High-end scanners

[Bruce Watson]

Frank, Of course, abolutely, you have to learn the scanner and the software to get the most out of it. You'll also over time adjust your film exposure and processing to take more advantage of the scanner. And yes, you can probably make some machines outperform others if you make the effort. I don't doubt it for a minute. I wasn't trying to downplay operator experience.

The question is, how good is good enough? The individual photographer has to decide. Since you can't gain a year's experience (which wouldn't be enough) with every machine you might be intererested in, you have to find another way to make a decision.

It's similar to how you buy a view camera. You research the market, talk to people, ask questions, look at your bank balance. You try to get your hands on a couple and try them out. Then, when you buy, you try to get the biggest bang for your buck.

All I'm telling QT really is, drum scanners aren't that hard to use, fluid mounting is worth the effort and not that hard to do, and the bang-for-the-buck quotent is very high since drum scanners are suddenly so inexpensive.

I'm also telling him and anyone else that wants to know that it might be wise to find out if the quality level works for what you want to do. For you, drum scans aren't worth it. For me they are (but not at the level of a Crosfield 6250, or a six ton Hell 3400). For what we do, we are probably both right.

[Henry Ambrose]

Lots of good advice given. I wonder what end product you have in mind? Print reproduction for magazines? Photographic prints up to 16X20? Murals?

In the mid 90s I tested the Polaroid and Nikon 2700 ppi 35mm film scanners against a Howtek 7500. I was pretty sure the Howtek would smear the little desktop machines. What I discovered was that up to about 600 percent enlargement I could not tell any difference! Above 600 percent the Howtek won with the difference growing as reproduction size grew. Those early desktop film scanners are pretty poor compared to today's machines but made very good scans within their meager limits.

Hogarth's last post contains the very best advice - you really should test to determine just how "good" a scanner you want/need. I suggest you start with the end product you desire and work backwards to find what machine is needed to make that happen.

Resolution will probably take your first attention but don't overlook speed and ease of use, they are also major factors. After you do a few dozen scans and get over the "new toy factor" you'll hate sitting there waiting for the scan to finish or wrestling with a film holder or mounting system that requires three hands to make work.

There is something about looking at my latest photograph on screen at high magnification that I still find THRILLING! There are not many things in photo world as juicy as a big fat sharp 250MB scan of a piece of large format film. As wonderful as that is I want it to happen as easy and fast as possible.

[Martin Miller]

Microtek just announced a new flatbed scanner, the Scanmaker i900. It is a new design with a drawer for glassless film scanning up to 8x10. The main specs are: 48-bit color, 4.2 DMAX, and 6400x3200 ppi optical resolution. Silverfast will be bundled with the pro version. Price will be "sub-$1000". See http://www.microtekusa.com/images/broi900.pdf

I suspect that Epson won't be far behind replacing their Expression 1680 with something similar. The Epson online store has been out of the 1680 for some time now. I can think of only one reason for that at this late stage in its marketing. This new generation of flatbeds may be worth a look. - Martin (www.millervisuals.com)

[Jonathan Brewer]

What about the Fuji flatbed scanners that the Flaar folks are raving about, are they good enough to get you in the ballpark of what you can get from the drums and top flatbeds of other manufacturers?

[John Hennessy]

This may be a question the answer to which everyone but me already knows and therefore it has not been asked. But if PMT scanners are inherently better than CCD scanners, why are none still made? Of course there is huge market for the cheap scanners but there seems to be a good market for more expensive digital equipment. QT, as an example, is ready to spend 20 times more than a Epson 3200 costs.

Another way to put my question is: what is the relationship between PMT and a drum and CCD's and dedicated film or flat-bed scanners? If PMT's are always better and if there unbreakable relationship between PMT and a drum platform, couldn't Imagon or Microtek make one for $5000 to $15000?

I'm from Missouri; so show me!

[Jeff Moore]

To follow up on Hogarth's comment with respect to the imploding market for drum scanners:

Recently (I believe it was on this site, but possibly another,) a Denver newspaper advertised two, yes two, Crosfield scanners and supplemental software and computer hardware for FREE. The scanners were there for the taking for anyone willing to pick them up. I enquired about these scanners and was sincerely interested. The person I spoke with informed me that the original purchase price for these scanners and the associated hardware and software exceeded a half-million dollars!! And they were giving them away, literally. And from what I gathered, not having much luck at it.

After my initial contact with the party to make them aware of my interest, I spoke to three different people in my area (Nashville, TN) about these scanners. Each of these people has over twenty years experience in the pre-press industry, two as scanning technicians and the third as owner of a mid-size publishing firm. All three were generous in their time answering the questions I had (I have zero experience drum scanning). Without hesitation or equivocation, all three of these very experienced individuals advised me to take a pass on this equipment, even if the scanners were to be delivered to my doorstep for free (I was going to incur considerable expense in transporting them from Denver to Nashville.)

The reasons for their opinions were numerous, but the common themes among each of their answers were:

1) The very high cost of maintenance, both in parts and the very specialized labor to perform repairs, which are not a matter of if, but when.

2) The questionable liklihood that parts would even be available in the very near future, when the inevitable breakdown occurs.

3) The learning curve for quality drum scanning, compared to something like an Imacon, is long and steep. While not that steep to get fairly decent results, the road to exceptional results (especially with b/w negatives, my primary interest) is an undertaking of several years experience. One of the techs I spoke with told me emphatically that it was literally impossible to get results from a 4x5 b/w negative out of a $250k Crosfield scanner that could be attained with an Imacon, no matter how many years experience the user had.

4) Drum scan manufacture is clearly a dying industry. The owner of the publishing firm predicted that, with advancing CCD and CMOS technology, the last drum scanner, by any manufacturer, will be made within the next five years, if not sooner. All of his company's drum scanning equipment has been gathering dust for almost two years now.

5) I primarily shoot 4x5 b/w and color transparency, then scan (currently on a Polaroid SprintScan 45 Ultra), and print my color work on an Epson 7600, with ImagePrint and UltraChrome inks; my b/w work is also printed on a 7600, using the StudioPrint RIP and PiezoTone inks. The publishing firm owner wasn't too familiar with these processes. However, the two techs both were, as they are both fairly serious hobbyist photographers. Both of them insisted that the newest Imacon scanner would give superior results--FOR MY PURPOSES--than the Crosfield scanner or almost any other drum scanner, for that matter.

In addition, Larry Danque, scanning technician at Cone Studios, gave me similar advice, especially with respect to the training/learning curve issue, and the maintenance/parts availability issue.

Again, I have no drum scanning experience myself. But the above opinions are from people who should know of what they speak.

Needless to say, I heeded their advice and took a pass on this equipment. Unfortunately for me, the Polaroid will have to do for now, as the Imacon is not in the budget...but, I'm saving my pennies!

I hope this information proves useful.

[neil poulsen]

Just as a comment, it wasn't that long ago, maybe a couple of years, that people who brought digital topics to this forum were really on the fringes of the conversation. It was like, "What the heck is that guy doin' around here? Is he lost?" More than one thread lamented the life threatening effect that digital could have on LF.

It's amazing how quickly things can change.

[QT Luong]

Henry, I plan to scan for reproduction up to 50x70 fine art (ie a 10x enlargement from 5x7). All prints won't be reproduced at that size, but at least some will, and if buying a high-end scanner I want to be able to "scan once for all".

Martin, certainly consummer/prosummer flatbeds scanners will improve, but for now they are pretty much at the low end. For now, dedicated toaster-shaped film scanners (ex: Polaroid Sprint Scan 45) are definitely better. Then the Imacons ("virtual drum" CCD scanners) are definitively better than the toaster film scanners. Many seem to think that the high-end flatbeds mentioned in my original question are better than the Imacons. And then, many seem to think that those are bettered by PMT drum scanners.

Jonathan, that scanner is the Fuji Lanovia. In the Seybold report cited by Doug, it ranked well, but below the two others I mentioned.

Jeff, I also saw those Crossfields and inquired about them. They are 1800dpi scanners, and it wouldn't be surprising that high-end CCD scanners could better them. This, combined with the difficulty of their installation and their sheer size and weight, made me also pass on the offer. However, more modern drum scanners have a much higher resolution (at least on paper), and a much smaller footprint. Last time I saw a comparison between an Imacon scan and a drum scan was on Michael Reichman's luminouslandscape.com (the comparison was indirect, as each of them was compared to a 1Ds file). The drum scan was significantly better.

[Frank Petronio]

"Back in the day" (mid 1990s) and even today most scans for print repro were between 20 to 50 mb. Even a full tabloid spread (17 x 22) would rarely top 200mb. I just did a 14 x 44 billboard that was printed at 9 dpi. It is only the fine art market that prints large and high resolution. This being a relatively small market, it is unlikely that a drum scanner manufacturer will create a product ideal for it (e.g. high res, long tonal range, modern software). I take the Flaar and Luminous Landscape reviews with a grain of salt - instead, look at what the most successful photographers are using in the real world - chances are you will come down to using an Imacon for a host of quality and productivity reasons.

That said, 45 Screen ai scanners sell for only $2500 around here - for that price you could hire a tech to come and give you lessons (and buy a couple of old Mac 9500s to drive it.)

[Bruce Watson]

I've added way to many words to this thread. But....

Just to answer John's question about PMTs and why they aren't being built any longer: money. PMTs are expensive to build themselves, and expensive to use optically and electrically. The whole design of drum scanners comes about because you could only have three sensors and you had to move the film past them (x and y coordinates) somehow.

The design of scanners using CCDs is way more flexible. CCDs are orders of magnitude cheaper to build and use, so cheap that you can build an array of them thousands of sensors wide and use this bar of sensors to make a single pass across a negative. If you can do a single pass like that, there isn't much point in using a drum - might as well use a flatbed to simplify the design even more.

So... why not PMTs? Costs too much.