USB and Firewire, making the FS-4000 comparatively slow.

On the upside, the FS-4000 had more accurate color, and the slide scan seemed sharper or more in focus. On the downside, there were little black dots all over the place. They looked too small to be dust particles, not patterned enough to be electrical noise, and in the wrong places to be sharpening artifacts. FARE was on.

On the upside, the LS-4000 did not have black dots or much noise at all. On the downside, detailed areas seemed either out of focus or a bit blurred by ICE. The scan was too yellow, although that would be easy to correct. Worst, there was a section with jaggies that might have been caused by a faulty stepper motor.

Bottom line: neither seems ready for prime time. The Nikon doesn't really seem worth $700 extra. If Canon could solve the black dot problem, the FS-4000 would be my choice for scanning slides. Unfortunately I shoot mostly print film, and didn't get a chance to scan any negatives.

I purchased the CanoScan D2400uf a few weeks ago directly from Canon. I have been looking for a modestly priced scanner that would give me acceptable images from both 35mm and 120 transparencies. Before this scanner, that meant $1500 or more. But I am thrilled with the scans I am getting from the D2400uf. Keep in mind, I am only looking for monitor scans, to load unto the web and to put on CDs, not printable images. But the ones I have scanned have printed very well. (Haven’t tried any negs yet). I have experimented with many flat bed scanners, and have had ok luck with my 67 trans, but never have I been able to get a quality scan from a 35mm slide until now.

I have just registered a new domain name that I will be transferring my web site to next month, but in the meantime I have loaded a few images scanned with the D2400uf there if you would like to see them. www.davidrdaniel.com.

dave

Ed Hamrick posted in the comp.periphs.scanner newsgroup that Canon will be sending him a loaner FS-4000 soon so he can implement Vuescan on it. That should make it easier for folks to compare it directly with Nikon LS-4000 hardware.

I've put my first impressions of the FS4000US on a web page, http://www.normankoren.com/Tutorials/Canoscan4000.html. Summary: It's extremely sharp (sharper than you could get with a conventional darkroom print), has good color, but has a few annoying software quirks. I can live with them, but I hope Canon or Hamrick fixes them soon.

I just bought this scanner, and have none of the problems with the "black dots" I've read much about here and elsewhere; Canon must have fixed it in recent batches. When using FilmGet (the included software), I'm getting extremely sharp, accurate scans -- I'm very impressed. I'm having a harder time getting satisfying scans from Vuescan, which I need to play around with more to make sure it's not user error.

Bottom line so far is: I love it. I just ordered an Epson printer so I can make 13" x 19" prints. I can't wait for it to arrive.

I got this scanner early August, and I haven't got the black dot problem either. It is very sharp and I love it. :)

One problem I had had was the red hue. I checked several review sites and it seems FilmGet gives the red hue. After playing with photoshop several times without the satisfaction, I finally got FilmGet's adjustment fix the red hue problem. Now, my every scanned image looks soooo cool. :)

I haven't had a chance to scan any slide, though. I am taking my first slide film right now. I gotta go out and finish this roll soon.

Anyway, this scanner is pretty darn good so far.

I'm also having focus trouble with vuescan on the fs4000. Even single pass scans give an elongated grain effect resulting in a blurry image that is no where near as sharp as the canon software.

Ed Hamrick just fixed the FS-4000 pixel smearing problem (where each pixel seems blurred up and down 1) in version 7.1.11 of Vuescan.

I´ve just sold the FS 4000 again.

I havent had a scanner yet that blocked the shadows as much as this one. And it´s just too slow at 4000 dpi.

There is no difference to 2700-2900 dpi scans in the final prints, only if i had used sharp single focal lens / tripod / Velvia, the scan is somewhat better. The files are huge (100 MB+ at full color depth).

I just didn´t have enough slides of the necessary quality to justify the drawbacks. I don´t think 4000 dpi is really such a hot thing, because 35 mm just isn´t, i get much better results using Pentax 645 and the cheap Epson 1640 SU. 35mm is dead anyway, results up to 8x11 inch are better from the Canon D30, at 400 ISO even much better.

My following comment comes after several months of experience with a Nikon 4000 ED scanning mainly E100S slides.

Nikon scanner optics have a very shallow depth of field, so autofocus accuracy is very critical for them. Since some slides and negatives have large areas with no image details, Nikon scanners have to use the emulsion grain to focus on! So Nikon scanners use collimated light to make the grain pop! Then they include GEM software to make you feel like you are really sophisticated and really got your money's worth. And while GEM is running, you can't use the scanner. Also, the CCD sensor in this scanner has signal bleed and many high contast slides are not worth scanning with it, IMHO.