I Built a Mechanical Digital Camera With a Nipkow Disk
Вставка
- Опубліковано 31 лип 2021
- See how I built a mechanical camera using an old record that I turned into a Nipkow disk.
The principle was already patented in the year 1884 by Paul Nipkow and realized 40 years later by John Logie Baird. Now I reinvented it, using a state-of-the-art stepper motor and a Raspberry Pi. Although the images are blurry, distorted and black and white, you can neverteless recognize what´s on them.
Camera: amzn.to/3QiWzmf
Consider to support this channel on Patreon and get updates about ongoing projects and preview videos!
/ docvolt
Buy me a coffee: www.buymeacoffee.com/docvolt
Paypal: www.paypal.com/paypalme/docvolt - Наука та технологія
Best conclusion ever!
at the last of video
does your produced photos the black and white
or the colored ? i don't understand
You can understand very well how the Nipkow Disc works. And your enthusiasm.
Hey hi! Just wanted to say that I teach engineering students, and will be sharing this as an illustration of how Nipkow disks work (in the context of confocal scanning microscopes) with my students. They will love this! Thanks for sharing this.
Nice to read this. Hope they enjoy.
@@DoctorVolt Indeed they will!
It would be interesting to see that made with an infrared temperature sensor. you could do thermal imaging.
Dope.
Love the tuxedo kitty's expression. He's like "Dad! The resolution of that thing will suck! At least use a vidicon!"
Sure, the images are hard to make out, but the teapot at least looks pretty good!
Great Work!!!
5:16 FREDERICK FREDDY FAZBEAR!!???!!??
Awesome!!! Great work
I love your assessment :D
Super cool!
nice work man!
If you redrill the holes to take (via a push fit) optical quality acrylic balls (making it a so called bead disk) you will solve your light and resolution issues. Interesting concept (still camera).
Yes, there are many other ways to make this camera better. For example drilling more holes, using a photomultiplier instead of the photodiode etc.
I love the cat ;)
Great video! I am not educated in such things but I still felt like the info was easy to understand. The big picture at least. You know, you ruined that poor record though! Joking. Cheers from South Texas
Use a photoresistor it can be thousand times mire sensitive then photo diode, also you can try with photon multiplier tube from old night vision device, that can increase sensitiviti to mental high level. Good luck !
I already thought of using an avalanche diode. But These are rather expensive, though.
@@DoctorVoltwhy not LDR ? Or get broken drum scanner brand like "screen" or "linotype" there are using photon multiplier sensor... and scan using same principle like your device but extremely sensitive. Old night vision googles are 20$ in thrift stores, or broken on ebay. Cheers M
@@marcsmithsonian9773 I tried LDRs as well. These have a bigger surface than photo diodes and thus can collect more light. But it turned out that they react very sluggishly on poor light condition. So image quality with photo diodes was way better than with LDR.
Shouldve gone down the flying spot scanner route
Very cool. Do you have any idea how you could make a sensor like this to identify color?
Yes, with a color sensor instead of the plain photo diode we could even take color photos.
now feed the output into one of the AI denoising/upscaling algorithms for a 4k picture xD
Did you think of using an LDR rather than a photodiode? And an analogue amp front-end to make it more sensitive and perhaps give a better range, by mapping values to something more linear, to compensate against the non-linearity of your sensor? If you switch to an LDR you might not need to re-map, they may be more linear in a certain range. Baird managed live video with whatever awful light sensor he had back then, some sort of vacuum tube, possibly using selenium.You also might try using R,G,B leds, pulsing each one in turn for each pixel as the Nipkow disc moves round, assuming your sensor is fast enough. With enough light, it'll likely be quicker in response. Then you'd have colour! Alternatively, have 3 LDRs with a colour filter in front of each one.
In fact a larger LDR might be more sensitive, too, or perhaps use a couple in parallel. Whatever, you'd just need to focus to a wider spot at that point. It would be dimmer, but you fix that with strong lighting, the same way they did back in the early days of TV.
Really, Baird gets far too much credit, when his invention is just applying a Nipkow disc to it's intended purpose. Nipkow deserves more than half the credit for Baird's television. It's the working element, nothing would work without it. Ultimately though it was Farnsworth who made television that was actually practical, Baird's idea didn't last long, before the world dumped mechanically-scanned TV altogether.
Most of the records I owned made weird noises but no music. ;)
maybe some filter would reduce the noise to enable more fps? If recorded objects don't move too fast, the signal will repeat with the same frequency that frames repeat, while I think noise frequency distribution won't depend on it. Also you can cover some edge to get repeatedly pure noise to measure it's parameters.
I already tried different filters, iterations etc. in GIMP. But nothing really improved the quality of the pictures.
I can't believe the photodiode is so slow. What was the original sensor type those days?
The photodiode is so slow because it has to make do with very little light.
maybe if on the nip disk
is make quite small openings somewhere around 400 it is possible to obtain a sufficiently sharp image?
Yes, I know that there were Nipkow disks with up to 441 lines resolution, until the electronical cameras came up.
@@DoctorVolt do you have a link or a photo of this 441 lines nipkow disk ?
I read about it in an article. second.wiki/wiki/filmabtaster
@@DoctorVolt thanks
replace a diode with a pyrometer and your optics with reflectors and it could be a nice thermal-vision system... ;)
No, it will always show temperature of the disc
Nix bad. 100 ano alt tec. Ja Mir frund. Danke zier. Some times used to xmit "wanted" photo!
when you can’t afford a webcam:
I from paris
If you put a PIR sensor in place of the photodiode you could see IR....imagine if you put an 5GHz antenna in place (and modded your setup a tiny bit :-p ) you could see in 5GHz
But I'd have to focus the IR or 5GHz image on the nipkow disk somehow.
Hallo. Kommst du aus Ostdeutschland?
Nein...
@@DoctorVolt ich dachte wegen der Schallplatte.
Stimmt. Habe sie mir vor langer, langer Zeit in Ostberlin gekauft.
Speak french! Please
Amazing