It looks like you're using an Ad Blocker.

Thank you.

Some features of ATS will be disabled while you continue to use an ad-blocker.

Help ATS via PayPal:

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

page: 12
33
share:

posted on Aug, 11 2014 @ 05:15 PM

originally posted by: neoholographic
Again, you have no idea what you're talking about.

You said:
The vibrations caused by 3000 lbs+ objects, moving at high speeds (relatively), will be far more powerful than any vibration caused by the sound of a human voice.
Who said the vibrations caused by a human voice are different than the vibrations of a car driving by??

Physics.

The sound is recreated as a whole so there wouldn't be vibrations coming from a car passing by that's any different than the vibrations being created by a human voice. The same way a mic will pick up the conversation with cars passing by if you had a criminal wired, is the same way this technology will pick up what's being said.

Yet, microphones have baffles to minimize the simple disruptions caused by wind..., but this technology, which has no baffle will not be affected by the magic of...you?

The researchers developed an algorithm that combines the output of the filters to infer the motions of an object as a whole when it’s struck by sound waves. Different edges of the object may be moving in different directions, so the algorithm first aligns all the measurements so that they won’t cancel each other out. And it gives greater weight to measurements made at very distinct edges — clear boundaries between different color values.

So they're recreating the sound similar to a microphone but using visual data. This will give them an advantage and the algorithm in the future may even be able to distinguish between a human voice and say a horn blowing in the background and then have the algorithm just play back the sounds of the human voice that's being recreated.

That's nice. You have yet to explain how they can isolate a vibration created by a sound wave from a vibration created from a physical action.

I will give you a hint...currently, they can't.

So again, you have to actually try and read about the technology and you will know why they will look for applications in law enforcement and forensics through this technology.

I'm going to go out on a limb and state that the only person in this thread that is not familiar with this technology is...you.

posted on Aug, 11 2014 @ 05:16 PM
ALSO, BEFORE ANYONE RESPONDS TO MY POST!

Please debate against something I actually said instead of making things up as you go.

This thread is filled with long diatribes that are debating against a claim I never made. These posts will get 3 of these:

It doesn't make any sense to have a debate where I have to spend post after post pointing out that you're debating against something I never said.

posted on Aug, 11 2014 @ 05:21 PM

A debate implies that both sides are informed enough to hold intelligent discussion. Currently, there's you making baseless claims based on incoherent and uninformed understanding and everyone else trying to point out exactly why you are incorrect in your conclusions and reasoning.

This isn't a debate. It's a debacle.

posted on Aug, 11 2014 @ 05:38 PM

in all honesty, I don't think anyone knows what you are saying.

posted on Aug, 11 2014 @ 07:56 PM

originally posted by: neoholographic

You said:

What I mean is, it detects how much (amplitude) and how fast (frequency) the plant or bag is moving by looking at it.

THIS ISN'T WHAT YOU INITIALLY SAID:

5. The video camera looks at those vibrations, and then analyzes the frequency and amplitude of the vibrations of the bag.

No it doesn't. I agree the camera detects these vibrations to be analyzed by the technology and then they recreate sound.

This isn't what you initially said and I suspect you realized your era and used DETECT but the sad thing for you is I SAW YOUR ERA TOO LOL.

For the purposes of this layman's version of things, "looks at" and "detects" are being used interchangeably. This is no error here.

The video camera and software looks at/detects the motion of the bag. It "sees" and detects the tiny vibrations the eye cannot see by analyzing minute pixels changes.

HOWEVER, the camera and software is still seeing and measuring the amplitude and frequency of those vibrations,

Now let's go onto your other contention.

This video camera "microphone" is extremely similar in concept.

When did I ever say they were not similar concepts?

Again, I ask WHAT ARE YOU DEBATING? WHY ARE YOU DEBATING AGAINST SOMETHING I NEVER SAID??

I said this can be more effective than a microphone and this is why their looking at applications in law enforcement and forensics.

I never said they were not similar concepts.

YOU GUYS HAVE A BAD HABIT OF DEBATING THINGS THAT I NEVER SAID.

And I never said you didn't think they were similar concepts. I'm was simply summarizing everything so that we are all talking about the same thing.

Please show me where I said "what you don't understand is that this is similar in concept to a traditional microphone". You can't, because I never said that. In fact, I have no idea what it is you think; that's the problem I was having -- I couldn't tell what it was you believed was at play here with this technology. Your too busy ranting and yelling that I can't tell how you think this technology works, nor can I tell if you see its current limitations.

THAT'S why I carefully summarized my thoughts on this, just so I could lay down a baseline for this discussion. By carefully laying out what I believe to be true about this technology, we both know where I am coming from.

With a microphone sound is picked up by the mic. With this technology you will have more flexibility because you can use different objects throughout the room to use to recreate sounds from the room and I'm sure this is why applications in law enforcement and forensics are being looked at.

Let's look at a case of John Gotti.

John Gotti would talk to his people outside of their hangout because they wanted to avoid their conversation being picked up by a microphone....

...With this technology, the cloths they're wearing, their bodies, the trees and walls around them become microphones.

This is why they will look at this technology for applications in law enforcement and forensics. It will be like putting a mic on criminals without turning someone to where a mic.

Like I said, it's technology that can be very useful and you guys need to try and understand what your talking about.

Yes. If this technology is perfected, it could be useful for that purpose. But it would need to be further perfected; they would still need to find a way to cancel out background noises and background vibrations that could mask the voice. The problem still remains that objects such as leaves on trees are not efficient microphones. Right now, I bet a parabolic microphone would do a better job in picking up specific voices coming from a specific direction.

This video camera method simply cannot (yet) hone in on JUST the vibration of a leaf caused by the target voice, considering that a leaf in the real world wold be experiencing other motions. It seems to me that they have not yet figured out how the to sort through all of the other vibrations to key in on just the voice. People walking past the plant and wind blowing will impart micro vibrations (even smaller than 100th of a pixel) that this system could think was sound.

By the way, I doubt the walls and their bodies will work like a diaphragm the way a chip bag or a leaf would. Rigid surfaces are probably not going to vibrate due to sound waves that way a bag or a plant does.

edit on 8/11/2014 by Soylent Green Is People because: (no reason given)

posted on Aug, 12 2014 @ 05:58 AM

Wait until you discover your thoughts to also give off vibrations as does everything you touch.. I wish they stop now before it is to late... Throwing your trash away will soon reveal your entire life.. If they can manipulate the object to the point they can measure previous vibrations.. Only a matter of time.. Even the ground you walk on gives off your vibrations.. And is easy enough to see with enhanced pattern recognition..and proper lighting.

I already see a future where your mind is scanned just by stepping into the outside world.. Handheld devices now to punish you for thoughts.. This is not just paranoia.. Water has a tendency to hold onto vibrations and detect them at the greatest distances as it is...

Also judging by the video they already can separate the "noise".. Wtg MIT.. Giving the people more " security". It only takes about 10minutes to rip any vocal wave from music or any interference if you don't believe me on that check out any software that allows you to edit lines like mml. You don't think they can do that with tones.. Go play with audacity.. This technology already exists.. They just discovered more ways to invade your privacy.

So well done MIT your kids shall thank you when they are bullied at school for being a nerd and wish doom and gloom on someone in the teenage years even of they never would act on it.
edit on 12-8-2014 by BlackArrow because: (no reason given)

posted on Aug, 12 2014 @ 07:23 AM

originally posted by: BlackArrow

Also judging by the video they already can separate the "noise".. Wtg MIT.. Giving the people more " security". It only takes about 10minutes to rip any vocal wave from music or any interference if you don't believe me on that check out any software that allows you to edit lines like mml. You don't think they can do that with tones.. Go play with audacity.. This technology already exists.. They just discovered more ways to invade your privacy.

So well done MIT your kids shall thank you when they are bullied at school for being a nerd and wish doom and gloom on someone in the teenage years even of they never would act on it.

Removing audio relies on having a good clean signal in the first place. They are simply not at that stage. The sound recorded under perfect experimental conditions are better quality than the first known voice recording from 1860 - but not by much!

Also, it is still insanely difficult (if not impossible at the present time) to completely isolate echo outside of a very specific set of conditions. Talking in a room with a strong echo might be enough to make this technique unusable for now.

This experiment is an interesting demonstration to combining two or three technologies. It's a proof of concept, but very far from being usable in the way that some people in this thread seem to think. In fact, the limitations to the concept may be beyond our current technological ability to overcome outside of perfectly controlled conditions.

Not entirely sure how you jumped from an interesting application of existing well-known principles and effects, to mind-reading and bullying, but this has already become the thread where anything can happen.

posted on Aug, 12 2014 @ 11:34 AM
a reply to: EvillerBob I would easily say they will have rthe tech down in less then five years. If they release it to the public is another story. I don't like the fact that with all of studies directly into sound like pushing and pulling objects and now using anything to record.. U must have forgotten they already have satalites in space that can record you within full details. The cameras released to the public are only a fraction of the quality that is out there. As for the higher ups no problem.

Don't you think that is amazing three discoveries powerful ones all sound oriented were discovered less then a week apart?

With sound study progression and military or government funding this can be fully operational in a matter of days. Especially since every country appears to be dabbling with sound and vibrations at the same time. The lens for quality is already there as I am sure you are awareness that SATs themselves can count the hairs on the back of your hand if you access the gov ones.

While isolating sound takes time removing Eco etc. What do you think music people are great at? That industry alone has had the power to manipulate sounds all along. I use to sit there and splice sound waves into separate tracks by going through and removing sounds out of key or unwanted instruments for a video game.

You think it is difficult to remove Eco's. Please I dare you look up mabimml music editor and load any midi it looks just like the charts on the video. Play with it and you yourself can split a track if you have a good ear for it. Granted because it is free and open source it is limited to midi fruiti loops is another bit by bit music editor. These programs are used to create music in a game that has been out for the last 10years.

I guess until you toy with it yourself you won't truly be able to see that a 10yr old would be able to clean up the translated tracks from video to audio in a few hours.

If you look even further into the vocaloid community the same thing is done by splitting a track to make sounds that make noises that are comprehensive to singing.

The audio community lacks nothing when it comes to removing background noises.
The military and gov lacks nothing when it comes to the equipment needed to take the required frame rate video. Stop watching webcams on a 10second refresh timer. They have a computer that is capable of creating as universe within itself that scientist have used to debate if we are even "real".

Don't turn a blind eye to the massive research going on with soundwaves and vibration considering everything is just pouring out one after another at this point.

Yes maybe today they are not merged but as early as tomorrow in the wrong hands it could be.

posted on Aug, 12 2014 @ 11:46 AM

originally posted by: BlackArrow

The audio community lacks nothing when it comes to removing background noises.

You grossly overstate the technology available. Successfully removing background noise is not only entirely source dependent and a destructive process but except for optimum cases is often all but impossible at a satisfactory level. Removing a bit of hiss from a recording or isolating a crude vocal from a recording under optimum conditions is one thing, attempting to accurately isolate a complex signal from a noisy environment is an altogether different beast.

Anyway, none of this is really on topic since the technology demonstrated makes no claims of this sort.
edit on 12-8-2014 by GetHyped because: (no reason given)

posted on Aug, 12 2014 @ 03:27 PM

originally posted by: GetHyped

originally posted by: BlackArrow

The audio community lacks nothing when it comes to removing background noises.

You grossly overstate the technology available. Successfully removing background noise is not only entirely source dependent and a destructive process but except for optimum cases is often all but impossible at a satisfactory level. Removing a bit of hiss from a recording or isolating a crude vocal from a recording under optimum conditions is one thing, attempting to accurately isolate a complex signal from a noisy environment is an altogether different beast.

Anyway, none of this is really on topic since the technology demonstrated makes no claims of this sort.

To be fair I did start by saying about five years. As the basics are there.

posted on Aug, 12 2014 @ 03:31 PM
Also the bullying bit as you called it was a sarcastic form knowing all to well anything created with good intentions usually turns around to be abused and harms others.

posted on Aug, 12 2014 @ 04:36 PM

originally posted by: BlackArrow
U must have forgotten they already have satalites in space that can record you within full details. The cameras released to the public are only a fraction of the quality that is out there. As for the higher ups no problem.

Oh they have amazing quality cameras, that I don't question. But the camera isn't the bottleneck. It's what the camera is observing. Microphones are custom built to respond accurately to sound. Bags and plants - not so much. You could have the world's best quality cable, but it won't help if you plug it into the world's worst microphone.

originally posted by: BlackArrow
You think it is difficult to remove Eco's. Please I dare you look up mabimml music editor...

I took up that dare. You might need to give me a link, though, because StartPage isn't bringing me any useful results. If it can cleanly remove echo from audio I would be incredibly interested, because it's a battle that I face on a regular basis, especially when clients decide they can handle their own audio recording.

originally posted by: BlackArrow
The audio community lacks nothing when it comes to removing background noises.

Except an easy and effective way to aesthetically remove echo. Believe me, I've looked. There are ways if you're lucky with the nature of the recording, and one of those ways might be useful with this technology if you have multiple items to analyse, but the rule of thumb is "the only effective way to get rid of echo is not to have it in the first place."

You can certainly take steps to reduce the noticeable effect of the echo (with mixed degrees of success) but at the end of the day you're still trying to apply them to observable data collected from a random object, rather than the output of a proper microphone.

originally posted by: BlackArrow
I guess until you toy with it yourself you won't truly be able to see that a 10yr old would be able to clean up the translated tracks from video to audio in a few hours.

I've done more than just "toy with it", including dealing with both video and audio recorded in situ (and often less than ideal conditions and equipment) and trying to bring it up to as close to broadcast-standard as possible. It's not my main job, but it's a core part to another business I have on the side.

originally posted by: BlackArrow
If you look even further into the vocaloid community the same thing is done by splitting a track to make sounds that make noises that are comprehensive to singing.

Turning data into sound (the purpose of vocaloid) isn't the issue, it's having a clean enough source of data in the first place!

posted on Aug, 12 2014 @ 04:39 PM

originally posted by: BlackArrow
Also the bullying bit as you called it was a sarcastic form knowing all to well anything created with good intentions usually turns around to be abused and harms others.

Fair enough, it's not always easy to pick up on these things based on written text. I agree entirely that we as humans have a special gift for amplifying the worst aspects of any promising technology

posted on Aug, 12 2014 @ 05:36 PM
As for the toying with I ment in reference to the programs sorry if that didn't come out look up mabinogi followed by mml. This program was used for midis often when you load something like death note in the program had every instrument on one track. If you YouTube it you can see examples of mml music creation. People would sit there with a one dozen line track with every instrument mixed on multiple or single recordings depending on what you used. You can search fruit I loops on YouTube as well to see tracks often over layered by each other. But what I ment with oneprogram such as that you can remove individual keys example piano and keep the guitar even after converting mp3 to midi format granted non free programs have an even superiors targeting that eases the burden via scale and pitch with a prediction code.

It might be called 3ml but it is on a Chinese page. 3ml

But remember this basic software is honest ancient.. There are much more advanced ones now that do a lot more this software is free. But if you play with it via converting and importing you can see what it is capable of. Mml is the raw data of sounds. Sharp eye converts sheet music to sound that is right a picture scanned PDF to music. Granted that is a very old software too don't remember the last time it was updated but there are others out there.

The method I believe is similar to the actual detrections they use at MIT to pick up vibrations but on an obvious advanced scale.

posted on Aug, 13 2014 @ 06:12 PM
a reply to: Soylent Green Is People

We seem to be lost in translation here because you don't understand the technology. My advice is that you take the time to actually understand what you're talking about. You said:

Yes. If this technology is perfected, it could be useful for that purpose. But it would need to be further perfected; they would still need to find a way to cancel out background noises and background vibrations that could mask the voice. The problem still remains that objects such as leaves on trees are not efficient microphones. Right now, I bet a parabolic microphone would do a better job in picking up specific voices coming from a specific direction.

This video camera method simply cannot (yet) hone in on JUST the vibration of a leaf caused by the target voice, considering that a leaf in the real world wold be experiencing other motions. It seems to me that they have not yet figured out how the to sort through all of the other vibrations to key in on just the voice. People walking past the plant and wind blowing will impart micro vibrations (even smaller than 100th of a pixel) that this system could think was sound.

What background noises and background vibrations are you talking about? You then said something that sounds even sillier. You said the camera cannot hone in on just the vibrations of a leaf caused by the target voice???????

THIS IS JUST NONSENSE!!!

First off, the technology recreates sound as a whole so there isn't any other background vibrations or background noises that are not picked up by this technology.

That technique passes successive frames of video through a battery of image filters, which are used to measure fluctuations, such as the changing color values at boundaries, at several different orientations — say, horizontal, vertical, and diagonal — and several different scales.

The researchers developed an algorithm that combines the output of the filters to infer the motions of an object as a whole when it’s struck by sound waves. Different edges of the object may be moving in different directions, so the algorithm first aligns all the measurements so that they won’t cancel each other out. And it gives greater weight to measurements made at very distinct edges — clear boundaries between different color values.

Like I said, all you have to do is read about the technology and you wouldn't make statements like this that have nothing to do with the technology.

Sound hits an object like a bag of chips. This includes all sounds. There are no special background vibrations or background noises that are not picked up by this technology.

So if a person is talking, the wind is blowing and a car is driving by these noises will hit the object being used as a whole. The technology then measures these vibrations through VISUAL DATA. They do this by running successive frames through their technology. This is why the better the camera when it comes to fps the better the sound that's being recreated.

There's not background vibrations or micro vibrations coming from the wind or anywhere else. The vibrations are coming from VISUAL DATA IN FRAME.

So of course they can use a leaf or anything else caught in frame because they're running successive frames through their the technology and the algorithm measures the small differences between frames and uses that to recreate the sound.

It doesn't matter what motions the leaf would be experiencing as long as it could be captured in frame. As long as the leaf isn't moving at superman speeds where the algorithm can't make out the small differences as it reads successive frames then of course it could use a leaf.

They also did other experiments.

In other experiments, they extracted useful audio signals from videos of aluminum foil, the surface of a glass of water, and even the leaves of a potted plant. The researchers will present their findings in a paper at this year’s Siggraph, the premier computer graphics conference.

They could video tape the ocean and recreate sound as long as it's not a tsunami. Even then it would have to be moving fast enough not to be captured in frame. This is because the ocean isn't moving fast enough not to be caught in frame while being taped by a camera that has 5,000 fps. So they can then run successive frames through their technology and the algorithm will pick up these small differences which are vibrations less than 100th of a pixel and recreate sound.

At the end of the day, you don't understand the technology. If you did, you would realize how silly you sound.

posted on Aug, 13 2014 @ 08:13 PM

originally posted by: neoholographic

At the end of the day, you don't understand the technology. If you did, you would realize how silly you sound.

He understands it perfectly well.

If someone was talking while a radio in the same room was playing extremely loud music (loud enough that you couldn't tell what the person was saying if you were just using a normal microphone and listening in) do you think this technology would let the scientists work out what the person was saying?

posted on Aug, 13 2014 @ 09:45 PM

So if a person is talking, the wind is blowing and a car is driving by these noises will hit the object being used as a whole.

Great. Now all you have to do is stop talking and show how this works. Good Luck!

posted on Aug, 14 2014 @ 06:40 AM

Again, you are failing to grasp the technology here. The technology is analyzing vibrations in the spatial domain and converting them to vibrations in the time domain. The speed of the camera is nothing more than a function of the Nyquist limit. They even state that upper frequencies are problematic and result in a noisier signal hence faster FPS does not necessarily equate to a better quality signal. With all this in mind, masking and maintaining a good SNR is the same problem as any other recording technology. Again, we're just repeating the same thing to you over and over and you continue to demonstrate your failure to grasp the concepts behind this technology.

The Dunning–Kruger effect is a cognitive bias manifesting in unskilled individuals suffering from illusory superiority, mistakenly rating their ability much higher than is accurate. This bias is attributed to a metacognitive inability of the unskilled to recognize their ineptitude.[1]

David Dunning and Justin Kruger of Cornell University conclude, "the miscalibration of the incompetent stems from an error about the self, whereas the miscalibration of the highly competent stems from an error about others".[2]

en.wikipedia.org...

posted on Aug, 14 2014 @ 11:38 AM
Im waiting for the day when they can turn sound in to images.

posted on Aug, 14 2014 @ 11:42 AM

You already can. It's called a spectrogram.

top topics

33