View Single Post
Old 08-23-2012, 10:48 AM   #6
BigV
Goon Squad Leader
 
Join Date: Nov 2004
Location: Seattle
Posts: 27,063
I had to watch it a couple of times to get that too. This is what I've come up with.

Remember he said they fire a *pulse* of laser light, just one trillionth of a second long? It's about yay big (holds fingers an inch apart). That's the little light bullet you see wafting through the Coke bottle. Also illuminating the table, scattering and refracting through the label and the cap. I love the little coda where the bottom of the bottle winks after the bullet leaves the cap and the frame goes dark. Anyhow.

Remember the screen he displayed that was a 3x3 grid of blue fields with various arcing rainbows on it? That was the "raw data". That is what each "frame" looks like before recomposition. Keep this in mind.

Remember the animation of the seeing around corners demo? This is where it all comes together. In that animation they showed the key piece, where the packet is fired from the camera, bounces, splashes all over everything, pieces of the splash hit other things, splash again, repeat until one or two or three lonely photons luckily make their way, independently, asynchronously back to the camera. Remember that?

Well, put all those together, that's how we see the "video" of the pulse's stately march through the bottle. What they kind of gloss over is this: they fire many, *many*, *MANY* of these trillionth-of-a-second pulses. The first one produces lots of data, so does the second one and the third, say you have 15 trillion of these datasets, a trillion to the second for the full fifteen seconds that the bullet/pulse takes to go through the bottle (I'm just estimating that time, work with me). They COMBINE these datasets "computationally" to produce the end result we see here.

It's just like your flying photos, glatt. You took one frame/dataset, you took another frame/dataset, you combined them and the computational result is what you posted of your daughter whirling your son in a superhero/supervillan move. This is doing the same thing. Remember the animation? Only two photons made it back in that animation. It doesn't give a LOT of information. But, if they did another pulse (just a trillionth of a second later) they'd have another two photons or maybe three. A trillion is a really big number, and even if you're only getting couple photons per pulse, if you go bzzzzzt on the PULSING laser you're gonna get a lot of them. Now assemble them. Now you've got your picture around the corner. Or you have your (fraction of a) frame of a movie.

It's a neat trick, and they don't emphasize the many multiples of small pulses it took to make the movie. It looks like ONE pulse moving through the bottle. It's not. It's a picture of a kabillion pulses, all IDENTICAL, but at imperceptibly small increments apart, glued together. Like a movie "moves" when all the pieces are really just a bunch of stills.
__________________
Be Just and Fear Not.
BigV is offline   Reply With Quote