Review Details

Super 8/8mm Film Scanner

Product Rating:

Product Review (submitted on June 14, 2021):

I transferred dozens of reels of super 8 and a few 8mm reels successfully with the MM100. But, I'm extremely demanding - and stubborn. [I swear a lot, but I keep plugging.] The quality of the results was fine, especially given that the source was just ~8mm home movies. But, the learning curve was long.There is only one reel motor. So, to rewind, you have to swap reels. [No big deal.] But, there is no mechanism to provide different torque values for play vs. fast-forward, such as a clutch, so the take-up tension is the same when 'rewinding' (fast forwarding with the reels reversed) or playing/transferring. So, when transferring, it pulls very hard on the film, which requires the pressure plate at the gate to press pretty hard on the film. This means that the sprockets will likely get damaged after a few tries. And, the pressure plate adjustment is very finicky, and there's a tendency for the framing to vary. I got around it by not using the take-up reel at all. I loosened the pressure plate at the gate to be very gentle on the film and just let the film fall into a clean bucket on the floor next to the table that the MM100 was on. Worked very well this way. [And, I often rewound by hand.] The framing was very stable when I skipped using the take-up reel.This thing transfers one frame of film to one frame of video. But the frame rate for video is normally 30fps, and for Super 8, it's 18fps; and, for 8mm, it's 16fps. So, if you want the video to run at the correct speed, you have to correct it with FFMPEG.Using FFMPEG is no picnic. Typical of software from the UNIX/POSIX/Linux world, to figure out how it behaves, you have to do a LOT of experimenting and/or read the source code. For example, I found that, when splicing two videos with the same frame rate and time scale ('video_track_timescale'), if I didn't specify the same time scale for the output file, FFMPEG would use some other value, which screwed up the running speed. I could find no documentation of any of this. It was all trial-and-error. [I posted some of this to an FFMPEG forum. and had people asking me what I had figured out.] If you abbreviate a time value in any way (leaving off zeroes), FFMPEG is almost guaranteed to misinterpret the value, do the wrong thing, and not report a problem.This could be a big opportunity for the Wolverine folks. I would guess that they're familiar with all of the things that I had to figure out by experimenting with FFMPEG. If they were to document it, it would make things a lot easier for people who buy this.I assumed from the start that splices wouldn't make it through the gate. So, I separated them and cleaned the goo off of the film. If appropriate (related scenes), I spliced the scenes back together digitally with FFMPEG.As I got better with this, I did more work in parallel. I'd transfer one film, take out the SD card, move the file to a PC that had FFMPEG, return the card to the MM100 to start another transfer, and then use FFMPEG to fix the running speed, split up unrelated scenes, etc, on the first video.I kept copies of all of the intermediate files (in case anything went wrong), and needed a fairly complex file naming scheme to keep track of it all.Using this is a LOT cheaper than having professionals do it. [And, this device is much less expensive than professional equipment.] And, I doubt that the quality is much, if any, worse - given the source. But, doing a good job requires a lot of patience and ingenuity.