Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Directional (ultra)sound speakers

  1. Nov 10, 2008 #1

    A while ago, we did a project at my university to construct a directional sound speaker; eg a speaker that does not emit sound in an omnidirectional way as usual but in a straight line (or at least as much as possible).

    The goal was to research different approaches, choose the most appropriate one in our case and try to construct a working speaker.

    We had found two approaches:
    1. Using multiple sources and creating an interference pattern which causes the sound to destructively interfere at the sides of the speaker, and constructively interfere at the front.
    2. Using ultrasound as some kind of 'carrier' for the audible sound; the theory being that ultrasound does not spread out in space so much as audible sound does.

    The second option seemed to work much better, judging from devices that were on the market during our research, but also seemed way to complicated for us at that time. So we went with option 1.

    We used 52 speakers (cheap piezo speakers) and arranged them in a hexagonal pattern, the crucial part being that each speaker was half a wavelength (6 cm) apart from its neighboring speaker.
    This can be more easily understood if you imagine a line of speakers instead of a 2-dimensional pattern: the sound will then only be directed in the plane of the line.
    If the speakers are half a wavelength apart, then speaker 1 and speaker 2 will cancel each other out at the sides, while speaker 2 and 3 will also cancel out, etc...
    The following image portrays this effect for 2 speakers (warning: it's a fairly large image):
    The red circles represent troughs, while the black circles represent peaks. The resulting sound wave has been drawn right in front of the speakers (double height) and at the sides (canceling out exactly).
    It can easily be seen that in front of the speakers, the troughs and peaks meet up with other troughs / peaks, and the result is again a wave alternating between troughs (red) and peaks (black), while at the sides, the peaks meet up with troughs and the troughs meet up with peaks thus canceling out.

    Anyway, we managed to get it to work in the end. The effect is fairly obvious and there is a very sudden increase in sound level (about 30-40 decibels at most is what we measured) right about 20 degrees in front of the speaker (0 degrees being right in front of it, 90 degrees at the side). There is a very sharp 'edge' where the sound suddenly increases quite a lot; we have measured about 60 dB at the sides (90 degrees) and about 110 dB right in front.

    However, while the difference in sound level is impressive for a 'simple 52-speaker' design, the 60 dB at the sides is still very much and the sound is still completely audible.

    Now finally on to my question (hehe). Last night on TV I was watching a show called Future weapons and a similar sound-directing 'weapon' was shown. Actually 2 were shown: one had properties similar to ours (except it went about 155 dB at the front so it was simply MUCH more powerful) and another was not so powerful but it had PERFECT directing abilities.

    My question is about the latter: it was a square, very thin speaker, about 20 cm squared, and could direct audible sound perfectly. The guy aimed it away from the camera, turned it on and you heard nothing. Then he aimed it at a wall and the sound could be heard coming off the wall (it reflected). When he finally aimed it right at the camera the sound was crystal clear, like it had only just been turned on.

    Now, at first I thought it was using the same principles as our design (interference pattern) but they said later it used ultra sound; sadly they did not elaborate how it worked...

    So: does anybody know anything about directing sound using ultrasound? How can it be explained that the device uses ultrasound, yet the sound is audible and at normal frequencies? Obviously the sound we hear is not the ultrasound but what is the purpose of the ultrasound then?

    I have searched around (including the research we did in our project) and found a few products on the market (audio Spotlight and hypersound something off the top of my head) that claimed to work well, but nowhere could I find more than a very vague description of how exactly it worked.

    I'm simply interested in this and I am wondering if it is something very hard to understand or not. I may find that it is beyond my knowledge of sound but please try me :p

    So if anybody knows anything about this, any links where it is explained maybe, could you please point me there?

  2. jcsd
  3. Nov 11, 2008 #2
    Hi Nick89 and everybody!

    No precise idea how ultrasound should carry or focus normal sound. In both cases, it needs a non-linear effect, either during propagation or at the ear or microphone.

    You mention a wall... If the show was within a room, then it was very probably a crook, as sound reflections are very important against solids (almost no absorption) and any propagation there involves hundreds of paths with many reflections, so no directivity can be observed in a room.

    To get a high directivity, letting many speakers interfere is the hard way... Using a large reflector (parabola) is easier. Reuse a satellite TV dish and put your single loudspeaker at the focus.

    If you want very low side lobes, then the "illumination function" is paramount. Synthetic aperture (interference of many speakers) is much harder than a parabola precisely for that very purpose. Techniques are well known for electromagnetic antennas: use an offset primary source, illuminate the dish with a smoothly vanishing intensity at the edge. You may also cut corrugations at the edge of the dish - so that the dish's surface doesn't end steeply - and even add a baffle, possible absorbing and corrugated. Some telescopes also incorporate similar features.
  4. Nov 11, 2008 #3
    Thanks for your response.

    I don't know what you mean with it 'being a crook'..? Do you mean it was faked? It was certainly not fake. It was indeed in a room but he pointed the speaker out of the (open) door when he pointed it away from the camera, so the sound was not reflected.
    When he pointed it towards a wall in the room, the show host said he could hear it coming from the wall (obviously you can't hear that effect through the tv, but you could only hear the sound very slightly). When he finally pointed it straight towards the camera it was as if he had only just turned it on (but I'm certain he had it on all the time).

    With our directional speaker (many speakers + interference pattern) we did observe very little directing effect inside a room, but as soon as we took it outside it worked much better, so the directing did not work inside.

    However, I believe this was because our speaker was not nearly as 'directed' as the speaker in the show. If the speaker would be perfectly directed there would only be one reflection before it reaches you (obviously, after that it will reflect further but you will probably hear the first reflection much more).

    We did consider a parabola in our first approach but decided not to do that. It would have been very hard to construct (a satellite dish as you suggest did NOT work at all, it was probably much too small) and we could not think of a way to bring the speaker in the focus of the parabola, without impeding the sound waves because there's a giant speaker in the way!

    We could have used a single piezo speaker which is very small ofcourse, but even then, the speaker would have to be right in the middle of the parabola, exactly where you want your sound wave to emerge from.
    Also, besides directionality, we also wanted it to have a bit of a punch, so it could theoretically be used as a non-lethal weapon. Ok, the 110 dB we reached (at very close range btw) is not nearly enough to cause someone to drop his weapon out of pain in his ears, but I am sure there are much louder speakers we could have used (we just needed to stay within budget)...
    There's no way you're going to reach an agonizing sound level with a single piezo speaker...

    What you say about ultrasound rings a bell! The non-linear effect is something I am sure we came across during our research, we just didn't understand what was meant by that. Can you elaborate?
    Last edited: Nov 11, 2008
  5. Feb 7, 2009 #4
    http://www.holosonics.com/technology.html [Broken]

    I was searching for the same thing, took me a while to find it but there it is! :)

    Google is my friend :))
    Last edited by a moderator: May 4, 2017
  6. Jan 15, 2010 #5
    I found this nice pdf - a thesis on "Sound from ultrasound : the parametric array as an audible sound source."

    http://dspace.mit.edu/bitstream/handle/1721.1/7987/53014459.pdf?sequence=1 [Broken]

    If that url does not work, it lives on this page: http://dspace.mit.edu/handle/1721.1/7987

    Very interesting reading, enjoy!
    Last edited by a moderator: May 4, 2017
  7. Jan 26, 2011 #6
    I'm wondering, with this technology in hand, can your invert this technology into a Directional microphone? Also, I would like to know if the panel can be made into fabric-like of medium sizes. I'm not looking for a very small sizes, unless if the sizes of the single speaker happens to be very small and can break apart into a fabric.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook