Hard Light Productions Forums
Off-Topic Discussion => General Discussion => Topic started by: Klaustrophobia on December 13, 2011, 07:09:18 pm
-
1 TRILLION frames per second camera capable of observing moving photons. yeah.
http://chime.in/user/Bill/chime/83832090542510080 (http://chime.in/user/Bill/chime/83832090542510080)
-
Oh man, I saw this earlier today. It's nuts. How does that even WORK?
EDIT: Oh, I see now... sort of.
-
...did you look at the video? Cos they kinda tell you how.
-
:warp:
-
Huh. I think the video and caption must be misleading, because to "see" a photon, it would actually have to hit the camera. But instead, photons are portrayed as particles emitting light. But... light is photons.
:confused:
-
However. . .
YOU CAN SEE LIGHT MOVE O_O
EDIT
For context, if there was a travelling bullet in the scene, it would appear to be still.
-
Yeah, you don't see photons move past the camera, but you can see how it bounces off different surfaces at different times, and that's ****ing awesome.
-
I think camera's are going to be a lot more different, and powerful, in a a few years. Did you guys see this one?
http://www.lytro.com/living-pictures/179/175
I wonder if they would do away with blue screens. You could make your film, and remove everything from a certain distance.
-
Huh. I think the video and caption must be misleading, because to "see" a photon, it would actually have to hit the camera. But instead, photons are portrayed as particles emitting light. But... light is photons.
:confused:
Put another way, if you were to direct a laser through a light fog, with this camera you could visibly see the beam extending.
-
Put another way, if you were to direct a laser through a light fog, with this camera you could visibly see the beam extending.
If I'm seeing/understanding this correctly, the scene shown at 1:45 through 1:58 is basically exactly that, except using an (I assume liquid filled) coke bottle instead of fog.
Edit: OHHH SHIIIT, what would it look like if they used this camera on the double-slit experiment? :eek:
-
This is too awesome
-
Link forwarded. Work filter prevents me watching it now. Lunchtime though :yes:
Thank god my X10 has Flash player 11 ;7
-
Edit: OHHH SHIIIT, what would it look like if they used this camera on the double-slit experiment? :eek:
You broke my brain. Congratulations.
-
I don't think this would change anything (though I'm not denying this would be an interesting experiment), as the beam is normally invisible. "Observation" of a photon is something different than observing the beam move. If you can see the beam at all, you're already "observing" photons in quantum sense. Of course, that's a simplification. HT will most likely come up with a more precise explanation.
Now that I think of it, I have to try doing the double-slit experiment and spraying something in the path of the laser. Though I might have to get a better laser, since the cheap pointer I have might not get through anything that would make the beam visible.
-
Thinking about this a bit more, if one were to use this camera setup with a double-slit experiment, you'd see the diffraction pattern show up on the surface, starting off directly behind the slit (shortest travel time for the photons) and spreading outward. If you filled the intermediate space with fog then you might be able to watch the 3D wave fronts spreading out from the slits, which would be pretty epic.
Unfortunately the camera would be useless for observing the particle nature of the photons in the experiment, even if it were sensitive enough to detect a single photon strike. First off the camera only records one slice of the scene at a time (thus the scene must be illuminated the exact same way on multiple runs, which fails for individual photons traveling through a slit).
Secondly, as Dragon points out, it wouldn't be able to measure the path that an individual photon took to make the diffraction pattern. It would only tell us the last thing that the photon was reflected off of before reaching the detector. If the photon was last reflected from the surface behind the slits (where the 2D interference pattern shows up in the classic experiment) then we have no knowledge of which slit the photon passed through. If the photon was last reflected off an air or fog particle, then we have no knowledge of where the photon hit the surface to build up the interference pattern, or even if it ever landed on it at all!
That all said, I'd still love to see it done in a thin fog-filled apparatus, just to be able to watch the wave fronts spread. :)
-
The guy that owns Playboy just ordered 69 of these. :pimp:
-
However. . .
YOU CAN SEE LIGHT MOVE O_O
EDIT
For context, if there was a travelling bullet in the scene, it would appear to be still.
Someone needs to combine these two things. NOW. :eek:
-
It's the worlds fastest slowest camera apparently.
Not sure if it could capture a bullet.
-
I think most of you are misunderstanding the video... that, or I am. :p
If I understand things correctly, this camera (array) can only effectively photograph physically unchanging scenes, because it does so over a period of time. The guy explaining it mentioned an array of 500 cameras, but more importantly, he showed how the camera captures a one-dimensional "scan-line" of the scene each time, necessitating the rotation of the mirror slightly after every exposure in order to capture the entire scene.
Combine all that with their laser that sends short pulses of light, and what you have is a camera that simply captures one "scan-line" of each laser pulse. Send 1080 laser pulses with the mirror angle adjusted between each one, and you have a single full-HD frame of a laser pulse. Repeat that 30 times, with the timing between the laser pulse and the camera's exposure lengthened each time, and you have a 1-second-long full-HD "video" of a laser pulse.
In short, it's not a hi-speed camera per se - it's an incredible combination of exact repeatability, very precise camera-laser syncronization, and mirror control.
Note: I'm not sure where the array of 500 cameras comes in... possibly they're chained together to all take exposures of the same scene over the course of 500 nanoseconds (or whatever), resulting in a 16.66667 second (500/30) video of a single "scan-line". Repeat 1080 times for full HD.
-
It sounds to me that they're indeed chained (note, they're not 500 cameras, but 500 sensors, meaning they share almost everything short of a actual receptors and control circuits. I'd describe it as a high-speed laser scanner. As the spokesman mentioned pictures taken with it, I think that images it produces are actually still.
In general, it's a brilliant device. I hope that I'll manage to go to MIT and see that thing someday. I'm planning to study physics on one of the best universities in Poland, so it's possible, if I'll only be good enough for that honor. :)
-
it can image any deterministic event that can be precisely timed and repeated.