best flatbed scanner for scanning medium format chromes?
prints would be 11 x 14 at the most.

thanks,
Mike

The most relevant question is, "What's your budget?"

I'll make a stab at explaining the differences:

If you can afford a drum scan (or scanner!), by all means, they typically outperform all the other devices (assuming a competent operator). Prices start at about $20,000.

If you are interested in doing your own, the Imacon series of "virtual drums" is considered by many to be the next best performer. They start at $6,000.

Staying with dedicated film scanners, the next level is Nikon and Minolta for about $2,000.

Then you move to flatbeds with prices around $500.

Each device has its pros and cons.

Drum scanners are labor intensive (wet mounting) and materials intensive (fluid, mylar sheets, tape), but the technology is superior (lasers and photo multipliers) that can reputedly see through unexposed film and resolve well below the actual film grain.

Dedicated film scanners, which includes the Imacon, Nikon and Minolta are considered the next best alternative as they usually don't have glass between the film and scanner array, and due to their smaller size (relative to flatbeds), the physics are in their favor and manufacturers supposedly put a little extra $ into the scanner array in pursuit of higher samples per inch (often misrepresented as dpi) and Dmax (range of tones that can be discerned). Because of the physical limitations (actual pixel width) of the scanner array and diffuse light source as opposed to the drum's pinpoint laser, the performance is not quite equal to the drum.

The most common complaints with dedicated film scanners include maintaining film flatness (the Imacon sometimes suffers along the edges, the Nikon's holder requires snipping medium format frames in twos to avoid puckering, and still doesn't hold it completely flat in some instances) which, because of the very shallow depth of field the scanners are capable of, can lead to out of focus areas on the transparency. Banding in the shadows (multisampling eliminates some of this, but takes longer to perform a scan).

The flatbeds are the most versatile scanners, scanning both reflective art and transparencies. There is much debate about these scanners because the manufacturers often blur performance distinctions with misleading claims about dpi and Dmax. Some folks prefer flatbed scanners that have transparency drawers, thus eliminating the glass between the scanner array and the film. Some people advocate wet mounting on the glass platen (kits are available to do this). Lots of debate also about the software that drives the scanners, too.

With flatbeds, the resolutions is again limited (widthwise) by the actual pixel width of the sensors on the scanner array (think Xerox copier) that moves veeeerrrrryyyy slowly across your image by virtue of stepper motors. Lengthwise, the resolution (samples per inch) is limited by the smallest increment that the stepper motor can shift the array (by slices of a pixel width). Hence you'll often see resolution represented in two numbers (say 4800 by 9600). The most important of the two numbers is the smaller one (which is the actual number of pixels across the sensor array).

But, having said all this, the performance gap between the $500 flatbeds, dedicated film scanners and the drums narrows every day as the technology and software improves. I currently own an Epson 4870 and a Nikon 8000, and was trained to scan on a TANGO drum scanner. I have been doing some testing lately comparing results from the same piece of film, scanned on each, with surprising results. I'd be happy to email some comparison files to you for evaluation. The tests were made with slices of a 35mm transparency blown up to 20x24 inch print size.

If you are buying your first scanner, you're coming into this at a good time, because you are going to get much higher value for your dollar than if you'd bought something five years ago.

Regardless of which you decide to use, knowing how to manipulate the software (setting white point and black point, etc.) will get you much better results than the default settings.\

I have also been doing some testing regarding sharpening the file at the scanner level. Common wisdom says it is not wise to do so, but given the results I have seen, I am beginning to question this. For instance, I find that a file scanned at 4800 spi on the Epson with a Medium level of USM (unsharp mask) produces a file that I perceive to be equal to the Nikon at 4000 spi without any sharpening.

In summary, in the hands of a skilled operator, with prints that are 11x14, it will difficult for most people to discern the differences in quality of the image file at the scan level between all of the previously mentioned options.

If you search the archives on this site and http://www.largeformatphotography.info/ you will find numerous threads discussing this same topic.

Good luck. I hope this helps.

I'd be happy to email some comparison files to you for evaluation. The tests were made with slices of a 35mm transparency blown up to 20x24 inch print size.


I would be interested in seeing those comparisons. Can you post them here?

Don Bryant

They are a little large. 8 mbs for four files. I'm not sure Sean would appreciate something that large using up his valuable storage space, especially since in this environment, it is considered a "dark art." ;-)

I hate to go much smaller with the files because then the comparisons are harder to discern. I can email them to you. It is not bad if you have a cable modem or DSL. If you want them emailed to you, shoot me a pm with your email address and I'll get them off to you.