Hi all,
Just wondering about the feasibility of this.
Basically would it be possible to use a Arduino plus plenty of head scratching etc as a bat detector?
Bat echolocations are above human hearing most of the time (in the UK, between 18 to 110Khz).
Bat detectors are already available which use a variety of systems to get the sounds to an audible range... either comparing the incoming frequencys to an internal frequency (so you hear the difference.. difference between 45khz and 43khz = 2khz...), extending the incoming signal by 10, for example (100khz = 10 khz).
Would it be possible to hook up an ultrasonic mike to the arduino and then process the information coming in using its processor to make something useful come out of the other end?
It'll be quite a learning curve for me, but that sorto the point!
I see what you mean, but the point would be what a microcontroller would then allow you to do.
For example, setting up a recording operation onto a minimum noise threshold and either/or a timer or a minimum noise stop, so that anything that went by could be recorded and potentially analysed on a computer later, storing info on an SD card or USB.
Some bat detectors use zero-crossing analysis of sounds in order to produce small computer data files which can be analysed to determine species. Others use more memory heavy .wav files to get a better sonograms - it'd be interesting to see what'd be possible.
The idea is, if you hook up a microcontroller, would this allow you to play around and add this extra functionality perhaps?
I reckon someones going to do it with microphone/smartphone setup sooner or later anyway, I just like the idea of figuring something cool out... any thoughts?
My immediate thought is that the Arduino is way underpowered for any sort of analysis, and a DSP architecture would be better.
The ADC is too slow to sample the sound directly, so you' have to divide the frequency down first.
Recording too would probably hit the buffers of how fast you could write to external media.
You might use it to analyse repetition rates, but that would be about the limit of it.