David Makin
|
|
« on: August 18, 2012, 11:33:58 PM » |
|
|
|
|
Logged
|
|
|
|
cKleinhuis
|
|
« Reply #1 on: August 18, 2012, 11:43:56 PM » |
|
OMG how crazy is this ?!?!?
|
|
|
Logged
|
---
divide and conquer - iterate and rule - chaos is No random!
|
|
|
cKleinhuis
|
|
« Reply #2 on: August 18, 2012, 11:53:18 PM » |
|
crazily enjoyed this video, and the speaker is funny by itself - i love indians - he ist talking about recording light at the speed of light, you know from when this video is ?! the most crazyness is how they apply this technique to look around the corners, awesome technological stuff
|
|
|
Logged
|
---
divide and conquer - iterate and rule - chaos is No random!
|
|
|
FracZky
|
|
« Reply #3 on: August 19, 2012, 01:28:13 AM » |
|
What is next... I'm so glad to all these new things happening in my lifetime. I don't think often appreciate the things that advanced... when I was boy something like cellphones was protrayed in SciFi movies and I was in awe... now you laugh at analogue phone sets?? Cool!
|
|
|
Logged
|
Make your 2D Dreams, 3D Reality...
|
|
|
LMarkoya
Strange Attractor
Posts: 282
|
|
« Reply #4 on: August 19, 2012, 01:34:21 AM » |
|
That is pretty wild David...and hard to believe as its hard to figure how a shutter can be faster than the speed of light....... Thanks for sharing
|
|
|
Logged
|
|
|
|
|
taurus
|
|
« Reply #6 on: August 21, 2012, 01:41:04 PM » |
|
...and hard to believe as its hard to figure how a shutter can be faster than the speed of light...
there is no such shutter, i guess not in this universe. i can't say, that i understood even approximately, but i saw a record a while ago, speaking of hundreds (or thousands?) of identical shootings in a row, only with some time shifting between light pulse and camera recording. the trick is to synchronize theese time-shifts with the given precision, to computate the illusion of continuity afterwards. this is not made of a single take (how could it?)
|
|
|
Logged
|
when life offers you a lemon, get yourself some salt and tequila!
|
|
|
cKleinhuis
|
|
« Reply #7 on: August 21, 2012, 02:01:06 PM » |
|
yes, such shutter cant exists, they combine many shootings, especially because of the amount of light that can be grabbed, he is talking about how they did it in the speech, but hence it is recording at the speed of light, you as human do not realize a delay, especially when making photos of around the corners ... it is awesome hightech ... as a side note, i think the facebook community is rather dumb, i posted it in the "i fluffing love science group" and received just 3 likes ... from a group which consists of +700.000 members ... i am glad this received some response here
|
|
|
Logged
|
---
divide and conquer - iterate and rule - chaos is No random!
|
|
|
Jesse
Download Section
Fractal Schemer
Posts: 1013
|
|
« Reply #8 on: August 21, 2012, 05:57:13 PM » |
|
... the trick is to synchronize theese time-shifts with the given precision, to computate the illusion of continuity afterwards. this is not made of a single take (how could it?)
Yes, but this will not explain everything because you might start the recording of many sensors in a synchronized manner, but you cannot stop each photon collecting in a picosecond, so you would record photons from starting time until nearly end of experiment. This would hardly to be deconvoluted in afterwards processing to form good images. And i read in the follow-up links something about projecting the timeline onto the ccd-line somehow, what would make more sense to me. Of course, in reality the produncing of such animations is a very complex thing with here 500 sensors that can only record a 1d line and with a rotating mirror the 2d view is archieved by the "slowest" "fasted" camera.
|
|
|
Logged
|
|
|
|
LMarkoya
Strange Attractor
Posts: 282
|
|
« Reply #9 on: August 21, 2012, 06:55:36 PM » |
|
"there is no such shutter, i guess not in this universe. i can't say, that i understood even approximately, but i saw a record a while ago, speaking of hundreds (or thousands?) of identical shootings in a row, only with some time shifting between light pulse and camera recording"
This can make some sense, but the triggereing then would have to be the the speed of light....maybe possible with many cameras...but he shows a single camera and lens in the video if I am remembering right.... It may be a very, very fast camera, which then uses simple morph software to interpilate frames to make the video as it stands At face value its a little mind boggling
|
|
|
Logged
|
|
|
|
taurus
|
|
« Reply #10 on: August 21, 2012, 07:38:40 PM » |
|
... you might start the recording of many sensors in a synchronized manner, but you cannot stop each photon collecting in a picosecond, so you would record photons from starting time until nearly end of experiment.
as you know, i am more a philosopher, than a scientist, so i don't know the answer, but i can imagine, that it is possible to generate a light pulse of one picosecond (ok, i can't even imagine a picosecond - i can only count digits). we have light emitters, that can produce single photons and between two ticks of a 3ghz processor light travels around 10cm, so why not such a short light pulse? edit> weak argument, i see myself...
|
|
« Last Edit: August 21, 2012, 08:04:03 PM by taurus66 »
|
Logged
|
when life offers you a lemon, get yourself some salt and tequila!
|
|
|
Jesse
Download Section
Fractal Schemer
Posts: 1013
|
|
« Reply #11 on: August 21, 2012, 08:04:38 PM » |
|
as you know, i am more a philosopher, than a scientist, so i don't know the answer, but i can imagine, that it is possible to generate a light pulse of one picosecond (ok, i can't even imagine a picosecond - i can only count digits). we have light emitters, that can produce single photons and between two ticks of a 3ghz processor light travels around 10cm, so why not such a short light pulse? It is becoming very technical i fear, but that is what is triggering me Have build some cameras with ccd sensors by myself, so my little clue is related upon this.. the main problem here is not the light pulse, what is of course possible to generate such a short light impulse. The problem is to record only photons in such a little time interval as one picosecond and discard all photons that arrives later or earlier. You must have a device that you can trigger also whith a picosecond short pulse that is only collecting photons in this short time. This might sound more easy than it is actually (but i am not sure if such a device exists, that is what i would like to know), usually camera sensors got far longer times to clean registers and move charges into a light shielded register - you would also collect the amount of photons arriving earlier or later on a scale of nanoseconds up to microseconds, and no time seperation would be archieved. The fastest timerecording elements commonly used are in oscilloscopes... ok, seems that 25 picoseconds are possible nowadays, maybe thay used ultrafast diodes with such a recording technology.
|
|
|
Logged
|
|
|
|
hobold
Fractal Bachius
Posts: 573
|
|
« Reply #12 on: August 21, 2012, 08:13:13 PM » |
|
You must have a device that you can trigger also whith a picosecond short pulse that is only collecting photons in this short time. As far as I understand, they shoot their photos in a dark lab. The only photons that are around at any given instant of time are those of a single pulse. The sensors do not just record the amount of light, but also the time at which it arrives (i.e. it is a movie camera rather than a still image camera). A new light pulse is not sent into the scene until the current one has completely faded. With near-perfect synchronization between light source and camera, they can record an image pixel by pixel, because their scenes are always completely static.
|
|
|
Logged
|
|
|
|
Madman
Fractal Molossus
Posts: 678
|
|
« Reply #13 on: August 21, 2012, 08:53:41 PM » |
|
The shutter in this case is the laser that illuminates the subject a number of times. I think you can compare it with taking a picture with a very slow film (i.e. very low ISO/DIN/ASA value) of a droplet falling from a faucet with the aid of a stroboscope; the camera is open, while the flashes from the stroboscope "freeze" the motion. If you know the frequency of the droplets falling and you increase the stroboscope frequency to a multiple of the droplet frequency, you can make a movie by combining every nth frame that you record, and you will have a movie of a droplet falling very slowly. One might wonder though, if Heisenberg will start to play a role if these pulses are becoming so short, that the number of photons is almost countable... Anybody any ideas on that?
But it sure generates pretty pictures!
|
|
|
Logged
|
All I want is a warm bed, a kind word and unlimited power (Ashleigh Brilliant)
|
|
|
taurus
|
|
« Reply #14 on: August 21, 2012, 09:16:26 PM » |
|
the main problem here is not the light pulse, what is of course possible to generate such a short light impulse. The problem is to record only photons in such a little time interval as one picosecond and discard all photons that arrives later or earlier.
i realized just after writing, but didn't want to delete. it is just like photographing a gun bullet with normal exposure time. maybe the solution of avoiding "light smear" is, that they only shoot one image-line per light pulse (the thing with the slow rotating mirror) anyhow, in the end - i'm not that thrilled by detailed technical solutions. i'm rather the one, thinking about new cases of use
|
|
|
Logged
|
when life offers you a lemon, get yourself some salt and tequila!
|
|
|
|