[OT] Standing waves (noise cancellation effect)

Sep 30 2008 | 12:45 am
Hello Folks!
The project from the outer limits has come upon me. First it was a multichannel installation with the sound circling around the speakers. I convinced them to use 5.1 surround, which has worked out fine.
Now, they want to have two speakers playing standing waves (i.e. matched frequencies) that cancel each other out. I understand how oppostie waves would cancel each other out, but how does one produce matched frequencies that would cancel each other out? Is it even possible. I'm guess that they want two frequencies to be heard in one part of a room, and then have it not heard in the other. Can it be done with any old sound? How would one process a particular sound to have this effect. Is there a mathematical formula that could be imposed in a patch?
I recall asking a friend a long time ago, that if opposite waves cancelled each other out, then couldn't you make an electronic device that would take sounds that were going on in an environment, and broadcast back the opposite waveforms in real time. This was a really long time ago, when we lived next door to a couple who were always fighting, and after fighting would blast hip hop music that would keep going until all hours of the morning, Of course, we laughed at the idea.
Any information would be highly appreciated.
Best, John

• Sep 30 2008 | 1:45 am
Quote: johnhudak@gmail.com wrote on Mon, 29 September 2008 18:45 ---------------------------------------------------- > Hello Folks! > > The project from the outer limits has come upon me. First it was a > multichannel installation with the sound circling around the speakers. I > convinced them to use 5.1 surround, which has worked out fine. > > Now, they want to have two speakers playing standing waves (i.e. matched > frequencies) that cancel each other out. I understand how oppostie waves > would cancel each other out, but how does one produce matched frequencies > that would cancel each other out? Is it even possible. I'm guess that they > want two frequencies to be heard in one part of a room, and then have it not > heard in the other. Can it be done with any old sound? How would one > process a particular sound to have this effect. Is there a mathematical > formula that could be imposed in a patch? > > I recall asking a friend a long time ago, that if opposite waves cancelled > each other out, then couldn't you make an electronic device that would take > sounds that were going on in an environment, and broadcast back the opposite > waveforms in real time.
Probably not - the moment that any transducer could recognize the waves that are going on in the room, they'd have already happened, so any cancellation effort would be delayed slightly by the time it output the opposite wave. Buuut, I'm pretty sure this is how noise reduction headphones work, so who knows? I think the NR headphones may only work because the microphone is on the outside of the headphones, and the speaker is closer to your ear than the microphone, so you already have a delay (albeit very, very short), from the time the sound reaches the transducer to the time it reaches your ear. In an acoustic space, where a sound could be coming from anywhere, and the transducer could be anywhere else in relation to the listener, I doubt it would ever work.
As to standing waves.... It seems like it'd be something fun to experiment with. Perhaps two facing speakers, one playing the inverse of the other one? You'd only get the null effect in a very small area of the room, directly in between the two speakers, though.
I think your best bet might be to research something I've seen called Hypersound - it's extremely directional sound, created using stacked ultrasonic transducers. When the sound from the two ultrasonic waves interacts, it turns into audible sound, but very, very directional. I have absolutely no idea as to the algorithms needed to produce the two ultrasonic waves, though. or how to control them.
• Sep 30 2008 | 1:56 am
On Mon, Sep 29, 2008 at 9:45 PM, mushoo wrote:
> > As to standing waves.... It seems like it'd be something fun to experiment > with. Perhaps two facing speakers, one playing the inverse of the other > one? You'd only get the null effect in a very small area of the room, > directly in between the two speakers, though.
yes! this is the idea. but how does one get the inverse of an audiofile? that is what they were trying to explain to me...that you would get the null effect in only one spot (maybe a painted "X" on the floor?). but i don't really understand how to get the inverse. it certainly isn't the reverse.
thanks for the thoughtful response!
• Sep 30 2008 | 2:01 am
Most audio editing programs (even the cheapo ones) have an 'inverse' function. If you wanted to do it to something coming in from max's ADC, it'd be something like:
[adc~] -> [* -1.0] -> [dac~]
But I would also run your regular, non-inverse signal through a [* 1.0] object, just to keep the (very minimal) signal delay the same (running a signal through any sort of processing is going to invoke some form of delay, though it may be something as small as .01 milliseconds or some such.)
• Sep 30 2008 | 2:14 am
Quote: MuShoo wrote on Tue, 30 September 2008 13:45 ---------------------------------------------------- > As to standing waves.... It seems like it'd be something fun to experiment with. Perhaps two facing speakers, one playing the inverse of the other one? You'd only get the null effect in a very small area of the room, directly in between the two speakers, though. ----------------------------------------------------
This is an old recording trick to deal with singers who won't wear headphones. You stand 'em between 2 speakers wired out of phase (the easy way to do it!), and put the mic at a null point exactly equidistant from each speaker.
Works best with point-source speakers, ie a single fullrange driver, or Tannoy's dual-concentric speakers or such. You get pretty good cancellation as long as you don't have any significant wall reflections, which really spoil it. Note the null "point" is actually a line.
Your head is a stereo receiver though - if you can hear the flyback transformer whine of an older TV or anything else with a high-pitched tone, you can achieve a cancellation of the (15kHz?) tone by subtly turning your head. There will be positions where your ears are receiving out-of-phase signals due to their differing distance from the single source.
This is the primary way we determine sound directionality, which is why panning techniques relying on delaying one side of a signal sound rather more realistic than simple attenuation of one side (panpot).
It also makes it more difficult to achieve a good null for human listeners with out of phase speakers - the listener must be oriented with the axis of their ears perfectly along the null line, and even then it's not as good as you'd get with a more point-like microphone.
• Oct 07 2008 | 8:51 am
> Now, they want to have two speakers playing standing waves (i.e. matched > frequencies) that cancel each other out. I understand how oppostie waves > would cancel each other out, but how does one produce matched frequencies > that would cancel each other out? Is it even possible. I'm guess that they > want two frequencies to be heard in one part of a room, and then have it not > heard in the other. Can it be done with any old sound? How would one > process a particular sound to have this effect. Is there a mathematical > formula that could be imposed in a patch?
If you're not already aware of it, you should check out La Monte Young's work - his 'Dreamhouse' music is entirely based on this principle (and has been for decades!). See his book 'Sound and Light' written with Marian Zazeela for detailed technical discussion of what he does and why...
But you don't need to do any processing to get the effect you're talking about. Use two sine-tones (low ones work well) and project one from each speaker - there will be areas in the room where they cancel, other areas where they reinforce and still other areas where they appear to produce extra tones that aren't either of the originals. Pointing the speakers at odd angles (for instance at the walls) and turning it up loud both help. Experiment with different interval sizes: Young concentrates on numerically simple frequency relationships, but dissonant intervals produce their own kind of effect.
In his Dreamhouse music, Young often uses several harmonic partials (3 - c. 20), projected from multiple-speaker systems: as the listener walks thru the space, different sets of tones cancel/reinforce each other, so what you hear depends greatly on where you are. Higher tones can also produce dizzying phase-differential effects so they appear to be panning wildly around the room (tho' of course, they're not actually moving at all).