Limited by my scanner: each frame takes 2 1/2 minutes to scan, frames need to be manually straightened if a restore is to take place. Because the scanner is only able to scan 22 frames in a single go, I need to be available every hour or so to place new film in the scanner, ensure it’s in the proper place, and hit “Scan.” This single thing alone makes this a full-time job, as it can take up to a full work day to scan a single long filmstrip. Archival-quality 35mm scanners are prohibitively expensive and little, if any, testing has been done with filmstrips on motion picture scanners to see if they’d even work with pieces of film this short.
You can scan still photography on a motion-picture scanner, but in order to do that you need to splice all your strips together into large reels, for example 1,000ft.
Acquire better scanning hardware: The sweet spot seems to be a BlackMagic Cintel at approximately $35,000, simply because every other machine in the market starts at almost a million dollars. However I have not specifically set this as a current goal because I still need to cover operating and household expenses, even if a Cintel were just to magically show up on my doorstep this afternoon. From the samples I’ve seen of Cintel transfers, it is able to register each film frame in exactly the same place, which would eliminate the crop-and-straighten pass for restores.
The other machines do not cost $1,000,000+!
Please don’t take this the wrong way, but you want a still photography scanner not a motion picture scanner. A BMD Cintel is not suited to you needs at all, and it would not be cost effective. Professional scanning on better machines than it generally starts at .20/ft, I presume your filmstrips are around 2-5ft in length each, so to put them through a motion-picture scanner you’re going to have to splice hundreds of them together just to make small reels - heck just a 300ft reel would consist of a minimum of 60 film strips. If you did all 3,000 filmstrips and they average say 4ft each then it would cost you about $2,400 to scan that way - although it would be an awful lot of work for yourself as you would need to: spice them all together, and then break them back down, check the scans are not clipped (ordinarily this would be checked for you - but you can’t expect 60 different “clips” spliced together to be individually checked - you’d need to do that yourself).
The still photography scanners cost a small fraction to purchase or build yourself compared with a motion-picture scanner.
The old still photography scanners (like Fuji Frontier SP3000) are not suited either. They were only designed to scan negatives, and you have positive print film.
If I were you I’d look into the Filmomat 135 Autocarrier. Setting that up perfectly to get the film as “square” as possible would require you purchase some SMPTE resolution film. You can DIY everything yourself if you prefer. There are significantly less challenges with still photography scanning compared with the motion-picture scanners. You can buy functionally the same camera that the Cintel uses for about $600 used, or you can buy something much better for not that much more and you’re not limited to global-shutter cameras which saves you $$.
The optical perf stabilisation you mention (that Blackmagic does in the Cintel’s hardware) can be done in Fusion.
I can’t speak to how perfectly flat a table-top system like the 135 Autocarrier can get the film. The motion-picture scanners have sophisticated film transport modules that provide constant even tension for the film, whereas still photography scanners load the film in as a strip with no lateral tension. Your best bet would be to get someone that already has one to do some sample scans for you. The geometric imperfections would be less important for you anyway, as mentioned it would be more about getting the frame as “square” as possible so that you don’t have to rotate the film digitally at all.
If you have strips that show excessive geometric distortion you always have the option to do them on a commercial motion-picture scanner, doing a few 300ft reels that way won’t break the bank - it’ll just be a lot of manual work on your end to do that.
I hope that helps!