📽️ 🎞️ 📸 ⤳💾✨ DIY 🛠️ 8mm / Super-8 Film Scanner, Part 1: Overview
HTML-код
- Опубликовано: 25 июл 2024
- 📠 This is a video about film, my project using a projector. 😉 The goal: "scan" film using a mirrorless camera + macro lens set-up, making a 4K+ RAW digital photo-to-video workflow a possibility.
Commercial film scanners exist, but I wanted to try building my own, and also have the benefit of best "theoretical" quality. With an uncompressed source, I have the ability to edit scanned footage in a lossless way before exporting a final video - either in lossless, or compressed-only-once format.
Certainly, 4K raw etc. could be considered overkill. Most exports I've done thus far have been at 2048 x 1536, given the 4:3 aspect ratio of film.
Notwithstanding - in the words of Tori Belleci from the TV show, Mythbusters: "If it's worth doing, it's worth overdoing."
12/2020: Part 2 is available, as well. • 📽️ 🎞️ 📸 ⤳💾✨ DIY 🛠️ 8mm...
03/2021: Part 2.5 with Q&A, and some updates: • 📽️ 📸 💾✨ DIY 8mm Film S...
04/2021: Part 3, finally! • 📽️ 📸 💾✨ DIY 8mm Film S...
📠 HARDWARE MODS
Using a Canon film projector as my base, I added my own stepper motor, 12V automotive LED light and Arduino hardware/software controllers to orchestrate the entire sequence.
📠 PROCESS
The process, roughly, is "advance the film one frame, and take a picture" ... thousands of times, up to three times a second. At 18 frames per second, a 3-minute Super-8 video has 3240 frames. 😅 With my current set-up, that's a little over 64 GB of raw .ARW files. It's a lot of data, but it's worth it!
📠 WHAT'S HERE, WHAT'S NEXT
Part 1 covers most of my early work and findings, in the process of getting digital copies of film. In the next part, I plan to dig more into the camera, film, and post-processing workflows that worked for me. Stay tuned. 📺
📠 WHAT'S UP WITH THE FAX EMOJI? 🤷
Y'know - so you have the *facts*. WAKKA WAKKA. Sorry. 🤣
Comments, critiques and questions are welcome! I've had fun both working on this project, and making this video. I'm not an expert by any means, on any of this stuff; I'm a software engineer by profession. Hardware and tinkering as such is not my usual hobby. Like many other things, I consider it a work in progress. 😅
🙇 CREDITS
I owe thanks to a number of people, videos and web sites for inspiration and guidance.
Ken Eckert's adventures in using a real scanner to digitize film:
• How to Convert 8mm/Sup...
Related studies in motorizing projectors, etc.:
keneckert.com/kenfilms/telecin...
Dronebot Workshop on Stepper Motors:
• Stepper Motors with Ar...
Curious Scientist, Arduino + TB6600 controller:
• TB6600 and Arduino - W...
🎼 MUZAK 🎵
I'm asking for forgiveness, not permission, on some background tunes. It seems that RUclips's bots haven't picked up on my selections at -24 dB volume. 😅😇
Fantastic Plastic Machine - "The Girl Next Green Door" (1999)
• Video
Piero Umiliani - Lady Magnolia (Royal Belleville Orchestra - Kid Loco Remix) (1998)
• Lady Magnolia (Royal B...
⏲️ Timeline:
00:00 - Introduction
00:58 - "Step 1": Stepper Motor + Arduino
01:40 - Variable Speed Motor w/10K Pot
02:49 - Widening projector Film Gate
04:14 - Removing film shutter "fan blades"
04:52 - Installing the Stepper Motor
05:46 - 12V LED Light Source
06:33 - LED diffuser, mount
07:15 - "IR Sensor"-based Camera Trigger
08:04 - Working Prototype!
08:22 - Camera set-up (teaser)
08:46 - Thank You, sample "Star Wars" Super-8 clip / end-of-film sequence
09:27 - THE END
Very cool. Building my own right now. I'm unclear on what you're using to trigger the light and camera shutter circuit. Is it a Arduino timing you've set based on the speed of the stepper motor?
Hey Joe, thanks. I didn’t go into the specifics of the circuitry, but I plan to get more of those details and the Arduino code up in a future video and on GitHub or similar.
There was a screw holding a metal ring on the projector “drive shaft” that I tied a bit of material to. It could be anything solid, the idea is just to have some shape that blocks the U-shaped infrared trigger / gate - a light emitter / collector pair of sorts.
The IR trigger is connected to the Arduino, and an analogRead() is close to 0 when open, 1023 when interrupted (or vice-versa, I forget which at the moment.) When the value changes, I then trigger the shutter. To keep things simple, I also ended up keeping the LED light on all the time - it’s been fine, no issues with heat - and that also helps to reduce one more timing issue, worrying about the LED light being at full brightness or what have you otherwise.
The timing of frame position -> trigger is adjusted simply by loosening the screw and repositioning the ring so the attachment sweeps through the interrupter after the projector “fork” has pulled the new frame into place in the gate. This projector has a fork that goes up, reaches in, pulls down and then out of the film sprockets. So I try to have the shutter trigger once the fork is out, and that gives plenty of time to get the photo.
The trigger for the camera, Sony a7riii here, seems to work if you hold the lines down for 10 msec. Doesn’t need to be too long (you might get multiple exposures,) so it’s best if you experiment and see what works for your setup.
@@scottschiller Thanks for the explanation!. Looking forward to the next video.
Most welcome! And yes, I need to get to work on Part 3 soon, I’ve had a few people show interest. I’m glad folks are finding this project useful or inspirational.
Hey Scott. Finally finished my own scanner build. I used many of your ideas so a huge thanks. I snagged the same Canon projector off ebay because of how easy it was to mod on the inside. I decided to not use an arduino and instead used a stepper motor controller signal generator.
smile.amazon.com/dp/B07HNSVMVH/ref=cm_sw_r_cp_apa_glt_fabc_ADNYDN89RSTASQF2SMJW?_encoding=UTF8&psc=1
I also wanted a camera that would take raw files (.cr2) so i found a used Canon 50D on Facebook Marketplace for $100. Yeah I'll kill the shutter mechanism but I'm fine with that for now. The cr2 format allows me to import it directly into resolve. I'm still working out the color pipeline but I'm leaning towards an Aces workflow since they have a Canon color transform already.
ruclips.net/video/vqEwg8Yxg7U/видео.html
8mm scanning projector imgur.com/gallery/JFAkKIy
Looking forward to seeing your future video on Resolve workflow.
Wow, this looks amazing! 🤩 Fantastic and very tidy work on the wiring and electronics; mine is very messy by comparison. I think your IR beam sensor approach looks great, and the simplified sans-arduino stepper motor idea is smart.
I started with the Arduino as a base because that’s what I was familiar with, but I like the idea of side-stepping it entirely. Also, your shutter trigger approach is cleaner because you aren’t going through an arduino for that, either.
Again, fantastic work! I’d love to hear more about the CR2 approach in DaVinci, also. Your timing is perfect; I just wrapped Part 3, it’s about 75 minutes long and will be “live” in 10 minutes. I also mention doing a Part 4 on electronics details, as well as doing the post-processing in Resolve - so it’d be great to hear what you find, how it works for you.
Finally: I recommend looking into Magic Lantern, the alternate (temporary) firmware for Canon DSLRs. There is a feature that I think allows complete silent full-frame capture, no shutter actuation that you might be interested in. Worth a look!
Just a shout out and a thank you for posting this series. I've just completed my setup using a combination of your and Fresh Ground Pictures approaches, a worn-out, reworked Sankyo Duplex 1000, and my Panasonic S1. I've refined the build and I'm pleased with it. Just finished my first test capture, post-processing run-thru, getting my head around Davinci Resolve 17 and applying Neat Video. I'm familiar NV from FCPX but your orientation to Davinci - the import of an image sequence, the opportunity to set a frame rate to 16fps or 18fps - things you cannot do in Final Cut.... All so helpful to make a switch. Thanks for everything that you have put out there! It kept me motivated to create my own when after a year of researching approaches I had just about given up.
Thank you for the kind words, Walter! One thing to note, on your S1 - make sure you have silent mode or equivalent so there is no mechanical shutter movement, so you aren’t burning the lifecycle of your camera hardware with all those pictures. It sounds like you’re probably aware, but I try to mention it to everyone just to be safe. 😉
“Fresh Ground Pictures” did a great job with their video also, I saw it just recently and it looks like they improved on a few aspects of my home kitchen-based layout! 🤣
@@scottschiller Thanks for the mention of the electronic shutter - yes, I have definitely implemented that. The big appreciating to the FGP approach was his incorporation of the mechanical trigger for the shutter. That made this process so much more accessible.
Fantastico!!!!
Very Cool!
Thanks for the idea. Build a scanner with same function. The quality is nearly perfect and after the process on the computer in a very good quality of 1080p or even 4k, because I scan with a resolution of more than 4k.The only thing that's not perfect is a little hotspot in the bright regions of the image. Have tested a lot of led bulbs. Some work better but no one works perfect. Overall thank you for the videos. That inspired me and I'm very happy with the results.
That’s great to hear, and thank you for sharing. I’ve noted some highlights being pretty bright myself, and I try to bring the LED lighting down (or increase the camera shutter speed) to avoid over-exposure, with the intent of preserving as much detail as possible. It can be a tough balance at times!
Wow 😮
Very cool Scott. I'm doing something similar right now. That piece of glass in between the bulb and the film was actually an IR shield. It protected the film from the heat of the old bulb. If you're running an LED it still gets hot but it's way less heat than the old bulb. The internal fan will keep it cool enough to not damage the film.
Thanks, Kevin! Thank-you for the IR detail, too; I was wondering what purpose that little glass piece served. I did testing with my 12V automotive LED set-up, and I found it hit about 110F - toasty, but not hot enough to melt the ping pong ball I had it "mounted" in or anything like that. Notwithstanding, I put a little fan in there to keep air moving just in case.
@@scottschiller Yeah, some sort of active cooling is good. I actually replaced the glass with a piece of diffusive plastic because you could see the individual LED elements lighting up squares of the film... I actually have a giant heat sink on the back of mine and a little Adafruit RPi fan running straight off the 5v arduino rail too. Then I only run the LED at 33v instead of the 37v it's rated for and it stays cool to the touch even after being on for hours. I'm watching part 2 now - it's amazing how similar our setups are!
This does sound rather familiar, yes!
My automotive LED has a built-in heat sink, and runs at 12V - so I use the same power supply as I do for the stepper motor. Early on, before adding diffusion, I was also noticing individual LEDs shining through the film.
Hi Scott,
Thanks for all the videos you made about the DIY 8mm Scanner, I admire them, they are very inspiring.
Although I have watched them all, I don't know how big the camera frame needs to be to capture the whole area around the frame even with the sprockets. I found out the purpose of doing this with Ken Eckert, whose work you mentioned in the Part 1 video . His The Magic Wand logically explains the principle of how stabilization of captured images works in Adobe Photoshop, but I don't know how it works in DaVinci Resolve?
I'm now starting to modify a Meopta N8/S8mm projector and after removing all the parts I saw in Part 1, my film gate is only 6.36 x 5.34mm and that doesn't meet the requirement for the camera to capture the entire surroundings of the film image including the perforations. On the N8 the sprockets is visible from 3/4 of the size , but on the S8 only a hint of about 0.1mm. No surroundings are visible on the opposite shorter (vertical) side - the image matches the film gate in both N8/S8 formats. I will have to enlarge the film gate symmetrically on the shorter sides, but how much? Does the whole sprockets need to be visible, or just half of the S8 as seen in the video in Part 3? The top and bottom film gate allows to capture the film image still with some margin. I'm asking as a beginner, but I have no experience with film post-production yet and will purchase DaVinchi after I get a faster PC. Looking at your videos, the perforation window is fixed, it doesn't move at all, it's OK.
I have a lot of my parents' old B&W 8mm film from the late 1940s, and the perforation is no longer in the best shape. So functional fixation throughout the transfer process is a very important topic for me.
I found that my projector’s film gate was both wide and tall enough to show the full width, including sprocket holes, and perhaps 1.5 frames’ worth of height.
It depends on the film, but the exposed image may extend across the full width - basically “behind” the sprocket holes or not. If the picture doesn’t go that far out, you can crop out the sprocket holes. I am not clear on what type of camera or process makes the difference between “full-frame” width and otherwise.
I think some folks have had luck filing out the film gate, carefully, to widen it. Your mileage may vary.
Oh, also - DaVinci Resolve is free to start with and you only need the paid version for certain features. I use the free one for all my stuff, and it’s perfectly sufficient.
Awesome Job, this is the best DIY on scanning I've seen. I am looking into ways to accomplish this myself and the only reservation I have is the film moving by sprockets but I can't get my head around that Mod. Anyway, not a question, just a comment, really enjoyed this .
Thanks for the kind words, and I’m glad you enjoyed this video! There’s a “Part 2”, and I’m working on a third as well. I hear you on sprockets vs. rollers or whatever else to pull the film. Thus far, everything I’ve scanned has worked via sprocket holes. Trying to move and align the frame without the built-in mechanical film gate would be a project in and of itself, but I know there are people who’ve done it!
@@scottschiller ONWARD and UPWARD ,,, I'll subscribe and also view those other videos ... Best,
thanks, learned so much.
Glad it was helpful!
Hi, Scott. Thank you for your inspiring videos. I am about to start a scanner project, and I will probably go in the lines of your own project. Your fine and detailed presentations will be extremely helpful... I have a question: I need to select a decent camera. The camera you use appears to be perfect. But it is a costly top-of-the-line camera. Is a 48 Mp sensor really usefull? What drawbacks if using a much cheaper 24 Mp camera such as a mirorless Sony ILCE 6000 ? Or have you a suggestion based on your experience?
Thanks for the note, and the kind words. Here’s my $.02 on the camera: I don’t think you need 42 MP or what have you, but having ability to shoot RAW format - and crucially, with a 100% electronic / motionless / silent shutter, is key. You don’t want to be burning out the mechanical shutter mechanism on a camera, shortening its lifespan (or killing it entirely) taking thousands of frames for 8mm/super-8 scans.
When looking for cameras, I would consider silent / motionless capture as a top priority, followed by RAW (lossless) image support (not only JPEG), and also the support / availability of a macro lens which will let you get a good close-up view of the film gate.
I should add - I did a quick search and it looks like the Sony ILCE 6000 has a mechanical shutter which can not be fully silenced, there will always be some movement when taking a photo. So if you were to use this or similar, you would be contributing to the actuations and mechanical lifecycle of the shutter mechanism.
I watched it all and got some great ideas. great project. I have seen some film dude in Italy is doing his own scans with 6k camera and his setup uses the stepper motor and led but it looks like he got rid of everything besides just the motor some reels and simple acrylic box for the led and film gate. Looks like he got rid of the film projector all together.This looks like a great little setup. If I had one I would be shooting way more instead of having to decide if its worth it and deciding its not....LOL
There are some neat builds out there, for sure. I’ve seen some where people have the film gate and little else, like you mention. I’m not very mechanically-inclined, so I took the route of modifying something that I knew was already reliable: a full projector.
Side note: I think you made a comment about that little piece of glass being a heat guard, but it seems to have disappeared. Let me know if you had trouble posting; I’ve had some comments vanish and I wonder if they got blocked, or if there was a bug. Thank you!
@@scottschiller thanks. I made a comment about the glass but I kept watching and saw that you covered that so it seemed a little redundant to mention it so I deleted it. LOL
Oh, ha! Thanks for clarifying. I was concerned there was some RUclips bug perhaps, and people were losing comments or something. As you were, then! 😆
Hi Scott thanks for this in-depth explanation and showcase of your scanner! I have built a 16mm scanner myself and I realise watching your videos, that the film runs really, really steady in your scanner! can you explain how you set up your stepper motor+potentiometer to always run the film just one frame? thanks in advance!
Hi Jonas,
I’ve been curious about a 16mm scanner as well - I bet there’s some great content out there, and at higher definition vs. 8mm.
I cover most of the stepper motor stuff here in Part 1. Largely, it boils down to the motor running at a speed much slower than a typical AC motor. I found that the projector “drive shaft” advances one frame for each rotation, so I used that as a trigger for the camera. I’m not very mechanically inclined, but my plan was to take advantage of the film gate and transport mechanism you get “for free” by modifying an existing projector.
A very fine work around to a home media challenge of the 21st century: how to effectively digitally transfer small gauge film in a high res' format. Unfortunately, a complex project such as this is a little out of the range of most home film-makers.
Thanks, Peter. I agree, this is a non-trivial hardware and software project that requires a certain amount of knowledge, and perhaps a “tinkerer” mindset. I am not an expert by any means, but I am happy to say I figured out enough to get some pretty good results, fairly reliably. It still takes a lot of time and effort to do the scan, but it’s really nice to have this sort of material digitized.
excellent job.
Now clear the table, your dinner is ready!
Ha! Thanks, and yes - it was nice to have this on my kitchen island for photography and some tinkering, but I was also happy to free up that space after making this video!
Fascinating! I'm using the same overall technique, photographing every individual frame as it passes the gate in my Elmo GS1200. I'm using the same Laowa macro lens as you have, feeding my Sony A7Rii. The results I'm getting are excellent, but I just wanted to say that your Laowa lens appears to be set at f/16, and I wanted to let you know that this actual lens performs very much better at f4, regardless of the magnification it's set to. Diffraction is the reason, as I'm sure you know. As you're photographing a flat plane with the lens perfectly perpendicular to the subject (the film emulsion), you don't need to sacrifice resolution for the added DoF that f/16 gives you.
This is great, Tom; thanks! I’ve been using the lens at f/8 mostly, but I did some earlier scans at f/5.6. I will revisit this and maybe try a side-by-side comparison.
It seemed like the DoF was so shallow on this lens, I thought f/8 would be a reasonable compromise giving me a better shot at having that razor-thin field of focus working better. I think you have a point though, there are trade-offs.
I found it most revealing (and quick) to photograph my 4k TV screen, using the 42.4 mp Sony, at all the Laowa apertures and at three different magnifications. There really is a sweet spot at f/4, though 5.6 is useable. Beware of using any smaller aperture, especially past the 3x magnification point. The DoF is indeed increased, but in our case we don't need that benefit.
This is a smart idea, thanks. I'll take a look at f/4 and reducing the magnification as much as possible. At present, I'm a shade over 3X. If I can get a sharp (focussed) image, exposure time can also be reduced and scan speeds can go up. My scanner can run mechanically up to 5 frames/second. It seems like the remote shutter on my a7riii "maxes out" around 3 frames/second, though I haven't dug much into this.
Bravo, congratulations, your project has nothing to envy the absurd professional cine scanners that cost (€ 12,000-35,000) The results are the same!
Thank you, Bruno! I do think those professional and high-end scanners are quite impressive, but they are certainly not cheap. My hope has been to get pretty good results using my own approach, with those pro scanners as a reference for quality.
@@scottschiller In what ways would you say a pro scanner is better than yours?
It’s likely the professional ones are more battle-tested, having been developed by companies for whom this is their business. They may have automatic cleaning abilities, and some can use rubber wheels instead of gears or cogs, and thus they can “play” film that a standard projector might not - e.g., if the film has shrunk somewhat over time and its sprocket holes are no longer consistent or the right size etc.
These are a few possibilities, anyway!
This is unreal Scott. Insanely well done. Am i correct in saying that i could manually feed (pull) the film through while the camera is continuously shooting?
Thank you! This project started as a rough build, but I’ve been able to get it working quite reliably after some trial and error.
The way I’ve set this up, the process relies on the stepper motor driving the mechanics of the projector, including the “fork” that pulls each new frame into place. By doing this, you get the benefit of the projector’s whole reel take-up and rewind process, as well as knowing when to “take a picture” of the new frame in place.
If you were to try to pull the film through by hand or some other way, you’d lose the convenience of the projector perfectly aligning each frame, holding it still, and knowing when to “scan” it. You’d have to re-invent all that stuff, and I wouldn’t recommend going down that path. The film gate and its mechanics are the most important part to moving and aligning frames to scan. I’ve seen projects where folks take only the film gate and “fork” bits out of a projector, and then make their own source and take-up reel mechanisms and all that.
Using the stepper motor approach, you get the benefit of all the existing functionality of a working projector - and I can’t recommend that approach enough.
For the stepper motor vibrating thing, I wonder if adding one of those "smoothers" that are used on 3D printers, to quiet down the steppers, might work for this?
I’m not familiar, but I wouldn’t be surprised if there’s a trick to reducing the frequencies that cause the most vibration. In my experience, most of the time I could avoid this by quickly ramping the speed up and down. 😅
Great project, and excellent results. I am using the gate from the same projector Canon S400, but after removing the cutout (3:32 in your video) the frame does not show the sprocket. Did you have to enlarge the gate hole? any advice on how to best do it? Thanks in advance.
Thanks, pm490! I'm intrigued by your question. Once I removed the frame "cutout" - the piece that moves depending on selection, 8mm vs. Super-8 - I was able to see the sprocket holes on both types of film. I didn't have to enlarge the hole in the film gate here, no. Once I removed the bit of glass (an infrared filter?) and the cutout on the back side, that was all that seemed to be needed.
Assuming the back side is uncovered at this point, is there any chance something is covering part of the film gate on the front? I could try measuring what I've got here, but I'd wager it's between 9 and 10mm across. For reference, 8mm/super-8 film dimensions: en.wikipedia.org/wiki/File:8mm_and_super8_and_double8.png
@@scottschiller Thanks Scott. It is odd but looks like the same projector model has different size opening. I see yours is rectangular, mine is virtually square and measures 6mm wide. From the video I see yours is wider. I do see a bit of the sprocket hole on 8, but the S8 is cropped. Thanks for confirming, I am debating to file it wider or leave as is. And in my case will have to do both the front and back, both have the same size. Thanks and again great work. I took the approach of getting the gate out and spin it with a stepper. The one in the video was the first run, once I decide on the filing will make a video of the current built.
Thanks for the follow-up, this is great info. And yes, it’d be interesting to see what you’re building!
Whoa. This is awesome Scott. DM me and I will send you a snap-fit case for arduino.
Hey ! Thanks a lot for sharing this project, I successfully developped my first roll of super 8 film today, so I guess I'll have to do on my own variation of your scanner :)
What power supply / voltage did you use for the stepper motor and the LED ?
Thanks
You are most welcome. I found a variable power supply online, maybe via Amazon, where I can adjust a dial to set the voltage. For Arduino stuff, USB @5V is great. For stepper motors, typically, you’re looking at 12V but check the specs for your set-up and controller.
Thanks for doing this, Scott. Very nice work.I may have missed it, but I would like to know what, specifically, is your "Z-type" tripod head and rail slider, and where did you get it. I am working on a 16mm system with Eiki telecine projector that has a five-bladed shutter. I would like to see if I can get decent results in real time to make the audio component easier
Sorry - I hit enter by mistake. The 5-bladed shutter at 24 fps eliminates flicker from the camera at TV rate - 30 fps (okay, 29.97), and allows transfer with audio in real time. Thanks for any info you can provide on the head and slider. Thanks also for the ancient film clips of Vancouver. And welcome back.
@@morrisonproductions4588 Thanks! It's good to be back. ;)
Regarding hardware, I found both these things on Amazon. There are a variety of brands and prices, so I recommend you compare and make sure the models you pick are compatible with your camera mount. Also, you'll want to make sure the Z-type can handle the combined weight of your camera + lens + slider. I picked a Z-type that can handle 6 pounds, and it's been rock-solid.
It seems sellers make the product titles super long and stuffed with keywords, so I've truncated these to include just the relevant bits.
Rail slider: "Neewer 4 Way Macro Focusing Rail Slider ... for Canon Nikon Sigma SLR" (I'm using a Sony a7Riii, its tripod mount is compatible with this one.)
Z-type "mount": "Neewer Upgraded Z-Flex Tilt Head Z Type ... Aluminium Alloy/Load Up to 6.6 Pounds"
It can be frustrating setting up the Z-type at first, getting the height and level set about right. Once set, you can tighten the screws and shouldn't need to change it. Mine has a little built-in air-bubble level which is nice, but it's best to have the camera and light in the projector so you can see how it aligns with the actual film gate. I show some of this in "Part 2" - as mentioned, it's arguably the "least-scientific" part of the process. ;)
I also flipped the slider mount around, allowing me to get the lens closer to the projector film gate. I showed this briefly in Part 2 as well, but didn't explain fully. To remove the top "ruler" slider and flip it around, you just need to take two screws off a bracket on one end, and use the gear control to work the top slider out of the assembly, then put it in backwards. I should probably do a segment on this for Part 3.
Hope this helps!
@@scottschiller Thank you, Scott. I will check out these items and should have some results in a week or so
Great! It’ll be neat to see what you put together.
Thanks for sharing this. It is really useful and I'm going to do it based on a Eumig projector. What is the specification of your stepper motor? I'm not sure what size to go for.
Thanks, Roger. I believe it was a Nema 17 that I got. You will want to make sure you get a stepper motor controller or driver that is compatible.
I got something like this from Amazon Canada. The driver/controller is a TB6600. You may be able to find better prices separately, etc. www.amazon.ca/Stepper-Longruner-Connector-Printer-Mounting/dp/B07FK8NRKL
Lot of modifications, But - Great job done 👍
Does not a DSLR shutter get worn out, after so many pictures? I am using similar concept, but recording directly a video, but the synchronization is a problem. Because, let's say 15min movie (16 f.p.s.) presses the shutter 14 000 times, so I am afraid to use it this way. What is you experience please?
You are correct regarding still photos! Indeed, photos on a DSLR that involves mechanical actuation will contribute significantly to wear and tear. I mention that in my Q&A video, with a suggestion about exploring a true shutter/actuation-free approach via "Magic Lantern" software for the Canon 5D and similar DSLR series. ruclips.net/video/hthVoeRgbos/видео.html
It's worth noting the Magic Lantern "silent" mode may be a very slow process, and so I recommend using a mirrorless camera with "true" silent (no moving parts) mode if you can for this sort of work.
Great video and really awesome project. I am looking to do something similar with 16mm. Old projectors are expensive and you never know what you're going to get when it's used. Do you have any recommendations for a 16mm projector within a budget? Everything seems to be geared towards 8 unless you go vintage. Sending all the film I have to convert to digital is too expensive. I've worked with Arduinos and I write software for a living, the part that will trip me up is the camera and projector portion of this if I try to replicate. Cheers!
Thanks for the comment and questions; I've had similar thoughts about 16mm, myself. It seems like vintage *might* be the way to go, but I think it would be wrong to irreversibly modify a vintage 16mm projector for scanning. The challenge is if and how to adapt one to line up a camera and an LED light source, work in a stepper motor and so forth to drive it similar to the 8mm/super-8 scanner.
Alternately, cameras might be a safer bet and less expensive than projectors - they have a transport mechanism, gate and so forth. But, you would probably be hacking one apart just as much. e.g. (and, "general disclaimer", this is a wild guess.) www.ebay.com/itm/295775545842
My background is in software engineering, but I was able to figure out enough (and have enough Arduino experience) to get things working for 8mm/super-8 - so in theory, someone smart should be able to apply the same approach to 16mm and other formats. 😉If you end up taking on this project, let me know how it goes!
@@scottschiller thanks for this information. This is very helpful. I was actually thinking why do I need a projector? Could I just make a light box or something similar to illuminate the frame with the camera positioned above? I’d obviously need to figure how to feed the frames and wind and rewind the reels but an actual completely functioning projector doesn’t seem like a necessity now. I really liked your hack with the LED and ping pong ball. I do not want to burn the film. The main video on my channel is of 16mm film I sent away to be processed. The company did an amazing job but it cost me close to $700. I’m hoping this project will get me acceptable results and I won’t be spending a fortune on each film.
It's hard for me to pay $300 for a crappy, single-use product to digitize my old home movies that no one will watch. I'd rather spend $500 and build something myself. It's nice to see someone else tried first. Nice job.
Thank you, and I agree! I had a similar thought process: It seemed wise to invest in building something that would use a proper mirrorless camera (which I already had), and a macro lens - plus, I learned some things and had some fun along the way.
Very great job, really! But how long will the camera survive? At 24 fps you already have 1440 releases per minute, which means 86400 releases per hour.
Will the camera survive this?
8mm films run at 16 FPS, whereas Super-8 is 18 FPS. Either way, you're correct in that's a ton of shutter actuations and mechanical wear and tear potential unless you have full silent mode, e.g., on a mirrorless-style camera like the Sony a7riii I used. I speak a bit about this in a follow-up Q&A. ruclips.net/video/hthVoeRgbos/видео.html
What lens setup did you use?
I used a Laowa 25mm f/2.8 2.5-5X Macro lens - and, I bring good news: I just got Part 2 of this up, where I go a bit into the camera gear and set-up detail. ruclips.net/video/aIJ8S4B6GQw/видео.html
Utterly brilliant! And the original Star Wars on super 8 at the end was just icing on the cake. Thanks for sharing.
Do you really "need" to slow it down with the stepper that much? I guess the camera's frame rate is the limiting factor.
Thanks for the kind words.
The speed you can run at is dependent upon a few things. Common bottlenecks:
* How fast your camera can write pictures to storage (SD/CF or USB to computer) before the buffer fills up.
* Camera shutter speed / exposure settings, and minimum time between shots / how fast you can trigger the shutter.
From what I’ve found, my Sony a7riii may not always capture all frames when tethered and attempting to shoot more than 3 frames per second. This may be an arbitrary Sony limit or safety thing, I’m not sure.
I do think 3 FPS is stable when not connected to a computer, but then you need a fast and high capacity SDXC UHS-II type card, to handle a consistent data rate of 60+ MB/s or what have you.
Generally, I’ve found that “slow and steady” is good. I can do 1.5 FPS tethered quite reliably, without skipped or dropped frames and without having to stop the scan.
Hope this is of interest!
@@scottschiller Thanks for the explanation. I recently did something similiar to shoot time lapse of my 3D printer, but there I just use the lowest resolution JPG setting (2.5K) , and the Panasonic G81 only gets triggered 1x minute or less.
ruclips.net/video/FXXJQMq8srs/видео.html
I guess you are shooting raw to be able to grade the footage later, but must have enormous amounts of data from an A7RIII. Does an 8mm frame have more usable information than about 720P, or 1080 ?
@@RotarySMP That's quite a video! Fun stuff. I don't have a 3D printer etc., but I can see how they can be useful.
For your case, JPG should be absolutely fine. For my project, I'm pulling in a bit over 64 GB for ~3.5 minutes of 8mm/super-8 footage when shooting using RAW "compressed" on the Sony a7riii, and Super-35 / 1.6x crop mode. I end up with an export typically around 4K, 2160p. This may be overkill, but I'd rather oversample vs. later realizing my original scan(s) weren't capturing enough detail.
From what I've read, 8mm might be 720p and Super-8 film perhaps 1080p, ~2K. It's unclear exactly how film translates into "vertical lines" of resolution. Either way, I think "over-sampling", scanning Super-8 at 2160p / 4K is OK as it's increasingly the standard for TVs, computer screens and high-DPI mobile devices where this content will be watched.
For typical 8mm / super-8 film, of course, the film's quality, age and how it was shot at the time - focus, lighting and exposure - all factor into the amount of quality / detail you can pull out of a scan. I think I've had some pretty decent results at times, but I've also had film that's both under-exposed and over-exposed between indoor and outdoor scenes on the same reel.
The intent with the raw process is that I can try to adjust, enhance and balance details "in post" in a lossless way after scanning, and then export a final .MOV/MP4 with only one generation of lossy compression.
Hopefully this is of interest! I'm not an expert, and learning as I go - it's been a fun process.
@@scottschiller Thanks for the kind words, and the explanation of your process.
Have you considered developing the film yourself? I have a problem with the $40 for film + $30 Developing + $25 scanning for 2.5 minutes. So far I haven't made the plunge into analog filming. Did plenty of B& still photography and darkroom work back in the 90's. Seems 16mm film stock is now about the same price per mnute as Super 8. I have been reading up everything on line on home developing :)
Another time consuming rabbit hole to run down :)
Have you also shot, developed and scanned any modern Super 8 film. I have looked at a fair bit,and rarely see anything approaching the sharpness of the old train films you have posted.
@@RotarySMP I hear that, film is an expensive habit. My DIY scanner's initial goal was digitizing my own family 8mm/super-8 archives, which I've done. Since then, I've purchased a few other things from eBay.
I'm interested in 16mm as well, I think it could be fun to try scanning that at 4K and have source material that is better-matched to that resolution.
hey scott building my own telecine machine, I'm kinda stuck on the stepper motor part what's the power supply you used for your stepper motor driver? and also what's that part called that gets connected to the vcc and ground part that looks like a plug head? I've been looking everywhere for that part
Great to hear you’re building your own! It can be a fair bit of work, but the results are absolutely worth it - and ideally it’s fun, too. 😅
I’m using a TB6600 stepper motor driver here, which the Arduino connects to and provides the signalling. You can also get separate modules for stepper motors that just have a little heat sink and a potentiometer that controls speed, if you want to avoid using an arduino for the stepper motor portion.
I’m not clear on exactly which plug piece you’re referring to, but the green rail on the TB6600 is often called a “terminal block” - with screws that clamp down on the connecting wires.
@@scottschiller I found the part that I was looking for, it's actually called "DC barrel Jack Adapter Female". Now that the Arduino parts are done just going to have to figure out how to disassemble my projector, it's a Fujiscope and the lens are not detachable, looks like I'm gonna have to drill it out 😄
@@theroebuck123456789 Ah, got it! I wasn't sure which part you meant, glad you sorted it out. You can get little terminal screw block-type versions for those also, I've used a number of those in various projects.
Best of luck with your build - I know some of that stuff can be tricky. I wasn't sure, but I tried to pick a machine that looked like it had a detachable lens piece and I lucked out. The side of camera lens butts right up against the frame of the projector ever so slightly, but I also didn't want to carve into the body.
Scott, A question for you; hoping you have the time to answer.
What are your expectations for shutter life-expectancy of your camera? My understanding is that most camera's have a shutter-life of somewhere around 200,000 shots.
If I had ten hours of film to transfer, at 18 fps, that would add up to 648,000 shots.
I am in the process of pulling together the hardware to build a scanner but have hit this potential snag that is adding to the difficulty of locating a suitable camera.
Any other recommendations you could offer on camera and suitable lens would be appreciated.
Hi Mike,
For this sort of project, in my opinion, it's *crucial* that you have a camera that has a silent shooting mode or similar such that there is no mechanical movement when taking a photo. I.e., your camera is able to use an electronic or otherwise entirely non-mechanical process to get an image from the sensor. When you press the shutter, there should be no noise. If something moves, you don't want that happening almost 650,000 times. 😅
I'm using a Sony A7riii here with the silent shooting feature enabled. It's a mirrorless camera, so it has less mechanical stuff than a DSLR to begin with - this is a good thing. With full silent shooting on this camera, there is no sound when it takes a picture - nothing moves internally, as far as I'm aware. That was a requirement for me, I didn't want to have this project put any wear and tear on my primary camera. So far, I have had no problems taking pictures with the camera with my usual lenses. Nothing is guaranteed, but I believe the lack of mechanical activity is a good sign.
Using a stock Canon 5Diii DSLR as a different example, I believe taking a photo requires a mechanical shutter or at least some sort of motion internally for each shot. The 5d3 does have a "silent" mode, but there is still some sort of shutter or mechanical motion. It is reduced, but there is still movement. That's going to be a lot of mechanical wear and tear on the camera and will shorten its lifespan, indeed.
If you have a Canon and are willing to try, I've heard the third-party "Magic Lantern" and/or other custom firmwares via CHDK might offer a feature to get photos without mechanical actuation. Magic Lantern calls their feature "Full-resolution silent pictures." If you aren't familiar, ML is its whole own free/open-source project and can be fun - but you also need to be careful, there is a risk you can damage or "brick" your camera.
Alternately, I have heard of some folks using cameras made for the Raspberry Pi hardware platform and using it with some success for scanning film. I believe those are entirely electronic, no mechanical bits to break down. ;)
Hope this is helpful!
@@scottschiller Hi Scott. Another question; if you have time and experience with this sort of thing.
Do you own/use a mechanical 8mm film editor/viewer?
I'm in the market and have been scouring ebay. Many of my old family films have some bad spots that need attention before scanning. I see lot's of promising ones but cannot find much in the way of recommendations on some of these old machines.
Some particular ones I've got my eye on are Mansfield 650's, 950's. Also P&B-88 and Kalart KE-8.
@@mikebetts2046 Hi Mike,
I was originally looking at viewers / editors as a possible transport mechanism for scanning etc., before I decided to go with a modified projector.
If you're thinking of spending time reviewing and cleaning individual frames with IPA or similar, perhaps a viewer / editor makes sense. I think you'd want something like this to be able to move through the film and quickly find potential dirty frames, etc.
"BAIA" and Elmo seem to be popular, but they can get up to over $100. I'm not sure it's worth spending that much. Check that the viewer can do Super-8 if you're using that, as well.
e.g.
www.ebay.com/itm/VTG-Baia-Reviewer-8mm-Film-Movie-Editing-Editor-RE-8-120/284195418047
Hope this is useful!
For films under exposed in parts?
For me, this project is a number of things. Mostly, preserving family memories in digital. In addition, it’s nice being able to get “up to” 4K scans of the source, regardless of the film’s relative level of detail. Being able to adjust color and exposure digitally and “losslessly” is also a big plus. I also had fun and learned things building this, of course!
HOW CAN I BUY A NEW LAMP FOR THIS?
I recommend searching for the model name, and then “replacement lamp.” Verify on the listing / product page that your projector is listed as being compatible or supported.
Don't want to criticise this great project but you should have used better light source. Cheap LEDs have terrible color spectrum and low CRI (Color Rendering Index). I would bet that using some high quality LEDs with CRI 95 (can be found in LED bulbs if you specifically search for CRI 95 LED light) would make the colors much nicer straight out of the camera. The car LED would have CRI 70 at best...
This is a good point! I was not fully aware of the CRI stuff at the time, although I did look a bit later. These were "Luyed" 3014 33-EX Chipset-based LED lights, and I don't see any indication of high CRI. If I do some more scans in the future, I will look at a CRI 95+ LED source for sure.
This is only possible for you!
Perhaps you mean, this is a big project and a lot of work? 😅 It is, but I am also not an expert with hardware!
I had seen other peoples’ DIY scanners online and working, e.g., ruclips.net/video/50MSidTNwTI/видео.html and they inspired me to try my own.
I feel like there must be a far, far simpler way to achieve the same stepper-advancing effect..
In retrospect, my approach here could have been much better. 😛 I think because of the Arduino angle, I went down the TB6600 path and ended up using a second Arduino for the "status" OLED display because its I/O caused stepper motor stuttering.
I've seen others using a very small standalone module, a pot on a postage-stamp-sized PCB that does all the speed control for a stepper motor - much more efficient.
Why the flying fuck hasn't someone made one of these commercially that isn't a jank pile yet.
Hah! I hear ya, this is a tough one. 😅It seems the Wolverine Pro is a common choice, but it seems like a compromise for convenience. Build quality is low, you get a baked MJPEG video at 30fps(?), and jitter between frames is common. I think to get more reliable frame placement, you need hardware from a real projector ... and that's the top of that slippery slope.
BTW: Nice Django cover! I enjoyed that. ruclips.net/video/w4SynylXvzg/видео.html
@@scottschiller the real kicker is just no control over anything, at least not that I can see - exposure is adjustable but colour balance isn't, so I have some reels that I just can't trick the thing into recording properly. It hasn't been great with longer reels either, and I've heard about inner components wearing out or melting down haha.
Cheers about Django, the guy was an alien!
Agreed. I think there are a lot of trade-offs with the WP, and plastic parts seem to be a contributing factor. There is a so-called "Wolverine" mod that might be of interest, but I think the machine is still fairly limited. I've got much more control with the DIY route, but it has been a lot of work. At the same time, it's also been fun and I've learned as I go, so I consider it a win thus far!
And yes, Django is something else. ;)
cheese grater.
That, a ping-pong ball, and the paper binder metal bits. It's hackish for holding the LED light, but it works! 😂
Hi Scott, I don’t know how to start my own thread. Do you have any documentation on how you did all of this? Thanks! Lewis
Hi Lewis, thanks for the note.
I currently have two videos of this “multi-part” series, and I’m working on a “Part 3” right now. I also intend to have some documentation or notes that people can reference, as well.
I plan to publish the source code and schematics for the Arduino / stepper motor and related hardware that I’ve put together, as well.
Let me know if you have any specific questions in the meantime, I’m happy to elaborate.
@@scottschiller thank you Scott. I’ve enjoyed your first two videos and look forward to the third installment. Will you be providing this info below the video? I am trying to build my own scanner so that I can continue to acquire 16mm films and digitize them myself. I restore and colorize these films and the cost of a professional service is prohibitive. I appreciate your getting back to me.
Hi Lewis,
I plan to provide more info and links with “Part 3”, and perhaps it will end up as a five-part series. I’ve got a few updates to the way I scan, and I still need to document my approach to “remastering” frames in Lightroom. I will probably have another section entirely dedicated to post-processing and exporting the video from DaVinci Resolve.
What you’ve essentially created is a Kinescope
Ha! Thanks. Perhaps it's like a Kinescope in reverse? Instead of filming a TV or screen _on_ the Kinescope, I'm projecting from it and onto a digital sensor - i.e., the camera.
I had to look up the Kinescope. That's a really interesting use of film, to preserve TV programs before videotape became a thing.
@@scottschiller Yeah, pretty much. The kinescope is really cool though, you can see the youtube channel Free The Kinescopes to check some out
Scott Schiller...contact me by reply. I have a projector that I will pay you to help me modify for frame-by-frame scanning.
Thanks for the offer, but no payment needed. I can advise based on what I've done, but I can't take responsibility for projects like these. There are a lot of small details involved, but I'm trying to document as much as I can in this video series!
IT. IS NICE. AS. HOBBY AND TO. KILL. YOUR. TIME, BUT. BETTER. THAN. UNIVERSAL. SCANNER. IS. NOTHING. YOU. GIFTED. AND. TALENTED INVENTOR BETTER. UPDATE. WOLVERINE. PRO. AND MADE. IT. SPROCKETLESS. LIKE. UNIVERSAL. WITH. NORMAL. PRICE. AND. EARN. A. LOT. OF. MONEY AND. APLAUD. FROM FANS. AND. PROFESSIONALS. GOOD. LUCK. MY. FRIEND!