Flatbed v. drum scans for lightjet prints - Microtek 1800F.

I would like to hear from experiences that you might have of using a flatbed scanner for making exhibition-grade color lightjet prints, and how the results compare to using a drum scan. What would be the difference in quality ? Neglectible, small, or very visible ?

I am particularly interested about results obtained with the Microtek Artixscan 1800F, but positive experiences with other scanners such as the Epsons would be encouraging, since it appears that on paper the 1800F betters them.

I don't expect the resolution to be a limitation. In the past, i have used a very cheap (now broken !) 600dpi flatbed to make a 12x18 from 5x7 and the print was sharp, easily beating anything done from 35mm. Even if the real scanner resolution is only 1200dpi, one could still make a 4x to 6x enlargement. What I am wondering about is the dynamic range and noise.

There are of course many variables at play here. As you have seen, much depends on the level of enlargement. My context here is from the point of view of 4x5, BTW.

If you are scanning to make 2x enlargments, probably any scanner you have access to will do the job. Consumer flat beds (Epson 2450 or so) seem fine up to about 4x enlargement to me. Better flat beds do visably better, I think. After about 4x, I can really see a difference when you go to drum scanning, especially if you oil mount.

What is "better?" What I see as you go from consumer flat beds to pro flat beds is smoother tonality, smoother transitions, somewhat better dynamic range, somewhat better apparent sharpness. It's sort of like removing a veil (thin, uniform) from between you and the print.

When you get to drum scaning, what I see is improved sharpness, improved smoothness (especially transitions), and much improved dynamic range. It's like removing a somewhat thicker veil from between you and the print. A lot of this is due to PMTs vs. CCDs. Quite a bit is due to oil mounting on a drum, especially the smoothness.

If you really push your transparencies (and you do, at least what I've seen on your website) and you want maximum quality, I think you are going to be happiest with a drum scan, especially at an enlargement in the 4x to 6x range and larger. The reason is, you'll see more detail in the shadows, and all most all of the smoothness of tone and transitions that you see on the light table.

Of course, YMMV. The easiest way to tell, is to scan the same negative multiple ways and make prints (same size of course). Put the prints on the wall together under the same lights, and have an apples-to-apples comparison. The one you are happy with is the one you are happy with.

To satisfy my own curiosity I did a comparison between how much detail that I could extract from a 4x5 velvia slide with my Epson 1640 and having a professional drum scan done. ( I used a 1/2" tall section of the film and got a 100 meg scan)

The following is a link to a power point presentation I did showing the difference. To my mind it is astounding. The name of the construction company on the crane is 1/10th of a mm tall on the film.

http://truckgenerator.com/subdomain/sueandneal/ppiscansvelvia_files/frame.htm

(It takes a while for the 12000 dpi to load over a phone line.)

I also took the drum scan into Photoshop and backed the resolution down until I got back to 1600 dpi. The difference between the drum scan and the flat bed scan was again dramatic.

My conclusion is that for posting to the web I will use my flat bed scanner but if I decide to go through digital to get prints, I will spend the money to get a drum scan.

I have also tried going above the optical resolution on the 1640 and find that all I get is a bigger file.

Neal

All I see here is a fuzzy image - no powerpoint presentation or any comparisons -

If you are looking for a reasonable alternative to drum scans that could save some money, you will not find it with the <$1,000.00 flatbeds. Period. The Epson 1680 Pro is likely the best flatbed you can buy, but still falls short of drum scans, but only at higher print sizes is it noticeable. For gallery quality, I would not even consider less than a drum scan.

i agree with matt, "gallery quality" to me means "highest possible standards" unless your particular asthetic is deliberatley low quality. i did some tests with my local repro house comparing flatbeds, "virual drums" (imacon), tango drum and really high end heidelberg (8900). the difference is real and there to be seen, by even the most untrained eye.

Tuan, I recall this subject was touched upon in photo.net lf forum at the time this one was not operating. You might try checking this out.

http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=004N6n

Matt,

You should be able to switch between different scans by clicking on the arrows at the bottom of the page, the titles on the left, or by clicking the slide screen and then using your left mouse button. It works for me,but I am still learning to build web pages, I may have screwed up.

Neal

I don't think the issue is resolution as much as dynamic range. I scan my 4x5 trannies on an Epson 3200 scanner and the resolution is pretty good. However when I want to make an exhibition print I rent time on an Imacon scanner. The sharpness is definately superior and the increased dynamic range make getting the color I want in the print much easier. There were some images where the highlights were completely blown in the Epson scan while the Imacon got good detail in the hightlights.

- Dan

Re Neal's webpage with scanner comparisons: it won't work in any browser that I regularly use (Opera 6.11 and Netscape 4.8). Looking at the source code, it was written by Microsoft Power Point. It will probably work in Internet Explorer but few other browsers. Microsoft has difficulty writing standards-compliant webpages. I've encountered plain-html webpages written by Front Page that show many dozens of basic html errors when checked by an html validation webpage.