Silent Sound Relication

Roger,

I'll post a document in alt.binaries.schematics.electronic that might be useful.

Bob

"BobW"

... Phil

OK. I have subscribed to alt.binaries.schematics.electronic, no problem.

Please let us know when the circuit is posted.

Thank you for your help.

Roger Tate

What are you using for the mixer? Seems that most use the nonlinear effects of AIR as the mixer.

If it's loud enough to compress air so it's significantly nonlinear, I wonder what it's doing to my ears.

Yes it is. You're not modulating the difference frequency. You're modulating the individual ultrasonics so that when they mix in the air, they result in audio we can hear. There's considerable "secret sauce" in how you pre-distort the modulating audio so it comes back the way you want it.

Hey Roger, Not to nit-pcik -- but I'm wondering if 'relication' is actually a word. Is that what you intended, or did you mean 'replication'?

I'm asking because when I searched Google (I'm always on the lookout for new words!), there are indeed lots of hits. However, the major dictionary sites don't list it. (?) Maybe this is a new word in the progress of being born? Here (USA), I don't think I've ever seen that word before.

Thanks.

Do you have a reference, where it is described in more detail? I've found this one:

but looks like it works with only one speaker and not with ultrasonic frequencies, and I doubt that you can hear the encoded signal.

I think it would be a good idea to try it in software, first. I've written a small command line tool for AM and FM modulation of WAV files (floating point encoded, which can be created and played back with the beta version of Audacity)

You could use my program as a base for implementing more interesting concepts. It compiles for Windows with Visual Studio (there is a free Express version from Microsoft), or with any other C++ compiler.

But if you need higher carrier frequencies, you'll need some external hardware and good ultrasonic speakers.

On a sunny day (Sat, 11 Dec 2010 16:36:57 +0100) it happened Frank Buss wrote in :

I remember from some earlier publication that the demodulation happens in the brain / ear because of non-linearity. The slightest non-linearity will cause frequency products and thus sound in the audible range. I would probably try myself with DSB first, as it is much easier to generate (diode bridge) then SSB, and should give the same results, but I have not tried. Audio amp, some 1A or more diodes, transformer, speaker. One channel the say 15 kHz carrier, other channel the audio, feed to bridge. No low level stuff, speaker connected to bridge. But, again, have not tried it, could work better with a few piezo tweeters...

That technology was demonstrated many times, but I have never seen it deployed in commercial installations. Perhaps, the required intensity of ultrasound is way beyond the safe limits. Besides, the path loss constant of the ultrasound in the air is really high (~meters) and strongly dependent on humidity, etc.

At one time I even ordered the "ultrasound speaker" from and paid somewhat $1k for that; but after a year of waiting I had to request money back.

They claim a non-linear mixing in the air. If it is due to non-linear mixing in the ears, that means SPL well over 100dB. You don't want to be exposed to that.

(diode bridge) then SSB,

This is exactly what they claim to do. Two intersecting beams, one with DSB signal, the other with Carrier signal, and some predistortion by DSP. The high intensity area at the beams intersection is where ultrasound is converted to sound.

I think it is one of those magic things that everybody had heard of, but it doesn't really work.

Vladimir Vassilevsky DSP and Mixed Signal Design Consultant

Yikes, 547 piezo elements to make a very, very bad speaker. The paper doesn't mention how much power is needed.

I wouldn't want to be blasted by 500+ ultrasonic transducers. I doubt that any museums or vending machines would care to go to this much trouble. I think I read about one public demo somewhere.

John

I have heard a system installed at University of Florida. Part of a system in a lounge area where they have 7 or 8 tv's mounted on a curved wall. The speaker system is mounted in the ceiling and point down. You must stand in about a 2-1/2 foot circle or you do not hear any sound. You can walk from circle to circle and hear a different tv. MikeK Found a few companies;

IIRC one of the big Japanese companies used it at a trade show. They modulated one beam and the other was used to demodulate. They were mixed by aiming the beams at a sheet of polystyrene, which then "sounded".

I first heard the amazingly localized sound from the mixing of two ultrasonic sound beams at the General Electric exhibit at the 1964 World's Fair. It was fairly new then, but well understood. The inherent nonlinear nature of sound propagation in an ideal gas seems to be unappreciated by most. See Fundamentals of Physical Acoustics by Blackstock for a good explanation. (Some understanding of partial differential equations is required - probably why so few people understand it.)

Glen

Air is pretty linear; the nonlinear part is the ear.

And I also wouldn't want to be hit by significant ultrasonic energy.

Thanks, Rich

Relication: Noun - The process of placing phony artifacts at an archaeological dig to achieve fame and fortune.

;-) Rich

Piltdowning: verb - The act of placing phony artifacts at an archaeological dig to achieve fame and fortune.

;-)

Aw, ya beat me to it. I wanted to use that as a reference: "see Piltdown Man"; I thought of it right after I hit "send." )-;

;-) Rich

I just realized that sound waves _do_ interfere, but that doesn't cause heterodyning, or does it? ISTR hearing two oscillators nearing each other in frequency, and hearing the beat note on top of them.

Now I'm confused, but I'm still a little skeptical that you could accomplish that with ultrasonics...

Thanks, Rich

ish

It's easy to have e.g. two wind instruments play a similar note, but what you're hearing is the amplitude beat, not the true LF subtraction of the tones (which would of course be infrasonic and inaudible, though it would be sensible if strong enough -- you'd be long since deaf from the din of so many oboes though :-) ).

It's funny... the ears are fundamentally a frequency-domain instrument (little hairs vibrating at various frequencies), but two frequencies which are spaced closer than the bandwidth of a given hair will not be differentiated. They seem to have the same pitch. Instead, only the apparent amplitude modulation is recieved, and only if the amplitude varies slower than the response time of the system (i.e., ~10ms). Which, bizarrely, is what you can see on the oscilloscope -- this type of signal actually emphasizes the time domain, rather than frequency domain, characteristics of the ear. Note also that the phase change (between amplitude peaks -- remember, this is essentially an AM suppressed carrier signal we're talking about) goes unnoticed, because the ear's amplitude sensing is incoherent (at least for some frequencies -- actually, low frequencies might have some coherency, which makes for interesting possibilities).

Tim

There is a guy that shoots an ultrasonic beam at a person hundreds of feet away, and can superimpose (or by other means) a human hearing range stream/sounds over that that allows you to hear a whisper that far away from the guy with the device, even in a high ambient noise setting.

It is accomplished through a modulation scheme that essentially AM modulates the ultrasonic "carrier".