1,000,000,000,000 Frames/Second Photography

[Dembadon]

I found this demonstration fascinating!

https://www.youtube.com/watch?v=SoHeWgLvlXI

 

[phinds]

I found this demonstration fascinating!

There was a discussion about this on this forum recently and as I recall there was some opprobrium regarding either the process itself or (more likely) the way it was incorrectly described. I don't mean that anyone thought it was a fake or anything like that, just that it wasn't quite what it purported to be.

Sorry I don't have a link to the thread.

 

[enosis_]

I love this stuff!

 

[Dembadon]

There was a discussion about this on this forum recently and as I recall there was some opprobrium regarding either the process itself or (more likely) the way it was incorrectly described. I don't mean that anyone thought it was a fake or anything like that, just that it wasn't quite what it purported to be.

Sorry I don't have a link to the thread.

Gotcha. I'll hunt it down.

 

[Greg Bernhardt]

Gotcha. I'll hunt it down.

I'd be interested in reading that too

 

[AGNuke]

This is a very high speed camera set up at MIT.

Technically, it's not a camera, it's an array of 500 sensors, each triggering at very short time period.

 

[Mech_Engineer]

I found this presentation interesting because at my previous job we were in the "state of the art" for high speed imaging in Department of Energy nuclear science. They are very interested in ultra-high speed imaging, and the best you can do in 2-D (frame) imaging is on the order of 10-50 million frames per second using tube-based individual frames and a single sensor. Digital framing cameras are worse, unless you interleave multiple sensors which can pose several other challenges. In either case, light collection is always an issue in ultra high speed imaging (which MIT solves by multiple integrations of the same scene).

For a fundamental leap to MIT's claimed 1 trillion frames per second (100,000 times faster) the MIT research is using "streak" cameras which I'm familiar with as well. See here:

http://web.media.mit.edu/~raskar/trillionfps/

By carefully synchronizing the pulsed illumination with the capture of reflected light, we record the same pixel at the same exact relative time slot millions of times to accumulate sufficient signal
...
To capture propagation of light in a tabletop scene we need sensor speeds of about 1 ps or one trillion frames per second. To achieve this speed we use a streak tube. The streak camera uses a trick to capture a one dimensional field of view at close to one trillion frames per second in a single streak image. To obtain a complete movie of the scene we stitch together many of these streak images. The resulting movie is not of one pulse, but is an average of many pulses. By carefully synchronizing the laser and camera we have to make sure each of those pulses look the same.

A streak camera is a 1-D time resolved recording device (you might be able to call it a 1-D "video" camera). Using a photocathode, it converts light to electrons in a vacuum and deflects it using electrostatic charge (sort of like an analog TV tube). The resulting electrons then hit a phosphor screen on the back side of the tube which glows and is read-out using a digitizer like a large cooled CCD sensor. It can give you a VERY fast time history of light on a single line. To reconstruct the video we see, they have to take thousands of lines of data over and over and add them together into an image. So this imaging technique can't be used on a single event, only 100% repeatable events.

They also don't tell you a streak camera is fundamentally monochromatic; the color images we see either had to be constructed using individual R,G,B data sets taken separately, or it was added in using software post processing.

Still, its a very pretty picture.