WOW! M.I.T. Researchers can recreate sound from objects in the room

Hope the busybodies like punk rock as here what theyd hear is The Clash .. Dead Kennedys .. SS .. the stalin and some of the upcoming punk bands in china and japan .. oh and me swearing at my miscreant macaque in japanese for stealing my beer ..

a reply to: neoholographic

This is pretty cool and takes spying to a whole new level. They actually recreated sound from the vibrations of plants and a bag of chips. They took video of the bag of chips as someone was singing in the room and they recreated the sound from the vibrations of the bag of chips.

It's impressive stuff for beginners. The high-speed camera would need a 100% static base to record accurately which would create issues in-the-wild as a spying gadget.

The locations are very acoustically reflective and the songs are very basic - lyrics highly enunciated. This would mean that identifying the signal in furnished rooms from nuanced, varied sounds would be one with diminishing returns.

On top of all that, the target rooms might have a window open or have the A/C running. All of which could cause a secondary and possibly more dramatic source of vibration that'd drown the speech signal.

I'd guess the real spooks have technology that'd be years in advance of the MIT work.

I think it's cool but certainly not something to worry about from a privacy standpoint.

Close the blinds. If there is a camera in the room you're already screwed anyway.

I'm fairly confident there is some cool stuff for eavesdropping.

a reply to: Kandinsky

For beginners? Let me see you do this.

The things you're saying don't apply. They're recreating the sounds based on video and they're measuing vibration less than 100 pixels based on video not audio. This is why they did this through soundproof glass.

So the plants nor the bag moved so how do you manipulate or interfere with vibrations on a level this small that's not based on audio but video of an object in the room even through soundproof glass?

a reply to: neoholographic

'For beginners' as in genuine surveillance technology will be far advanced. It wasn't intended to mean that anyone can do it themselves.

Everything I said still applies. Should that huge camera be set up on a surface that wasn't completely stable, the vibrations would prevent the accuracy of the high-speed footage by creating blurred footage. With blurred footage, they couldn't recreate the sounds.

If the object of the video was being moved by vibrations *other* than speech e.g. A/C, footprints, draughts, the footage would provide inaccurate results if the results were intended to recover the speech.

So, yeah, what they did was a success that proved their ideas work. However, you framed the success as something useful for spying and I was pointing out that (as it stands) it wouldn't work in-the-wild.

Now, if you can argue that a big camera, a totally static base and a situation arising where two stage actors speak clearly in an otherwise silent room occurs in the world of spying...I'll stand corrected.

a reply to: TheLaughingGod

This is scary! All that AND a bag of chips...

a reply to: Kandinsky

No it wouldn't. Again, how can you manipulate vibrations that small that can even be picked up through soundproof glass??

They weren't even using the fastest cameras out there when they did this. They recreated sound using 60 fps and it got better when they use a 2,000 - 6,000 fps camera but you have cameras that are much faster. Here's more from M.I.T. News:

Reconstructing audio from video requires that the frequency of the video samples — the number of frames of video captured per second — be higher than the frequency of the audio signal. In some of their experiments, the researchers used a high-speed camera that captured 2,000 to 6,000 frames per second. That’s much faster than the 60 frames per second possible with some smartphones, but well below the frame rates of the best commercial high-speed cameras, which can top 100,000 frames per second.

How are you going to manipulate sound when they're are talking about vibrations that can't be seen? Imagine if they had a camera with a higher fps than 2,000 - 6,000.

Here's what one of the Researchers said:

“This is new and refreshing. It’s the kind of stuff that no other group would do right now,” says Alexei Efros, an associate professor of electrical engineering and computer science at the University of California at Berkeley. “We’re scientists, and sometimes we watch these movies, like James Bond, and we think, ‘This is Hollywood theatrics. It’s not possible to do that. This is ridiculous.’ And suddenly, there you have it. This is totally out of some Hollywood thriller. You know that the killer has admitted his guilt because there’s surveillance footage of his potato chip bag vibrating.”

At the end of the day, there's nothing you could do. You're dealing with vibrations that are so small and they don't depend on audio to be picked up. This is why they did it through sound proof glass. You get a camera with better fps you get better audio.

If I'm talking on the phone and they're video taping my pillow, the vibrations they're using to recreate the sound are so small there's nothing I can do to stop it. I don't even see my pillow moving. How can you manipulate vibrations that small that don't depend on audio?

a reply to: neoholographic

I'm not making myself clear am I??

The vibrations on the plant leaves were caused by the sound-waves from the song or the voice. They were the only sounds to create the vibrations.

*If* there were other sounds in the room, the plant leaves would still vibrate, but the song or the voice would be less recognisable. *If* there was a breeze from an A/C, the leaves would be moving independently of the vibrations caused by the speech alone. All of these effects would still cause vibrations on the leaves.

a reply to: Kandinsky

Again I ask, why would this make a difference when the recreation of the sound depends on vibrations so small that they can't be seen? You still haven't shown why the voice couldn't be recognized based on the algorithm used to recreate the sounds.

originally posted by: EveStreet
a reply to: TheLaughingGod

This is scary! All that AND a bag of chips...

That's a good one.

a reply to: neoholographic

Because the micro vibrations would be affected by other micro vibrations.

Draughts, air con, other noises, footsteps upstairs, background music all cause micro vibrations. The video is based on a clinical environment not the noisy, shaky world the rest of us live in.

originally posted by: Domo1
I think it's cool but certainly not something to worry about from a privacy standpoint.

Close the blinds. If there is a camera in the room you're already screwed anyway.

I'm fairly confident there is some cool stuff for eavesdropping.

Back in the 1980s, the Americans and Russians were spying on each others embassies and intelligence buildings by reading the window vibrations of those buildings created by the people talking.

blogs.ft.com...

a reply to: neoholographic

Fantastic thread
I shared the video with my facebook group

They should use this tech to "listen" to stellar objects..

Ohh and UFO footage !!

originally posted by: TrueBrit
a reply to: MystikMushroom

The use of a long lens might extend the range of this technique, and at that point, not only would no conversation be totally safe, but even those who would have access to the sort of surviellance operation identification gear that intelligence agency personnel sometimes get outfitted with, would have serious difficulty in knowing for certain, that they were not being recorded, even if they were out in the sticks.

Every blade of grass, every window, every table, cupboard, bag of chips, empty plastic bag, every steaming cup of coffee could be a microphone, a microphone that you cannot baffle. This discovery invalidates every single anti-surveillance countermeasure, that I have ever heard of.

While it certainly extends the capabilities, I think it certainly doesn't invalidate "every single anti-surveillance countermeasure". It relies on the audio signal being the dominant force in an otherwise static environment. I doubt either the plant or the bag of crisps experiments would have worked outside with a stiff breeze or heavy footfall nearby.

It would be quite interesting if the best countermeasure was simply to hire a fat man with bad wind to walk around the room while you're talking

originally posted by: neoholographic
a reply to: Kandinsky

Again I ask, why would this make a difference when the recreation of the sound depends on vibrations so small that they can't be seen?

I'm not entirely sure which part of the explanation is causing difficulty, the limitations seem entirely obvious.

You can only detect the small vibrations because there is no other stronger vibration to overwhelm them.

If you are having a whispered conversation and you want to stop people overhearing, you can play loud music to drown out the soft whispers to anyone but intended listener. Applying the same logic, if you don't want the bag of chips to "give away" the vibrations caused by your voice, you do something to the bag to mask the small movements by creating bigger movements - whether that's using louder music, a breeze through the window, whatever.

The vibrations they are detecting are not some kind of separate vibrations that will continue to be detectable even if the bag is being moved, they are exactly the same vibrations caused by moving the bag. This system only works because the bag is not being subject to any other stronger vibration to override them.

a reply to: neoholographic

signal to noise ratio

originally posted by: TrueBrit
a reply to: Gaud27

I was thinking about how radical it would be to have optical pickups on my bass guitar!

New ways to spy, but also new ways to ROCK!

The possibilities are astonishing!

Yeah. you could combine it with this people.csail.mit.edu... and run through some amps.

Tactile transducers.
2nd.

THE BEST thing to use to distrup any kind of voice recognition, PLAY DEEP BASS, its sure to distrup any sound waves and cause any vocals to be drawn out

Interesting technology, I can think of several beneficial uses:

1) Could be used on flights / planes. With live video feed going back to a point on land and recorded for use if later needed. Might be useful to hear what was going on in the passenger cabins of plane or pilot cabin of plane when something goes wrong. They would no longer have to wait solely for the black box.

2) Paranormal Investigators: Could now use this to corroborate an EVP. If they have a video recording of an object that was in the same room and time as an EVP is/was captured, they could then use this video to play back the sound and see if the same EVP was captured from the object too. They would now have a second source to back up the EVP recordings.

3) Video footage of crime scenes where they could now try to capture the noises during the crime, it may or may not help shed some light on solving a case that they could not prior because the video didn't capture anything and didn't have any sound...or so they thought. It might just lead to a clue undiscovered previously and crack a case.

These are just some examples, I know the other side of this coin is negative uses unfortunately.

Just my thoughts...

leolady