8
$\begingroup$

I initially became interested in the idea of acoustically tracking birds when in San Diego where there are a lot of wild parrots. Their calls are very impulsive, and combined with how loud the birds are I hypothesize they would be straightforward to localize with a passive acoustic array and time delay beamforming.

I would like to develop this project further as a teaching demonstration, but have since moved to a place without parrots. Would time domain beamforming be effective with the less impulsive calls I hear from many songbirds?

I had in mind a simple delay and sum beamforming strategy. A broadband signal should allow for source localization where the angular resolution would grow with sensor spacing without aliased direction of arrivals (assuming coherent arrivals on each phone). I am hoping that an array designed assuming impulsive arrivals will greatly simply the manufacture and calibration process.

I have no experience working with natural sound sources, but in detection theory this question is addressed by the ambiguity function of the source signal. If the source signal has no delay sidelobes, a delay and sum beam former gives a unique angle of arrival without consideration of sensor spacing. Given the great diversity of bird calls, are there significant differences in localization performance predicted by the ambiguity function?

Improve this question
$\endgroup$
2
  • 1
    $\begingroup$ Can you clarify if you plan to do proper beamforming (coherent sum of delayed microphone measurements) of direction finding using TDA? $\endgroup$ Commented Jun 25, 2022 at 17:34
  • $\begingroup$ Please do not edit OP with another question. Better create a new question. $\endgroup$ Commented Jun 26, 2022 at 15:38

1 Answer 1

Reset to default
2
$\begingroup$

You should be able to do time-delay-beamforming for any signal that satisfies the spatial sampling theorem: microphone distance < frequency_max/2.

The direction to the sound source is then where beamformer response is maximal.

As you may know, the primary objective of a beamformer is to suppress signals/noise from other directions and to reduce the noise in the direction of interest where you get a cleaner version of the timeseries. This does not depend on the type of signal.

Improve this answer
$\endgroup$
3
  • $\begingroup$ Thanks for your response, it helped me to significantly clarify my own understanding of my question. I hope the edits that I made have clarified what I am trying to get at. $\endgroup$ Commented Jun 26, 2022 at 15:35
  • $\begingroup$ I think this is not quite a complete answer. In real-world scenarios the beamformer response can be much more ambiguous - this means many lobes, and it's not quite clear which one to pick. This will be even worse for signals with more ambiguity (e.g. simple tones rather than broadband). Thus I think the OP is correct that broadband impulsive parrot sounds should work well, and others (e.g. pigeons) less well. I don't have a complete answer since I haven't worked on this in practice. $\endgroup$ Commented Jun 26, 2022 at 15:59
  • $\begingroup$ Not sure if we talk about the same issue. A time delay and sum beamformer simply sums all sensors in phase maximizing the response in a given look direction. This does not depend on the waveform. In fact the waveform of the signal should be the same only less noisy. $\endgroup$ Commented Jun 26, 2022 at 16:01

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.