I’ve been shooting film since high school. And by far the worst part of it is scanning the negatives. Paying a lab costs quite a bit (usually), and scanning prints loses all of the detail that film offers. So I decided to start scanning myself. I have an Epson V700, which is a fairly high-end flatbed scanner that is known for excellent film scanning. However, I find the process to be a chore. 36 photos to scan on a single roll, a few minutes per image, several rolls to scan… an hour and a half is a pretty decent estimate for the time you’ll be spending committed to scanning. And that’s assuming your first scan is perfect without any adjustments to the settings. (Spoiler: it won’t be. Add another half hour to tune it in) It’s gotten to the point where I stopped doing film photography, and I have a backlog of over a dozen developed rolls that I’ve never seen the images from. Oh, and did I mention dust on the glass bed making everything worse, and focus problems, and… I’ll stop there.
Enter the DSLR scanner. This is not my invention. In fact, the concept predates DSLRs entirely. In the film days of yore, if one wanted to copy slide film themselves, they used a slide copying rig. These rigs were designed so that one could easily take a photo of a slide, thus copying it.
As soon as DSLRs came onto the market, people started experimented with using slide copying rigs on them. They never gained much traction until recently. I can only guess that it was because early DSLRs lacked the resolution to compete with flatbed scanners. My Canon 5D II makes 21MP photos, which is perfectly fine for 35mm negatives. And I happened to have an old Novoflex slide copying rig that I got from a camera show. My task was now to combine them. Inspired by Peter Krogh‘s Nikon adaptation, I used a Really Right Stuff plate to connect the bellows to the camera body. I used some small clamps to hold everything together before I committed to any permanent modifications. By chance, the Canon 50mm f/1.8 lens I was planning on using had the correct filter thread size to mount directly to the bellows. A 36mm extension tube gives it nearly perfect magnification, too.
All I had to do now was figure out how to hold the negatives repeatably on the front of the bellows. I used cardboard as a quick and dirty way to make a proof of concept film holder. I reverse-engineered the bolt pattern from the front of the bellows and drew up a piece in Solidworks that I could cut out quickly.
With everything clamped together, I could do some testing. I used a roll of black and white that I had just recently developed.
I moved the film from shot to shot by hand, (Having a live preview on the camera LCD made this trivial) and let the camera take care of exposure.
Not bad for a day of work. The images curves slightly at the edges due to the tension of the film pulling down on the film holder, and there are huge light leaks. But the cardboard did exactly what it was supposed to do: Hold the film mostly flat against the bellows so I could see if this was feasible.
What’s next? Now that I know this will work, I can go ahead with permanent modifications to the bellows. The cardboard will be retired in favour of a negative holder from an enlarger. (They’re purpose-built for almost the exact same function) And perhaps some reels on either side so the film isn’t flopping about.