Laser bouncing

[Nathelion]

I'm sure you have all heard of the cold-war spying technique of bouncing a laser off of a window thus measuring the vibrations of said window, which are then used to extrapolate sounds inside the room (such as private conversations).
As a project for physics class, my group is going to attempt to use this technique to see if we can "eavesdrop" on a room across campus. It's been rather difficult, however, to find any specific details on how this was (and is) actually done, such as the modulation used, the angle of incidence of the laser, etc...
Does anyone know of any websites/books/other sources where I could find out more? It doesn't have to be peer-reviewed or anything like that, just a technical description of the actual procedure.

Thanks!

 

[PottedMeat]

Heh I'd like to know too. I knew someone a few years back that made a laser eavesdropping device for his final EE senior project. I remember him saying that it worked, but for some reason it would only work a few feet from the window.

 

[irishScott]

Try the stickied "Searching for Information" Thread

http://forums.anandtech.com/me...=1731866&enterthread=y

 

[KIAman]

Building a laser eavesdropping machine is child's play... It's the filtering of every other noise in the world and only leaving the vocal contents that's the hard part.

Learn some programming and develop a noise filtering algorithim and you're set.

 

[Modelworks]

I recall reading a DOD file that specified various levels of security and one of the things listed is that for some high security discussions its not permitted for the discussion to take place in a room with windows, or even a room that is on the exterior of the building, for the above reason.

So it must be used quite a bit.

 

[wwswimming]

a lot of defense contractors have large conference rooms that
are basically EMI test chambers with a table and chairs.

i wouldn't be too surprised if the laser-acoustic technique was
available at Fry's.

is the laser pointed directly at the room in question, or at
an angle ? until you get a conclusive answer, i would pursue
both.

 

[irishScott]

Originally posted by: wwswimming
a lot of defense contractors have large conference rooms that
are basically EMI test chambers with a table and chairs.

i wouldn't be too surprised if the laser-acoustic technique was
available at Fry's.

is the laser pointed directly at the room in question, or at
an angle ? until you get a conclusive answer, i would pursue
both.

Requiring a direct, level LOS would be pretty impractical IMO, so I'd assume it'd have to be usable from an angle.

 

[gorobei]

my understanding of the process(after a couple of news reports and a pbs special) was that you shoot a laser at the window pane and hit some surface just inside the frame(blinds/curtains/sill/whatever.) You then use a high power camera lens to view the red dot of the laser. Software converts the micro-movement of the dot caused by the sound waves hitting the glass the laser is passing through into audio. The demonstration on tv showed the dot hitting the inside of the window frame of a room way up in a skyscraper, so I don't think you have to be perpendicular to the glass.

 

[wwswimming]

my understanding of the process(after a couple of news reports and a pbs special) was that you shoot a laser at the window pane and hit some surface just inside the frame(blinds/curtains/sill/whatever.) You then use a high power camera lens to view the red dot of the laser. Software converts the micro-movement of the dot caused by the sound waves hitting the glass the laser is passing through into audio. The demonstration on tv showed the dot hitting the inside of the window frame of a room way up in a skyscraper, so I don't think you have to be perpendicular to the glass.

it sounds like "line of sight" is good enough. as long as
you can see the red dot created by the laser, your
instrumentation has a fair chance of detecting its
modulation by the vibration of the window.

 

[jmcoreymv]

I actually did my thesis design on a system that could do this with a very good SNR. The system itself could be very simple to design or very complex depending on the distance and other constraints and SNR you are hoping to achieve. It was also able to process 36 separate locations simultaneously which could later be run through an additional algorithm for an even higher SNR.

 

[Biftheunderstudy]

I'm not quite clear on commercial laser eavesdropping devices but the actual ones use something called a laser interferometer. Basically it involves 2 high reflectivity mirrors placed in series with each other with a space in between that is known. Then a laser is sent through the mirrors and on to the window or surface. The return path of the laser from this surface then interferes with the incident light in a way that is related to the displacement of the surface ie vibration ie sound.
To reproduce the sound from the surface, a modulation is put on the laser and then the return signal is demodulated to get the sound.

This is at least a very rough idea of how to do it, in practice it is much harder and there is a lot more to it. I'm also not sure if they use this set up of interferometer as there are a few other configurations that might work better.