0

I have placed two microphones placed within 1 or 2 cm distance. They can receive and record sounds around them. I want to detect the distance of the sound source (specially human voice) and the direction of the sound source using those microphones. Is it feasible when the distance of the sound source is short (i.e. around 10 cm).

Fred
  • 9,562
  • 11
  • 32
  • 44
Sazzad Hissain Khan
  • 101
  • 2
  • 1
    two microphones can detect direction, not distance – jsotola May 07 '22 at 19:15
  • Any way to estimate distance from microphones? Even using more than two mic? @jsotola – Sazzad Hissain Khan May 07 '22 at 19:21
  • think about it ... a pair of microphones can be used to determine the direction of the sound source ... adding a third microphone gives you three pairs of microphones – jsotola May 07 '22 at 19:25
  • You could use the mics to determine the distance between each other... – Solar Mike May 07 '22 at 19:28
  • 1
    I suppose you could have multiple pairs of microphones arranged far apart and find the triangulate/find the intersection between directions indicated by each pair but they would need to be far, far apart or surrounding the target. – DKNguyen May 07 '22 at 19:29
  • What's the real problem you're trying to solve? Why do you want to measure the direction and distance of a human voice and in what environment? – Transistor May 07 '22 at 19:31
  • @Transistor I need to determine the distance of a human spoken voice, and so the distance of the speakers mouth from ears using earbuds worn. – Sazzad Hissain Khan May 07 '22 at 21:28
  • 1
    Yes, but *why* do you need to determine the distance? What use is knowing the distance from someone's ears to their mouth? Your question says the microphones are 1 - 2 cm apart - now you're mentioning earbuds which would be the width of a head apart. Can you see why your readers could be confused? – Transistor May 07 '22 at 22:04
  • @Transistor each earbud piece has two microphones built in. I am interested to estimate distance from one earbud not both earbuds. And the purpose of the question is confidential and I don’t want to disclose it here. – Sazzad Hissain Khan May 07 '22 at 22:09
  • What we are suggesting is that, regardless of the gizmo you're trying to invent, you can't get a distance with a passive sonar system. If we had even a general idea of why you are trying to get measurements of head size / shape, we could suggest much more fruitful methodologies to investigate. – Carl Witthoft May 09 '22 at 13:18
  • BTW, if you take a look at the wavelength of audible frequencies you'll see that you'd need to measure received signals to a tiny fraction of a wavelength to get even a vague measure of mouth to ear distance. – Carl Witthoft May 09 '22 at 13:19

3 Answers3

3

A sound source and listening device can only provide direction.

Multiple microphones is how passive submarine sonar works. This will give direction only. Note that microphones in a line can only provide the two mirror image directions, you can't discriminate from off-axis sounds from either side.

To get distance, you need an active sound emitter, and measure the time for the reflection to return. This is how radar works as well, but with radio waves.

Tiger Guy
  • 5,369
  • 6
  • 17
  • Well, in theory (hah) if you know the dispersion curve of the media (atmosphere) and the structure of the source signal, you could analyze the relative arrival time of different frequency components. I wouldn't put a lot of faith in this approach. – Carl Witthoft May 09 '22 at 13:14
1

If the distance you want to measure is small than maybe you can use one of the microphone as a sound source.

But using a active sound emitter would be a better option.

effective-energy-356
  • 45
  • 1
  • 10
0

For an unambiguous distance in 3d space you need 4 mics. The spacing needs to be a reasonable proportion of some function of the wavelength and the distance. The more mics you add the better the error.

Thinking about it there are at least two approaches. The first, which is generally applicable in open air, is to use two widely separated intensity probes (4 mics each), to get a vector from each to the source. Where the vectors meet is the location of the source.

The second in a room would only work with non stationary signals. Simply put a mic in each corner (ie 8 mics again) of the room and use the cross correlation between the channels to find the time delay. Then do some geometry.

Greg Locock
  • 144
  • 1
  • 7