Microtek Artixscan 120TF Scanner vs. Nikon Coolscan 9000

Don't have experience with the Nikon, but my 120tf is a fantastic scanner. Using Silverfast software, a 6x7cm trans scans in about 8 minutes at 4000dpi, from the time I feed the carrier into the scanner until the completed scan displays on my monitor. Lower rez scans take significantly less time, but if it's an initial archive scan at full rez, that's what you're looking at. I know some newer scanners have Digital Ice and so on, so that might be worth checking into, since it's not on the 120tf.

You're better off getting a used Imacon Photo Flextight. I've seen them sell on EBAY in the $2,000 - $2,500 range. The key to CCD scanners is the film holders. I used the Polaroid predecessor to the Artixscan for a week and then returned it before buying my Imacon, 3 years ago. The Polaroid/Artixcan was good, but the film holder, while decently designed, created flatness problems in some cases, and overlapped some of the image. The Nikon is reputed to have awful medium format holders, unless you buy the glass holder. The problem with a glass holder on the NIkon is (1) it creates 4 more surfaces for collecting dust and (2) you can't use digital ICE with glass holders. Dealing with dust is a big problem with CCD scans at high resolutions. You'll spend a lot of time dusting scans bigger than 2000 dpi, and a decent amount of time dusting scans at around 1600-2000 dpi. It helps if you do "virgin" scans; scanning the film immediately after processing. I don't care how well you've cared for your film; it collects dust quickly. Dust is also affected by the quality of the light used. The Imacon, while imperfect, is fantastic. It's designed around the film holder which flexes creating a virtual drum. It means the distance from lens to film plane is precise and doesn't vary. You get very sharp images. No glass used. Dusting is an issue, but much less than with the other medium format scanners. I'd of course prefer the newest Imacon 343, but that would cost you about $5,000. Also, the IMacon software is great, and easy to use, and constantly being updated and refined. Imacon support is also excellent. They are very responsive and actually solve your problems.

I second the choice of Imacon. I got a used Flextight Photo about a month ago and the scans are fantastic. No glass film holders required. There software is also very good.

I also have the Microtek 120tf since about a year now. I preferred it over the Nikon 8000 which had banding issues. If those have been solved with the 9000, I dont know. The 120tf also comes with Silverfast Ai6/IT8 (the Nikon does not), wich is in my opinion the best scanner softwere around. It therefore makes the 120tf the most cost effectiv MF scanner available.

Scans are perfect and are very sharp, easy to tweak in Photoshop, or in my case PWP. I have read somewhere a test, which measured the highest dynamical range, and it shows in very good shadow detail. I scan 35mm, Panoramas 24x66 and MF 6x9. For the 24x66 a glass holder is a must. But the optional glass holder has its problems. The dust issue has been addressed before, another problem are newton rings. In case of the 120tf glass holder there are two issues with the glass quality itself. First, the top glass plate has a matt (it is more structured than matt) bottom surface in order to avoid newton rings. This surface shows up in 4000dpi scans. Blown up to 100% in Photoshop, the surface appears as noise or grain. Second, the glass has a slight tint, which also is noticable in the blow ups. I worked my way around this problem by ordering a second 35mm filmstrip holder, cut out the sprokets, and cut some additional gaps in order for the scanner to recognise the film size (the same gaps as on the glass less MF holder. It works perfect, no more structural and tinted interferences in even colored areas like skys.

As to the glassless MF holder, if your film is perfectly flat, then there is no sharpness problem and Dof takes care of it (how to flatten film: lay it flat on a light tabel for a couple of hours with a book on top). Again in my case as I only scan 6x9, I added sprocets at the 9cm mark, just in case, with perfect results.

Digital ICE, no, the 120tf does not do it. But when you scan with glass less holders and have your film well dusted, you dont need it. It is faster to use the clone tool.

Jo,

I'm very interested in the Microtek 120tf and there aren't too many people out there that own it to report on it. I'd love to see an in depth review of the machine.

I'm most interested in finding out how to avoid Newton Rings, since I'll be scanning 6x6 and 6x7 color slide and B&W films, as well as the mysterious color registration issues (where the R, G and B layers are slightly shifted by a few pixels).

I would rather get the Microtek than have to shell out an additional $1000 for a comprable machine (additional software and glass holder) with the Nikon 9000.

Thanks Brett

If you are scanning medium format, you might consider the Epson 4870.

I should add that the reason I suggest this is that I have the Microtek 4000tf, which is the 35mm model, and I find the difference in scanning between the two (Microtek 4000tf and the Epson 4870) to be dramatic. I originally got the Epson only for medium and large format scanning, but I now use it for 35mm as well, especially when scanning transparencies as the color rendering is much better right from the start. Means less time spent in Photoshop. I have a number of b&w negatives that look like crap with the Microtek and beautiful with the Epson. Why, I know not. Maybe the Epson is a better liar. I do know that both the 4000tf and the 120tf have been around a long time and were originally sold as Polaroid scanners, recently upgraded to Firewire, so perhaps new in this case is actually better.

OK, I'm back with some fresh research (to me at least) and I thought I'd share. You see, for months I have convinced myself that the Microtek 120tf was the way to go. I was basing this on the (good) reputation of the Polaroid scanner that Microtek once made and the rather wishy-washy yet glowingly positive reviews it got from users that never seemed to show their work (suspicious).

So along come these threads, and as more people use the Nikon 9000 and the Microtek 120tf it seems that they are both equally capable scanners but two things still stand out for me, 1) this color shifting thing in the 120tf, and 2) which one REALLY does provide the sharpest image?

Here's what I found and maybe this is old info, but I found this link:

http://www.jamesphotography.ca/scanner_test.html

And it's very interesting when you break down the results. It appears that the Nikon 8000 resolves in the range of 26 MTF or so and the Microtek is down near 18 MTF - not a huge difference but a difference nonetheless. It also appears that you really need to get Silverfast Ai and a glass film holder to get the best out of any of these scanners.

Having said that and using B&H as a pricing example, the Microtek with glass holder and Silverfast (bundled) goes for $1930 USD (not counting the rebate), you also get a IT8 target which is nice, but you could obtain one through other means if you had to.

The Nikon, also at B&H (if it was in stock) with the glass holder, and Silverfast rings up to $2740 USD, about $800 more, but with it I think you may get better images. When shelling out close to $2000 already for a scanner, what's $800 more is my thinking? Well, maybe 8 more MTF!

By the way, Jorn, nice gord, but flatbeds are in a league of their own, if you know what I mean.

Happy shopping everybody! This is why I love photo.net so much - I think I may be a Nikon convert! Who knows, in 3 months time I may waffle back to Microtek. Someone prove me wrong.

BK

Not sure I understand what league of their own means about flatbeds, but Silverfast is frequently buggy and inconsistent. I have always had to imagine that stable versions were all on Mac machines. I finally got a stable version on my PC at version 6.2.Or4 and so stopped updating it. Now it works fine, but who knows?

Howard - your claim that you can't use ICE with glass holders is totaly false. With a glass holder (either DIY or seperatley purchased) the 9000 provides corner to corner sharpness with perfect colour channel registration that is the equal of the Imacon ( yes I have tested them ) with the added benefit of very, very effective removal of dust.

Brett: Newton Rings appear when you put film on an absolute flat surface i.e.glass, it becomes worse with high humidity. To overcome this the continuous flat surfaces have to be broken by putting microscopical grooves or somthing like this into the glass surface. But those grooves worke like little lenses and are transmitting the scanners lights in different intensities through the film to be scanned. There is no solution to it except absolut dryness.

As to reviews, well, those giants such as Nikon spend e lot of money on marketing. These people know very well how to make you want a particular product, and most of the press is not impartial, they relay on huge advertising accounts. Therfore you only find favorable reviews of the Nikons, Canons, Minoltas and Epsons. Go to Computer Darkroom, there you find a review of the 120tf. I did my homework and tested all 3 available MF scanners myself, same slide, same settings, etc. Thats the most objective proof you can do.

Jorn: The 4000 is an outdated scanner and has been on the market before the Polaroid 120tf. When Microtek took over the marketing of the 120tf they refined and improoved it and therefore it is realy up to date. The Epson 4870 is a different kind of scanner. Flatbed scanners have a very low resolution, mainly because of the glass sheet. Test results rate the Epson with a real resolution of 1700 to 1800ppi when set to 4000ppi. For MF it is ok if your printout is no larger than 10x12, the other disantvantage is the high storage size (it counts for 4000ppi, for the actual resolution of 1700ppi).

Jo,

Thanks for taking the time to post - your responses are well thought out and I (we) value your opinion, so thanks!

Since you have the Microtek, do you see any of the color misregistration issues that others have talked about? Where an inconsistent misalignment of the R, G and B layers don't match up (making for a less than optimal scan)? Some have reported it to be 3-5 pixels.

BK

I just posted this exact question a few weeks back. After much deliberation I went for the *Minolta Multi Pro*. Here is why.

It appears, that for the Nikon scanner the glass holder is a must for MF scanning. This means, not only the extra cost, but also as has been mentioned you have to put up with four extra surfaces. Now I do use the glass holder with curled films and would like to have the option available once in a while, but like ICE I wouldn't want it to be compulsory. Also, if it is anything like its predecessor (which I have used) then one has to often resort to "super fine" scanning to avoid banding which makes the scan times explode.

Microtek doesn't come with the glass holder either but the native holder seems to be better designed. But I would like to have the option as already mentioned. Moreover, it has no ICE. Again, I do not use ICE all the time, but when I need it, I need it. Spotting a 750MB or 1GB+ scan is no fun.

I had originally rejected the Minolta because of is poor reputation for scanning negatives. However, it appeared that the new version of the software has helped considerably. In addition, one must consider that people have harder time scanning negatives on any scanner since negatives do not give good colors off the bat but needs some fiddling. I accept this as a fact of life, and it is no different in the Nikons. Additionally, I had heard of the Firewire interface not working on Windows.

I decided to go with this one and here is the result after a week.

I think it is the best compromise. I have had no problems scanning negatives, or for that matter any other formats (but see below). The best feature of this scanner is its lack of noise and *extremely high dynamic range* (compared to CCD scanners of course). This appears to have the best dynamic range I have seen on any CCD scanner. It is definitely better than the Nikon 4000 that I have been using. The installation was a breeze, so far no problem with the Adaptec Firewire card.

What I don't like with this scanner is that the optics exaggerate grain (or some artifact that appears as grain when viewd at 100%). I am hoping that the Scanhancer will reduce that (most user reports that it does). In any case, if you sample down from a high resolution scan, the grain is not that discernible on smaller prints. Further, the auto focus does not work. You have to manually focus each scan. A pain.

Overall, all the scanners seem capable, and you will not be outright wrong with any of them. Just my two cents on the Minolta.

Brett

I have not noticed any misalignment as I never check my scans byond 100%. I have just looked up a few tif files and did not find anything, no banding either. Having said that, one must put into account that on and off a misaligned CCD can slip through to the production line, can happen to any manufacturer.

To Victor's comments: The Minolta was one of the two scanners I had in mind to buy (the other was the Nikon 8000), there was a lot of advertisment about them and I did not know anything about the Microtek. Although the Minolta produced very good visual results, not much different than the other two scanners, I decided against it because for 35mm film it has a resolution of 4800ppi, but for MF only 3200ppi and that's because there is a mechanical/optical realignment for the MF format (that might effect the auto focussing problem). Secondly, as said, the scanning software was lousy and Silverfast was not available (to which I was acustomed). Another issue was the so-called optional Scanhencer, a clever little device (a diffusor plate) to eleminate so-called peppergrain in high speed negative films. I dont need it for slow speed slide material, but my tests showed a bit of image softening of the scan (now proven by James Photography: it has an MTF of 19.65 without and 18.63 with the diffusor, a pixel rise of 4.21 without and 4.53 with (the 120tf: 18.4 and 3.39). The dynamical range, according to an independant optical test lab in Germany, the Minolta has 3.22D for color and 2.8D for b/w. The 120tf has 3.67D for color and 2.87 for b/w (the 9000 not available yet, but the 5000 has 3.08/2.28D, that confirms that nothing has changed since the 4000 in terms of dynamical range).

Hope that clears few questions and does not confuse you even more.

Jo

Jo,

Took me a while to get back here, but I'm back because I'm reconsidering (AGAIN!). I'm actually hoping the Microtek is all its cracked up to be instead of dumping an additional $1100 on the Nikon (factor in software, glass holder ect).

Why second thoughts? I forgot about Monitor/Printer calibration! Silly me. So I'm looking at the Gretag Macbeth Eye-One Photo which is a mere $1400 at B&H (ha! I thought this was going to be cheap!) So the Microtek needs the final test - an actual scan test. I'm actively looking for a local retailer to give me a test drive.

Thanks for the insight though, Jo!

BK