This one isn’t clickbait, but it is cheating. [Brian Haidet], the guy behind Alpha Phoenix, has managed to assemble movie footage of a laser beam crossing his garage, using a rig he put together for just a few hundred dollars. How, you ask? Well, for the long version, you’re going to want to watch the video, also embedded below. But we’ll give you the short version here.
Light travels about a foot in a nanosecond. What have you got that measures signals on a nanosecond scale pretty reliably? Of course, it’s your oscilloscope. The rest of [Brian]’s setup includes a laser that can pull off nanosecond pulses, a sensor with a nanosecond-ish rise time, and optics that collect the light over a very small field of view.
He then scans the effective “pinhole” across his garage, emitting a laser pulse and recording the brightness over time on the oscilloscope for each position. Repeating this many thousands of times and putting them all together relative to the beginning of each laser pulse results in a composite movie with the brightness at each location resolved accurately enough to watch the light beam fly. Or to watch different time-slices of thousands of beams fly, but as long as they’re all the same, there’s no real difference.
Of course, this isn’t simple. The laser driver needs to push many amps to get a fast enough rise time, and the only sensor that’s fast enough to not smear the signal is a photomultiplier tube. But persistence pays off, and the results are pretty incredible for something that you could actually do in your garage.
Photomultiplier tubes are pretty damn cool, and can not only detect very short light events, but also very weak ones, down to a single photon. Indeed, they’re cool enough that if you get yourself a few hundred thousand of them and put them in a dark place, you’re on your way to a neutrino detector.
“Femtophotography” is what you’re looking for…
This has been done over a decade ago
Femto photography of light moving and spreading through a soda bottle
Can even see the particle wave duality
https://youtu.be/EtsXgODHMWk?si=o6JwpDzu41gJerIR
You can even use it to see around a corner using a form of essentially raytracing and echolocation
Did you even glance at the video?
Per detector pixel the MIT solution was less expensive.
But not by much :-/
It’s always tough to hear people say “this PMT can detect a single photon, it’s amazing!”
That’s… not what makes it amazing. Your eye can detect single photons. It’s easy: it’s just a chemical reaction where one photon causes a state change.
The amazing part of a PMT is that it detects a single photon with almost no noise. Your eye doesn’t: that’s why your brain filters out signals less than “a couple” photons. Your eye’s still awesome, mind you: when we need to find light leaks in a PMT testing room we just went in with a couple people and shut the door and waited.
PMTs generally have noise – dark currents – of about 1 electron/nanosecond-ish (fraction of a nanoamp). But because that noise is mostly from thermal emissions later, it’s smaller than the signal from a photon, and you can clearly resolve the peak of a signal from a single photon.
Larger PMTs have more noise, and do worse, in general, at detecting single photons.
It’s the low noise that matters. PMTs are actually worse at capturing photons than, say, a CCD, which captures virtually all the light onto it. But they’re silent when nothing’s there, which is the real key.
If you were doing this with precision, you’d actually want to fit the peak signal you get for each shot, calculate the area, histogram it, and you could cut the noise even more.
Awesome comment – thank you !