Acoustic detectors are widely used for the non-intrusive study of wildlife. It turns out that the data collected by these detectors have some similarities to radio astronomy data, which I’m much more familiar with. I’m interested to see how applicable some of the astronomical observation and analysis techniques are and if they can be used in eco-acoustics. This blog describes my work to build a detector to collect and analyse audio data to try out some of these techniques as a proof of concept. I’m focusing on the use of multiple microphones (two initially) and cross-correlation of their signals to improve sensitivity and signal to noise as well as provide estimates of signal location and speed. Audio data at frequencies up 48 kHz are being recorded at present but the techniques will also work at higher frequencies..
The detector is based on a Raspberry Pi computer with a pair of digital microphones attached. The cost of components is relatively small, ~$100, making it a nice entry-level device for anyone who doesn’t mind wielding a soldering iron.
In the next post I’ll describe some work on using two microphones to improve sensitivity and show some initial results. Following that, I’ll look into methods to measure location and movement of signals.