As the art of film photography has gained once more in popularity, some of the accessories from a previous age have been reinvented, as is the case with [tdsepsilon]’s radar rangefinder. Photographers who specialized in up-close-and-personal street photography in the mid-20th century faced the problem of how to focus their cameras. The first single-lens reflex cameras (SLRs) were rare and expensive beasts, so for most this meant a mechanical rangefinder either clipped to the accessory shoe, or if you were lucky, built into the camera.
The modern equivalent uses an inexpensive 24 GHz radar module coupled to an ESP32 board with an OLED display, and fits in a rather neat 3D printed enclosure that sits again in the accessory shoe. It has a 3 meter range perfect for the street photographer, and the distance can easily be read out and dialed in on the lens barrel.
Whenever the revival of film photography is discussed, it’s inevitable that someone will ask why, and point to the futility of using silver halides in a digital age. It’s projects like this one which answer that question, with second-hand SLRs being cheap and plentiful you might ask why use a manual rangefinder over one of them, but the answer lies in the fun of using one to get the perfect shot. Try it, you’ll enjoy it!
Some of us have been known to dabble in film photography, too.
Thanks [Joyce] for the tip.
This is a cool project and offers the additional ability to scan a 3D surface data set by moving the sensor as people have done with the ultrasound modules.
For photographic use, one of the foreseeable problems is that of just which objects have been detected. Detection is easy, discrimination is hard. Current digital imagine devices (most of them in phones, much to the dismay of the dwindling ranks of camera geeks) have sophisticated algorithms to detect and hold focus on a particular object (a detected face for instance) or a selected area in the image displayed. This mimics earlier optical systems with a center focus spot to work with.
The wide field of view of the radar sensor (43° W 116° V) makes this problematic if you’re trying to do something other than detect the closest object and it will be interesting to see how it all, um, pan out in practice.
Oh, I thought this was going to be about taking photographs with radar, which would be good for me as I look better in the invisible wavelengths.
At 24 GHz the wavelength would be ~1.25 cm (~0.49″), so the 3D resolution would be of a similar order of magnitude. You could probably mount a sensor on two stepper motors and generate a hemisphere, or a bit more spherical, scan of a room.
Collision avoidance radar modules for commuter vehicles are now COTS bolt on hardware. I wonder how difficult it would be to turn a few into “radar cameras”. A quick search turns up articles about units operating in just about every band between 1 and 85 GHz.
That kind of spread should give ample room to find the best combination of frequencies for imaging.
I say all of this having only used marine and aviation search radar for the Coast Guard. As console operators we didn’t too far down in the weeds on radar and radio propagation theory. It was mostly about playing with the gain and declutter settings until we got a usable image.
I did something similar using a satellite dish and LNA. That’s at 10GHz-12GHz. It doesn’t transmit – it only picks up what is present. You’d be surprised how “bright” everyday objects ar at 10GHz.
https://josepheoff.github.io/posts/rfcamera
That’s really cool! Thanks for sharing.
Where’s the beef? Shalom
Two things:
1- 3 meters In street photography is not useful as most pictures are taken between 3 a 5 m. Distance (using 50mm on a full frame).
2- best results are done using the biometrics hardware ( eyes+brain) that every one is provided with since is borne.