Really enjoyed this video! I quite like the solution to the focusing you're coming to as, with a lot of the other bare sensor in the focal plane videos, it may actually work when using the zoom focus and maybe even the lens distance guides. Can't wait for the next part when you get to it! I'm particularly interested in your idea for using the actual motor drive in the camera to trigger the recording as this is something that's often skipped and really makes the "product" in the other videos feel like a retrofit!
Thanks! More work coming up on this, for sure. I will be experimenting also with a photodiode to sense when the shutter is open. This may allow me to synchronise as well as start and stop recording.
@@jennylist Now that would be super cool! It'd allow you to use the shutter speed selections (if your super 8 has em) and get the feel of "press the shutter button to film, stop pressing to stop" as opposed to what I've been seeing "open the side, press the power button, press the record button, shut the side and go" ! Can't wait for the next part whenever it's ready!
loving this project! I feel like perhaps using a macro lens reversed might be something to explore... I don't know if you've seen the "focal reducer" (aka "speedbooster") devices for video cameras that have been released the last few years. But, they seem to essentially be doing what you are proposing -- for a different use case slash imaging pipeline. My understanding is that these "focal reducers" are essentially macro lenses in reverse. I'm probably not explaining this well or 100% accurately, but, hopefully it gives you some stuff to look into if you are interested or if it sounds interesting. also, on professional video cameras there is (maybe was) a thing called "back focus" which was a necessary process to do minute adjustments in order to get the lens and sensor to focus correctly -- otherwise your parafocal lens would not hold focus. I just mention this to suggest that perhaps this kind of process or some ability to make small adjustments to the sensor position might be necessary occasionally. Like, it might be helpful to imagine this as part of the maintenance of the setup and allow for that kind of tweaking. excited to see where this goes!
I'm familiar with back focus from the Raspberry Pi HD camera, and it's a PITA. I am hoping that it's taken care of in the camera's optics, in that they focus on the focal plane where the film sits. If I focus on that point, I should be able to become an equivalent of the film.
Send in your video I tell myself that my Super 8 cameras will soon come back to life and I am very happy about it could you explain to us how you intend to distribute the product once finished in kit form ? Anyway have a great day your videos are great.
Good question, and the answer depends on many things. I'm not in the kit business. I have been in the past, I used to sell a range of little radio kits but Hackaday took up my time so I stopped. I'm not sure I want to get in the kit business again, it's not an easy one to be in because you also take on a significant support burden. Also, as you can see this project isn't quite there yet. If you look you'll have seen that quite a few others have tried, and not really come up with fully workable results, it's a hellishly difficult thing to get right, let alone put into production. So I'm not certain whether or not I'd ever sell these, or even if I'd end up with something saleable. It will in the end be possible for people to replicate most of what I've done and have a go at making one of their own, of that I'm certain.
People have used the Sony IMX bare sensor C-Mount PI modules for making film scanners like the degusse roller, I think using more approprate sized sensor in this inverted sence would yield great results as you could get a native focal depth out of the fixed lens on the body. Though global shutter sensors are more suited with the mechnical shutter nature of the camaras unless your stripping it out, but to get the flocal plane right you would need to put the sensor on a slider internally really but I dont think there needs to be more then 3mm of play based off the position of a super8 cartrige film position, then again just off the top of my head thoughts the small injection molded lenses on those probe/webcam style sensor modules are utterly terrible for this application but there is good micro lenses on the market if you just need to multiply the focal ratio, but thats the main issue I see with sensors smaller then the intended image area.
Surprisingly, that sensor is about the same size as a super 8 frame. The lens however is not very good, you're right there. Keep watching, I have some ideas I want to try there.
@@jennylist I've got 8mm and Super8 camaras so I will be, someone on the DomesDay86 community posted a link to this and we do raw FM RF capture and decoding for tape/laserdisc and even CVBS and then can DAC even pipe it back to analog, so its cool to see stuff being done with legacy film kit as I just picked up a new 8mm body this week its really nice to see 100ft and 25ft rolls are still around though.
@@jennylist Yeah the last hackaday article was really not well done, quite franky ld side and tape side are seprate but united under the core concept of FM RF capture which any publication should focus on rather then pushing a now 3 years out of date absolute edge case example of the post processing of the RF data as the first video people see... Really that whole article should have been fixed but no one updated it even though there editorial staff was provided with the content to do so, it angers me to this day as the wiki writer as its borderline confusing to people getting into whats the end all methods of digtisation, which alongside the Lordsmurf's missinformation campain becouse we put a foot down on his acts of fraud and the inflation trap DigitalFAQ has become a stronghold for SEO score still has a long way to go as more external content is made.
Was it just an accident that your camera's frame rate synchronized with the shutter? It seems like you could just have easily found yourself synched with the closure of the shutter and got a black image. Or anything in between! Very clever solution, though!
I've done a lot of experiments and I think it's an inevitable side effect of filming through a shutter. But one next step once I have it battery powered is to try different cameras. Right now the Minolta is the only one I can get the Raspberry Pi cables into the cartridge.
I wish I could do a Single 8 version too, however it's not possible due to the film gate in a Single 8 camera. You probably saw a teaser of my Single 8 cartridge project at the end of the video though, that's my bit to make the format no longer dead. A new video with Single 8 updates coming soon, however you can see past progress in my earlier videos.
Shouldn't the 18fps frame grabbing of the Pi be synced to the physical camera shutter? When out of sync you could get frames where the shutter is partly or even entirely closed.
I guess that your solution may need to be set up individually for any camera lens since you kept the lens of the camera module. It may be possible to remove the lens of the camera module. Any pice of optics added to the super 8 camera will alter the image and may also change the focal range, the field of view and may cause trouble with zoom lenses. If you want to try a bare image sensor in the cartridge you will need to set up it's distance to the super 8 lens very accurately to obtain focus
The idea is that it has to focus on the focal plane of the camera, which is the same relative to the cartridge for all cameras. Thus once it is set up properly for one it will work on all. Take a look at the previous videos in the series and you'll see why a bare image sensor didn't work. It's extremely difficult to get it in the focal plane, because of the thickness of its glass front.
That looks fantastic :D Instead of recording as a video file, could you use a spindle and something like a Hall effect sensor to take a single image such as a jpg and then process it into a video after? This might eliminate the shutter issue? I guess it potentially improve the video quality too. With regards focusing, take a look at a video called building a next level camera by DIY Perks. Note: I know nothing and understand even less so please bear this in mind with regard my suggestion :D
this makes very limited sense, but some kind of sense would make if you would at least add a rear projektion screen material between the super8 camera lens and your camera lens in the cartridge, which has some physical grain, and ideally not only one screen but a rotating set of different frames with different grain, e.g. an endless loop (agitated by the camera as normal super 8 film) of blank/white film. but still, it lacks the most important aspect of real film, the colour response curves, especially in conjunction with the subtractive process. unfortunately, there is no know "filter" material which emulates that physically. or does something like that exist?
Watch the previous video, I looked into the projection screen idea. Colour response is a software thing on this. As you can see this is an early prototype, and there's a *lot* more work to be done here.
@@jennylist ok, i will watch. colour response emulation in postprocessing doesnt work, and cant work, what is messed up and missed by the sensor can not be repaired and invented from nothing or by guessing afterwards in a way which comes close enough to the real thing, maybe someday with ai, but not with any available software tools, not even 300k$ color grading toolsets used by the largest studios worldwide, it just doesnt work this way. there might be physical ways before the sensor, photochromic materials etc, but not afterwards. no chance! you cant fool the laws of physics
It needs a way to sense the mechanical shutter speed and synchronize to it. This one does that, it also has the camera lens removed completely, with the sensor placed at the original film plane. ruclips.net/video/n1QlHnQutIg/видео.html
Patience, I'm working on it. Watch earlier videos to see why putting the sensor in the focal plane doesn't work. You'll find that many such designs remove the film gate.
Hi Jenny. We found your videos so rewarding, and were wondering if any prototype services may help? Such as custom PCBs, 3DP, CNC etc. If so, would love to supply freely and reach any YT collab together! (PCBWay luz)
Hi Zoey, I'm very much aware of PCBWay and familiar with your offering. If you look at my header image you'll see some conference badges which have PCBWay boards. This project doesn't have a PCB but I will bear your offer in mind.
Really enjoyed this video! I quite like the solution to the focusing you're coming to as, with a lot of the other bare sensor in the focal plane videos, it may actually work when using the zoom focus and maybe even the lens distance guides.
Can't wait for the next part when you get to it! I'm particularly interested in your idea for using the actual motor drive in the camera to trigger the recording as this is something that's often skipped and really makes the "product" in the other videos feel like a retrofit!
Thanks!
More work coming up on this, for sure.
I will be experimenting also with a photodiode to sense when the shutter is open. This may allow me to synchronise as well as start and stop recording.
@@jennylist Now that would be super cool! It'd allow you to use the shutter speed selections (if your super 8 has em) and get the feel of "press the shutter button to film, stop pressing to stop" as opposed to what I've been seeing "open the side, press the power button, press the record button, shut the side and go" !
Can't wait for the next part whenever it's ready!
loving this project!
I feel like perhaps using a macro lens reversed might be something to explore... I don't know if you've seen the "focal reducer" (aka "speedbooster") devices for video cameras that have been released the last few years. But, they seem to essentially be doing what you are proposing -- for a different use case slash imaging pipeline. My understanding is that these "focal reducers" are essentially macro lenses in reverse. I'm probably not explaining this well or 100% accurately, but, hopefully it gives you some stuff to look into if you are interested or if it sounds interesting.
also, on professional video cameras there is (maybe was) a thing called "back focus" which was a necessary process to do minute adjustments in order to get the lens and sensor to focus correctly -- otherwise your parafocal lens would not hold focus. I just mention this to suggest that perhaps this kind of process or some ability to make small adjustments to the sensor position might be necessary occasionally. Like, it might be helpful to imagine this as part of the maintenance of the setup and allow for that kind of tweaking.
excited to see where this goes!
I'm familiar with back focus from the Raspberry Pi HD camera, and it's a PITA. I am hoping that it's taken care of in the camera's optics, in that they focus on the focal plane where the film sits. If I focus on that point, I should be able to become an equivalent of the film.
This idea es just amazing!
Thanks :)
Send in your video I tell myself that my Super 8 cameras will soon come back to life and I am very happy about it could you explain to us how you intend to distribute the product once finished in kit form ?
Anyway have a great day your videos are great.
Good question, and the answer depends on many things.
I'm not in the kit business. I have been in the past, I used to sell a range of little radio kits but Hackaday took up my time so I stopped.
I'm not sure I want to get in the kit business again, it's not an easy one to be in because you also take on a significant support burden.
Also, as you can see this project isn't quite there yet. If you look you'll have seen that quite a few others have tried, and not really come up with fully workable results, it's a hellishly difficult thing to get right, let alone put into production.
So I'm not certain whether or not I'd ever sell these, or even if I'd end up with something saleable.
It will in the end be possible for people to replicate most of what I've done and have a go at making one of their own, of that I'm certain.
People have used the Sony IMX bare sensor C-Mount PI modules for making film scanners like the degusse roller, I think using more approprate sized sensor in this inverted sence would yield great results as you could get a native focal depth out of the fixed lens on the body.
Though global shutter sensors are more suited with the mechnical shutter nature of the camaras unless your stripping it out, but to get the flocal plane right you would need to put the sensor on a slider internally really but I dont think there needs to be more then 3mm of play based off the position of a super8 cartrige film position, then again just off the top of my head thoughts the small injection molded lenses on those probe/webcam style sensor modules are utterly terrible for this application but there is good micro lenses on the market if you just need to multiply the focal ratio, but thats the main issue I see with sensors smaller then the intended image area.
Surprisingly, that sensor is about the same size as a super 8 frame.
The lens however is not very good, you're right there. Keep watching, I have some ideas I want to try there.
@@jennylist I've got 8mm and Super8 camaras so I will be, someone on the DomesDay86 community posted a link to this and we do raw FM RF capture and decoding for tape/laserdisc and even CVBS and then can DAC even pipe it back to analog, so its cool to see stuff being done with legacy film kit as I just picked up a new 8mm body this week its really nice to see 100ft and 25ft rolls are still around though.
@@TheRealHarrypm I think we covered the raw digitisation of laserdisc and VHS for Hackaday. Very cool stuff indeed.
@@jennylist Yeah the last hackaday article was really not well done, quite franky ld side and tape side are seprate but united under the core concept of FM RF capture which any publication should focus on rather then pushing a now 3 years out of date absolute edge case example of the post processing of the RF data as the first video people see...
Really that whole article should have been fixed but no one updated it even though there editorial staff was provided with the content to do so, it angers me to this day as the wiki writer as its borderline confusing to people getting into whats the end all methods of digtisation, which alongside the Lordsmurf's missinformation campain becouse we put a foot down on his acts of fraud and the inflation trap DigitalFAQ has become a stronghold for SEO score still has a long way to go as more external content is made.
@@TheRealHarrypm ok. Do you have an updated recent write up of technical progress? Submit it to the Hackaday tips line.
Was it just an accident that your camera's frame rate synchronized with the shutter? It seems like you could just have easily found yourself synched with the closure of the shutter and got a black image. Or anything in between! Very clever solution, though!
I've done a lot of experiments and I think it's an inevitable side effect of filming through a shutter.
But one next step once I have it battery powered is to try different cameras. Right now the Minolta is the only one I can get the Raspberry Pi cables into the cartridge.
That’s sick! If you get it dialed in and to a quality you feel is good enough I’d love to talk about making the same for the dead single 8 format!
I wish I could do a Single 8 version too, however it's not possible due to the film gate in a Single 8 camera. You probably saw a teaser of my Single 8 cartridge project at the end of the video though, that's my bit to make the format no longer dead. A new video with Single 8 updates coming soon, however you can see past progress in my earlier videos.
Shouldn't the 18fps frame grabbing of the Pi be synced to the physical camera shutter? When out of sync you could get frames where the shutter is partly or even entirely closed.
I find once the software is set to grab at 18fps that it does a pretty good job of its own. But I will be experimenting with a photodiode sensor.
Just discovered your channel! Subbed! Wishing you the best! Take care! 😎🤘☮
Thatns!
I guess that your solution may need to be set up individually for any camera lens since you kept the lens of the camera module.
It may be possible to remove the lens of the camera module.
Any pice of optics added to the super 8 camera will alter the image and may also change the focal range, the field of view and may cause trouble with zoom lenses.
If you want to try a bare image sensor in the cartridge you will need to set up it's distance to the super 8 lens very accurately to obtain focus
The idea is that it has to focus on the focal plane of the camera, which is the same relative to the cartridge for all cameras. Thus once it is set up properly for one it will work on all.
Take a look at the previous videos in the series and you'll see why a bare image sensor didn't work. It's extremely difficult to get it in the focal plane, because of the thickness of its glass front.
This seems totally mad but, could a blank DVD be converted into a reel?
i don't quite understand the context. Heat formed into a cartridge shape?
@@jennylist I don't think I understand the context, either.
That looks fantastic :D
Instead of recording as a video file, could you use a spindle and something like a Hall effect sensor to take a single image such as a jpg and then process it into a video after? This might eliminate the shutter issue? I guess it potentially improve the video quality too.
With regards focusing, take a look at a video called building a next level camera by DIY Perks.
Note: I know nothing and understand even less so please bear this in mind with regard my suggestion :D
I could indeed do that. But I think I can eliminate the shutter artifact.
this makes very limited sense, but some kind of sense would make if you would at least add a rear projektion screen material between the super8 camera lens and your camera lens in the cartridge, which has some physical grain, and ideally not only one screen but a rotating set of different frames with different grain, e.g. an endless loop (agitated by the camera as normal super 8 film) of blank/white film. but still, it lacks the most important aspect of real film, the colour response curves, especially in conjunction with the subtractive process. unfortunately, there is no know "filter" material which emulates that physically. or does something like that exist?
Watch the previous video, I looked into the projection screen idea.
Colour response is a software thing on this. As you can see this is an early prototype, and there's a *lot* more work to be done here.
@@jennylist ok, i will watch. colour response emulation in postprocessing doesnt work, and cant work, what is messed up and missed by the sensor can not be repaired and invented from nothing or by guessing afterwards in a way which comes close enough to the real thing, maybe someday with ai, but not with any available software tools, not even 300k$ color grading toolsets used by the largest studios worldwide, it just doesnt work this way. there might be physical ways before the sensor, photochromic materials etc, but not afterwards. no chance! you cant fool the laws of physics
its rare to see someone advocating for transgender rights in the retro tech/hobbyelectronics community :)
Happy to oblige.
It needs a way to sense the mechanical shutter speed and synchronize to it. This one does that, it also has the camera lens removed completely, with the sensor placed at the original film plane. ruclips.net/video/n1QlHnQutIg/видео.html
Patience, I'm working on it. Watch earlier videos to see why putting the sensor in the focal plane doesn't work. You'll find that many such designs remove the film gate.
Hi Jenny. We found your videos so rewarding, and were wondering if any prototype services may help? Such as custom PCBs, 3DP, CNC etc. If so, would love to supply freely and reach any YT collab together! (PCBWay luz)
Fully aware of who you are, no worries.
Bravo bravo bravo
Thanks :)
Nice content! Not sure if any custom PCBs may help? Would love to supply free ones and reach any YT collab together if possible. (PCBWay zoey)
Hi Zoey, I'm very much aware of PCBWay and familiar with your offering. If you look at my header image you'll see some conference badges which have PCBWay boards. This project doesn't have a PCB but I will bear your offer in mind.