19
$\begingroup$

Has anyone shared their calibration of an AudioMoth automated sound recorder that has become popular as an inexpensive passive monitoring device with open-source firmware?

The manual specifications of audio quality include only the signal-to-noise ratio (SNR) at the one frequency of 1 kHz.

Many users may thus wonder about the calibration of a range of frequencies, meaning the voltage level (or digital level) that will be logged when using this device's microphone to record a specific physical sound level (dB SPL) at specific frequencies.

$\endgroup$
3
  • $\begingroup$ It depends on the specifics of the microphone which varies across models slightly (the Audio Moth seems to come from China in batches and have mics added and then cases added - all by small independent businesses who each do things slightly differently) . Do you have the serial number on the internal mic (the hole size seems to make a bit of difference too), failing that the model and case manufacturer that houses the Audio Moth module? $\endgroup$ Jun 26 at 5:21
  • 1
    $\begingroup$ They have been out of stock, but I am interested in getting some. Thanks for asking, @JiminyCricket $\endgroup$ Jun 27 at 15:50
  • $\begingroup$ All the audiomoth specs (plus technical info for the cases) are available on their site and/or in the user guide - openacousticdevices.info/audiomoth. These are the acoustic membranes that go across the hole in the 'official' case - groupgets.com/manufacturers/open-acoustic-devices/products/…. $\endgroup$
    – Carly Batist
    Jun 28 at 2:45

3 Answers 3

3
$\begingroup$

https://github.com/kitzeslab/audiomoth-performance/blob/main/report.md

I was wondering the same and found this report. Researchers for the University of Pittsburgh have essentially measured sensitivity, directionality patterns and frequency response of Audiomoths with different casings and in different environments across a wide range of frequencies. May it be of interest?

$\endgroup$
1
  • $\begingroup$ Could you add a summary of this report and say how it responds the question please? Or just put it as a comment of the question ;-) . $\endgroup$
    – Noil
    Aug 5 at 18:46
13
$\begingroup$

There is no simple way to calibrate an Audiomoth, as also the housing will have a significant influence on directionality and frequency response. This means that it should be calibrated by means of a soundsource (signal generator or similar plus a suitable wide-band loudspeaker) and a microphone with a known and trusted calibration chart (sound level meter able to work in the ultrasonic range or microphone with attached oscilloscope or some other recording device). With a setup in reasonable free-field conditions (away from the floor, walls and other reflective surfaces, ideally in a soundproof room) the Audiomoth and the microphone are mounted close together and a series of calibration sounds are played. These can be short pure-tone pips, or frequency modulated sweeps, depending on your analytical skills. The frequency response and clip level of the Audiomoth is then found for each frequency as the sound pressure level measured with the reference microphone minus the rms-level of the Audiomoth signal (assuming the wav-file is normalised to +/- 1).

$\ CL_{Am}(f)= L_{p,rms}(f)-20log_{10}(s_{rms})$

This gives the clip level at each frequency, which can then later be used to convert Audiomoth recordings to absolute sound pressure levels by reversing the equation:

$\ L_p(f)=20log_{10}(s_{rms})+CL_{Am}(f)$

Directionality is a separate issue, but essentially handled he same way: by calibration of the Audiomoth against a reference microphone, but with various angles of incidence. Note that the reference microphone should always face the loudspeaker, only the Audiomoth should be rotated (the reference microphone is directional too).

$\endgroup$
2
  • 2
    $\begingroup$ Jakob, are these equations in Matlab syntax? That's useful information to have for users. :) $\endgroup$ Jun 27 at 14:44
  • 2
    $\begingroup$ @etgriffiths I guess they are in Latex format for when the Bioacoustics SE will integrate this feature (bioacoustics.meta.stackexchange.com/questions/3/…) $\endgroup$
    – Noil
    Jul 9 at 23:14
7
$\begingroup$

Though this doesn't fully answer your question, I would point you towards this recent publication, which examines the use of a hydromoth - an audiomoth for underwater recordings. It doesn't calibrate the device, but instead focuses on acoustic metrics (e.g. acoustic complexity indices) that can be obtained from an uncalibrated device.

The abstract and intro read like a bit of a sales pitch, and don't really match the results section, which is where you should look to see a report on device performance. Some of it is worrying, including co-located devices recording different things (with one detecting/recording prominent dolphin sounds, and the other not, at the same time).

  • Lamont, T.A.C., Chapuis, L., Williams, B., Dines, S., Gridley, T., Frainer, G., Fearey, J., Maulana, P.B., Prasetya, M.E., Jompa, J., Smith, D.J. and Simpson, S.D. (2022), HydroMoth: Testing a prototype low-cost acoustic recorder for aquatic environments. Remote Sens Ecol Conserv, 8: 362-378. https://doi.org/10.1002/rse2.249
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.