Automate analogue film scanning with Raspberry Pi and LEGO
This automated analogue film scanner runs on a Raspberry Pi and LEGO bricks. BenjBez took to Reddit to share this incredible lockdown project, which makes processing film photographs easier.
Benj explains:
“When doing analog photography, scanning is the most painful part – RoboScan tries to make the whole workflow easier, from the film to the final image file.”
Mesmerising, isn’t it? We don’t know why we want it, we just do. We love it when new technology supports traditional methods with hacks like this. It reminded us of this Raspberry Pi powered e-paper display that takes months to show a movie.
How does it work?
The film roll is fed through the LEGO frame and lit by an integrated LED backlight. Machine learning detects when a photo is correctly framed and ready for scanning, then a digital camera takes another photo of it. RoboScan downloads the photos from your digital camera as soon as they are taken. Only 80 photos were used to train the Raspberry Pi and Benj has shared the model here.
But I only take digital photos anyway…
Most of us rely on our phones these days to capture special moments. However, we bet loads of you have relatives with albums full of precious photos they would hate to lose; maybe you could digitise the negatives for safekeeping using this method?
Best of all – it’s all open source and available on GitHub.
Thanks, Electromaker!
We love our lovely friends at Electromaker and we found this project through them. (They found it on Reddit.) They release a new video every week, so make sure to subscribe on YouTube so you don’t miss out.
7 comments
Bert
At first glance I was about to feel huge chagrin about not knowing about this project, because I just finished a long project of going through our family photo film negatives from the pre-digital area. Approx 125 rolls worth. However, our negatives were returned to us (from the film processor) cut into strips no longer than 4 images wide. I ended up using an Epson V600 scanner, typically 8-10 images at a time once set up.
For anyone considering doing this: don’t wait! Consumer-grade film deteriorates over time, and IME it’s a function not only of the brand of film, but the processor used (age and storage conditions are factors too, of course). Ours were 25 years old on average, and most had experienced color shifting that needed correction; a few had uneven and severe color shifting that was hard to correct. Similar admonition for those dealing with (positive) color slides. I just finished working through a cache that was, on average, 50 years old. Prints are affected as well, though they tend to fade more evenly. Anyway, if you’ve got some film or prints you want to digitize, don’t wait, make it a priority – they don’t get better, and often get much worse.
Benjamin
Benjamin, here, the author of the project. I agree with you, it’s important to scan and preserve the photos, some films won’t last forever. Plus it makes a cool trip down the memory lane.
So whatever the scanner, automatic or not, it’s always worth it!
Alejandro
Is it compatible with super8 video film? For video purposes…
Benjamin
The problem of Super-8 is that each frame must be at the exact same location as the previous, to prevent the movie from jittering. So the best is to use a dedicated super-8 film advance mechanism. For example, I found this: https://hackaday.com/2020/04/12/8mm-film-scanner-grows-into-a-masterpiece/
AndrewS
Similar-but-different to https://www.raspberrypi.org/blog/kinograph-preserving-1920s-movies-with-a-raspberry-pi/ and https://www.raspberrypi.org/blog/digitising-reels-pi-film-capture/ :-)
Ed Chapman
If you want to scan, archive, sync and collaborate on more content or convert images to pdf you can use a similar scanning app called camscanner premium unlocked, free download at TechBigs https://techbigs.com/camscanner.html
sergio
I Will one day use raspberry pi in my document projects, as of now I only use camscanner.
https://www.enttechub.com/2020/02/camscanner-mod-apk-free-download.html
Comments are closed