Fun for nerds

[RoGeorge]

 

[RoGeorge]

What's with all this recent fuss about measured CMB (Cosmic Microwave Background)?



 

[pidcon]

Filming the Speed of Light at 10 Trillion FPS

[pidcon]

World's Deepest-Penetration and Fastest Optical Cameras

[BU508A]

Which door?


"Can you pass the salt, please?"

[RoGeorge]

World's Deepest-Penetration and Fastest Optical Cameras

I didn't know about all the new applications.

It was intriguing that I couldn't find any good brief info about how it works.  Yes, the video mentions something about how it works, but doesn't teaches.  I've seen a few years ago similar videos with femptosecond cameras, and always wrongly assumed that the principle is similar with that of a sampling oscilloscope, but that is NOT how it works.

The genius of it is that the whole video (like the one with the light passing through a bottle, or the illuminated figurines) is computed out of only one single flash.  The video is not produced by taking repeated snapshots at each flash.  It is one single flash and one photo, and from there, the time evolution of the light is reconstructed (computed).  The movie is calculated, not filmed.  At least this is how I understood it so far.

Following the research papers, the working principles of femptosecond cameras originate in some apparently unrelated concepts that were started during WWII, then developed in the following years to improve (chirped) radar and sonar technologies.  At first it was about impulse stretching while keeping their shape (in analog domain).

Impulse stretching happens by passing the signal through a diffuser, then mixing the signal with a frequency chirp.  By the chirp type, ascending or descending in frequency, the original pulse can be stretched or compressed in time while still keeping its shape.  From there, it all went to building time lenses, or the opposite, to strongly sharpen wider pulses without destroying the vacuum tubes or the sonars, and so on.  Stretched signals were required to extract more details from the echoes, while compressed signals were required to sharpen the pulses sent to make echoes.  The sharper the sent pulses, the more precise the echoes.

Stretching is the first key concept, and from there there are many physics, RF, EE, optics and DSP tricks combined in making femptosecond cameras.  I just hope I didn't get it all wrong, still digging. 

So far it was a fascinating journey down the rabbit hole, thanks for bringing up this subject!   

[pidcon]

Hi RoGeorge, I'm glad you liked the video. I was watching a series of videos on mathematics, and slowly the video recommendations on YouTube steered me towards these high-speed camera research. I'm sure there's more interesting videos on other research areas.

[chris_leyson]

[RoGeorge]

How radioactive are bananas and other radioactive foods?



 

[RoGeorge]

 

[RoGeorge]

Steve Collin's Passive Dynamic Robot



No motors, no microcontrollers, just mechanics and gravity   

[RoGeorge]

Intuitive explanations for Seebeck and Peltier effects   

[T3sl4co1l]

http://markoapparatus.com/portfolio/machine-clock/

[RoGeorge]

Very long (3 hours and 5 minutes), but it doesn't make much sense unless watched entirely in one go.   

[RoGeorge]

Very short (15 seconds), but it doesn't make much sense unless watched entirely in one go.   



LATER EDIT:
In case you were looking for a high resolution version of the annotated thumbnail of the video, like the one seen before pressing play, use this link format:

https://img.youtube.com/vi/<insert-youtube-video-id-here>/maxresdefault.jpg

to get this jpg:

[RoGeorge]

The guinea pig one is wrong.   

[T3sl4co1l]

The guinea pig one is wrong.   

How so?

It's just a visual version of the old joke:

A mathematician and physicist are standing on one side of a room.  On the other side, a prize sits on a pedestal.  The host says: "Each time I ring this bell, you may move half the distance to the prize.  The first one to get it wins!"
Mathematician: "This is dumb, I'll never get there."
Physicist: "I'll get close enough..."

Or more homologically: an infinite series of patrons walk into a bar.  The first orders a beer, the second half a beer, and so on.  The bartender rolls his eyes and pours two beers.

Tim

[RoGeorge]

The number of the threads of fur is finite, and also quantified, therefore the halving process cannot go forever.

The infinite halving is possible for numbers from R, but it's not possible for finite numbers from N, for example.

[T3sl4co1l]

Ah yes, an example of this --
https://math.stackexchange.com/questions/250/a-challenge-by-r-p-feynman-give-counter-intuitive-theorems-that-can-be-transl

I can assert my continued correctness by suggesting the figure is a virtual guinea pig, for which the fur might indeed be infinite, or even continuous.  It's solid shaded, after all.

And also on the subject of approximation, no haircut or shave is perfect, and leaving a few hairs would be a perfectly acceptable outcome (only needing some 14 or so trips to reach).  If it went on for a long time, we might note that real hair also grows, so our job might never be complete; this would be best illustrated with a different animal, the yak, the process of shaving for which is, by one definition, never complete.

Tim

[RoGeorge]

https://math.stackexchange.com/questions/250/a-challenge-by-r-p-feynman-give-counter-intuitive-theorems-that-can-be-transl

Nice collection! 

My most recent favorite is not listed there:  Can you generate fair heads/tails tossing a slightly biased coin?  (von Neumann extractor)

[RoGeorge]

 

[RoGeorge]

 

[RoGeorge]

Smol osc demo at 12:25 

[BU508A]

Making sharp knives out of some very strange materials, e.g. eggs, Bismuth, mushrooms, chocolate, tofu ...

https://www.youtube.com/channel/UCg3qsVzHeUt5_cPpcRtoaJQ/videos

[RoGeorge]

Might be the next big thing