Ultrafast Camera Takes 1 Trillion Frames Per Second of Transparent Objects, Phenomena (phys.org) 36
After developing the world's fastest camera a little over a year ago, Caltech's Lihong Wang decided that wasn't good enough and started working on an even faster device. A new paper published in the journal Science Advances details a new camera from Wang that can take up to 1 trillion pictures per second of transparent objects. Phys.Org reports: The camera technology, which Wang calls phase-sensitive compressed ultrafast photography (pCUP), can take video not just of transparent objects but also of more ephemeral things like shockwaves and possibly even of the signals that travel through neurons. Wang explains that his new imaging system combines the high-speed photography system he previously developed with an old technology, phase-contrast microscopy, that was designed to allow better imaging of objects that are mostly transparent such as cells, which are mostly water.
The fast-imaging portion of the system consists of something Wang calls lossless encoding compressed ultrafast technology (LLE-CUP). Unlike most other ultrafast video-imaging technologies that take a series of images in succession while repeating the events, the LLE-CUP system takes a single shot, capturing all the motion that occurs during the time that shot takes to complete. Since it is much quicker to take a single shot than multiple shots, LLE-CUP is capable of capturing motion, such as the movement of light itself, that is far too fast to be imaged by more typical camera technology. In the new paper, Wang and his fellow researchers demonstrate the capabilities of pCUP by imaging the spread of a shockwave through water and of a laser pulse traveling through a piece of crystalline material.
The fast-imaging portion of the system consists of something Wang calls lossless encoding compressed ultrafast technology (LLE-CUP). Unlike most other ultrafast video-imaging technologies that take a series of images in succession while repeating the events, the LLE-CUP system takes a single shot, capturing all the motion that occurs during the time that shot takes to complete. Since it is much quicker to take a single shot than multiple shots, LLE-CUP is capable of capturing motion, such as the movement of light itself, that is far too fast to be imaged by more typical camera technology. In the new paper, Wang and his fellow researchers demonstrate the capabilities of pCUP by imaging the spread of a shockwave through water and of a laser pulse traveling through a piece of crystalline material.
Re: (Score:1, Insightful)
Re: (Score:3)
I have a camera... (Score:3)
Re: I have a camera... (Score:2)
Re: (Score:2)
I bet that with the support of a little post-processing you can get it to produce a trillion photographs that accurately capture transparent objects and their actions in mere fractions of a second.
Re: (Score:2)
Indeed, I have invented a camera that improves on this technology in two respects: it takes ten trillion pictures per second, not just one trillion, and it accurately captures not just transparent but completely invisible objects. Patent pending. Testing will be completed once I find a place to store all the film.
Jaws reference... (Score:3, Interesting)
Wang invented a pCup? Is it April 1st yet? (Score:2)
Takes photos of transparent things? Captures light?
I'm supposed to take any of this seriously?
Re:Wang invented a pCup? (Score:2)
"I think a 640 kCup ought to be big enough big enough anybody"
Re:Wang invented a pCup? Is it April 1st yet? (Score:4, Interesting)
A picture of two entangled photons [petapixel.com]. So yes, captured light.
And here's a picture of a single atom reflecting light [qz.com] (i.e. photons).
Re: (Score:2)
Re: (Score:2)
"Takes photos of transparent things? Captures light?
I'm supposed to take any of this seriously?"
Sure, think of taking a picture of the outside air......in LA.
If Henri Cartier-Bresson had one of these (Score:2)
If he were to frame one of his famous candid scenes and hold the button down for one second, how long would it have taken him to go through one trillion shots for that "decisive moment?"
Re: (Score:3)
My camera can take pictures of invisible things (Score:1)
In 100% HD.
2 girls 1 pCUP? (Score:2)
Re: (Score:2)
Not dCUP?
This is nothing new (Score:4, Funny)
The other day, some friends were insisting on sharing their vacation photos with us. Given the sheer number of photos, I'm pretty sure they were shooting at roughly 1 trillion frames per second...
Pr0n! (Score:2)
Re: (Score:2)
It can only photograph a 1 microsecond orgasm, so it's perfect for slashdot.
Re: (Score:2)
Vectorize it all (Score:1)
Perhaps the concept of "frames" will go away. You have photon receptors and they can send a signal whenever they detect a photon. No need to force them into a concept of frames. The same is true for movies/videos in general; just represent them as time-morphing polygons. Sure, frames sometimes are handy for editors to work with, but the final form doesn't need to be frames.
Same for pixels. Vectorize it all. That way it will better fit a multitude of viewing devices. And viewing devices won't need to do (inc
Re: (Score:2)
I had a related idea back in 2002 when I was working with image processing. I figured movies could be compressed as 3D wavelets; this would enable automatic balancing between spatial and temporal resolution, since we cannot focus on both at once. Little did I know that Ogg Tarkin was being developed at the same time with this exact idea.
When you store wave parameters, the output resolution need not be fixed. There are MP3 decoders that can output 24-bit sound from the originally 16-bit source, as they ar
Re: (Score:1)
I agree it would probably need to be broken into chunks for streaming, but roughly 10 to 30 seconds per chunk should work. Have smaller chunks at the beginning to improve start-up time. The cut point could fall on scene changes, if available in the threshold window.
Something in-between polygons and wavelets is color interpolated time-morphing tr
Actual Question about the article (Score:3)
Re: (Score:2)
The setup is awesome, but as you suggest, much of the work is in the reconstruction - fitting curves to each bit in each frame. At the end of the paper, it says "72 pixels × 512 pixels × 350 pixels took about 15 min for 50 iterations" on an "Intel Xeon E5-2670 v3 CPU (48 cores at 2.3 GHz) and 256-gigabyte random-access memory."
Awesome paper, misleading headline. (Score:2)
The setup doesn't take "1 trillion frames per second of an object," it takes "up to 350 frames captured at (a rate) up to 1 trillion frames per second."
Well-written paper! (Score:2)
What's going on (Score:1)
Have I mistyped the URL?
Ultrafast Camera Takes 1 Trillion Frames Per Secon (Score:1)