Light captured at 100B frames per second | Hacker News Reader

Light captured at 100B frames per second

motherboard.vice.com

166 pointsdavidbarker27 comments

I find it interesting that these cameras and videos claim to capture photons as they travel through the air. Don't the photons still have to bounce off something (toward the camera) in order to be imaged by the camera? In other words, can we really "see" the photons/light in the various positions of its apparent journey, or are we still looking at a portion of the photons that have reflected off other particles (such as dust in the air) and made their way to the camera?

Even without dust in the air, you have things like Rayleigh scattering. But yes, we only "see" the photons that were deflected from their original path.

> I find it interesting that these cameras and videos claim to capture photons as they travel through the air.

Where do they claim that?

A camera needs light to hit a sensor. It can't "sense" the presence of a photon at some arbitrary point in space.

That being said, I just bought a 5 mW green laser pointer, which is quite fun to play with. If you point it so the beam is in front of a dark background, you can actually see the light ray. So if we filmed this laser pointer, using the camera in question, we should be able to literally see the light ray move towards the object it's pointed at.

That's about 105 years of a normal movie if you're going to watch one second of it completely.

https://news.ycombinator.com/item?id=8709920 is the same story, posted 42 days ago.

As is https://news.ycombinator.com/item?id=8700239 from 44 days ago.

but this submission did not get enough upvotes nor was on front page. Why so

Because success, both for startups and for HN submissions, is highly affected by chance.

So HN is bascicially a startup simulator for side projects?

I'm imagining now a system where users with lots of karma can "invest" in posts they think should be on the front page, incubator style. If they make it to the front page it's an "IPO" of sorts and they get a return on the other users votes.

Because whether a story makes it onto the front page is a total crap shoot. It depends entirely on whether anyone is looking at the new page in the 30-60 minutes immediately after you submit.

And also the story I posted 11 days ago.

> With CUP, the photons necessary to take an image are blasted through a beam splitter and then through a tube that has several tiny mirrors. These photons are converted into electrons, which encode the data you want captured—namely the time and space data necessary to create an image.

Can photons actually be converted to electrons? I've never heard of anything like this. Very interesting.

I'm a Nature subscriber; here's some information from the paper:

> To record events occurring at subnanosecond scale, currently the most effective approach is to use a streak camera, that is, an ultrafast photodetection system that transforms the temporal profile of a light signal into a spatial profile by shearing photoelectrons perpendicularly to their direction of travel with a time-varying voltage.

It references:

  Guide. to Streak Cameras http://www.hamamatsu.com/resources/pdf/sys/e_streakh.pdf (Hamamatsu Corp., Hamamatsu City, Japan, 2002)

From that PDF:

> The incident light on the photocathode is converted into a number of electrons proportional to the intensity of the light, so that these four optical pulses are converted sequentially into electrons. They then pass through a pair of accelerating electrodes, where they are accelerated and bombarded against a phosphor screen.

The paper does present a new method, but they are still using a streak camera:

> To overcome this limitation, here we present CUP (Fig. 1), which can provide 2D dynamic imaging using a streak camera without employing any mechanical or optical scanning mechanism with a single exposure.

---

I do not know anything about the topic, so I can't answer any questions. I was fascinated by the paper back in December when it was published---I'm curious to see eventual applications.

More accurately, these photons then excite electrons to a higher valence band and when those electrons fall back they give off a tiny electric current.

(Either that, or another photon! see: http://www.reddit.com/r/askscience/comments/141pip/how_do_mi... )

>Can photons actually be converted to electrons?

They can, but that's not what they actually mean here.

In this case, it would be more accurate to say that the energy of the photon is being used to excite an already existing electron.

As for creating electrons from photons: if you collide two photons at the correct energy level, you can get an electron-positron pair. This is a very high-energy process, and not how cameras work.

They can ( or more precisely into positron electron pairs), but only at much higher energies ( so for example gamma rays). Most likely here some effect, for example photo-electric effect, where you can reconstruct the original photon from the electron, so that you only transfer the information of the photon.

One way is via Photomultiplier http://en.wikipedia.org/wiki/Photomultiplier essentially the photoelectric effect is used to create a cascade of electrons, the created current can be measured and related to the energy of the original photon (which is h * f, where h is Planck's constant and f = c/lambda is the light frequency.)

I've possibly missed the point of what you're asking, but... surely it's obvious that photons can be converted to electrons, given that's what a photodiode does and that's how digital cameras work?

>surely it's obvious that photons can be converted to electrons, given that's what a photodiode does and that's how digital cameras work?

This is a common misconception. Both of those use light to facilitate electron excitation, but neither of them actually convert photons into electrons. The electrons are already there.

Ah, so that's how I missed the point of what he was saying. I know how photodiodes work -- I was just struggling with the semantics of "conversion".

The photons are probably actually being converted into electric current, as a photodiode, photomultiplier, or solar panel would do.

If I remember correctly, these devices are capturing multiple photons over a similar path; not a single photon's journey.

every day feels more closer and closer to everything getting insanely cool all at once

Is the blotchiness due to uncertainty principle or sensor imperfections? Both maybe?

That is energy vapour surrounding individual photons.

poor statistics, I would guess.

when you're shooting that fast, there are few photons per frame.

Reply on news.ycombinator.com