Naked Science Forum

On the Lighter Side => New Theories => Topic started by: Expectant_Philosopher on 02/05/2011 03:33:17

Title: Can an image be derived from a sound recording?
Post by: Expectant_Philosopher on 02/05/2011 03:33:17

Can an image be derived from a sound recording?  Can you reconstruct the echo patterns from sound waves to image the room from which a recording was produced?  Would the image be distinct enough to produce a realistic picture of the faces in the room in which the recording was produced?  Can we think of the microphone as a lens in a radar-like camera?  Listening to a recording of a "live" concert, one can almost see the scene, the crowds, the musicians.  If the recording were analyzed against a database of reflective values and resonance factors could an image be reconstructed?  Can you see into the past?

Title: Can an image be derived from a sound recording?
Post by: MikeS on 02/05/2011 06:53:03

As I see it, it would not be seeing into the past any more than you are seeing into the past when watching a film.  It's a recording.

However, it's an interesting question.  In theory, with multi channel recording and the right software to decode phase and amplitude difference you could reconstruct something.  However, due to wavelength of sound the resolution would be very poor in comparison to light.

Title: Can an image be derived from a sound recording?
Post by: Phractality on 02/05/2011 17:36:19

Blind people do have varying abilities to "see" their acoustic environment. Also, there are devices worn on a blind person's head which emit ultrasonic pulses. The echoes are converted to a audible frequencies and directed to the persons ears. This enables the blind person to "see" the acoustic environment in higher resolution than without the device.

Teaching a computer to do what a blind persons does without thinking is the hard part. Some of that technology is available for such applications as imaging a baby in the womb. Sonar for submarines has come a long way. We are learning how dolphins and whales do it and teaching computers to do the same. Of course, most of that tech is top secret. It'll be a while before you can buy it at Radio Shack.

Title: Can an image be derived from a sound recording?
Post by: yor_on on 02/05/2011 19:21:39

I wonder?

It's about frequencies isn't it? Light has so many but sound is more limited it seems to me, if you imagine sound and seeing as two oceans then they have a different size, but you can also take each ocean by itself and find it 'infinite', that is, you may be able to 'split' it into a infinity of narrowed down ah, 'frequency's'? Well, as a guess this is. We have our definitions of that clear, but that doesn't state that with the right software there might be more depthts to it, as Phractality and Mike point's out.

Sonar is a very interesting example. I remember reading about experiments with dolphins somewhere? Able to detect minuscule stress breaks inside materials, or was that in a SF??

Da*n :)

Not sure actually, but assume for a moment that sonar would be the dolphins eyes. How sharp would its sight be?

Title: Can an image be derived from a sound recording?
Post by: Phractality on 02/05/2011 20:02:49

Bottlenose dolphins  (http://www.chelonia.co.uk/bottlenose_dolphin.htm) emit clicks with a peak frequency of 130 Khz, which translates to a wavelength of 11.4 millimeter (.4488 inch). Due to overtones, perhaps they can "see" details smaller than that.

(Calculations and speed of sound in water from Wolframalpha.)

Title: Can an image be derived from a sound recording?
Post by: JP on 02/05/2011 20:21:02

I think you're describing phased arrays of microphones, which can work for sound the same way lenses work for light.  You can take sound waves emitted from a source and reconstruct an acoustic image of the source somewhere else. 

http://www.acoustic-imaging.com/array.html

Also, this isn't a new theory you're asking about, so if you'd like, I could move this to the physics or the technology forum where it might get more views.

Title: Can an image be derived from a sound recording?
Post by: RD on 02/05/2011 20:27:29

Quote

Can an image be derived from a sound recording?

http://www.thenakedscientists.com/forum/index.php?topic=28995.msg301907#msg301907

Title: Can an image be derived from a sound recording?
Post by: Expectant_Philosopher on 04/05/2011 01:37:07

The Topics RD noted are on a slant from this topic.  I was thinking how sound seems almost holographic when I listen to a live recording.  If we could reconstruct from the sound frequencies the interference patterns of the sound waves could we then take a recording from the past like a hostage audio and reconstruct the room and the number of people in the room.  With improvements from experimentation could we then take some of the earliest recordings and reconstruct realistic images?

Title: Can an image be derived from a sound recording?
Post by: Geezer on 04/05/2011 07:07:26

I suppose if there was sufficient data in the recording it would be possible to reconstruct an image, but typical sound recording techniques don't really have much data, so the resolution of the reconstructed image would not be too good.

Title: Can an image be derived from a sound recording?
Post by: RD on 04/05/2011 10:40:31

From echo and resonance it may possible to tell the size of the room a recording was made in, and the reflectivity of the surfaces, e.g. http://www.freesound.org/samplesViewSingle.php?id=22715

Title: Can an image be derived from a sound recording?
Post by: Airthumbs on 04/05/2011 16:35:15

Sounds like Sonar to me.......

Title: Can an image be derived from a sound recording?
Post by: Geezer on 04/05/2011 18:12:24

PING!



Woops! Sorry. I hope I didn't deafen you.

Title: Can an image be derived from a sound recording?
Post by: imatfaal on 04/05/2011 18:17:30

This thread reminds me of another of the famous feynman video explanations; it starts off talking about a bug on a swimming pool inferring all the swimmers in the pool from the waves that reach it. 

It's a great watch

Title: Can an image be derived from a sound recording?
Post by: Airthumbs on 04/05/2011 19:08:16

Quote from: Geezer on 04/05/2011 18:12:24

PING!
Woops! Sorry. I hope I didn't deafen you.

For a second there I swear I could see you!   

PONG!

Title: Can an image be derived from a sound recording?
Post by: yor_on on 05/05/2011 07:34:49

Yes, that's definitely science in Practice.

Let me try...


Ping.