Summary
Our testing has found that the biggest variable on filesize is, not surprisingly, background noise.
When deployed in a quiet location we have found that both the configured Amplitude Threshold AND Gain setting control the resulting filesizes. Put simply, the higher the gain, the higher the Amplitude Threshold needed to reduce filesize.
We found selecting an optimal setting of Amplitute Threshold for a set gain to be very difficult as there is not a linear response, with the setting more of a “cliff face”.
Testing method
A set of 8 AudioMoths were set up to record UK bats for one night and placed next to each other in a very quiet location. The testing process we followed is summarised on this page.
On each night all the AudioMoths were configured with the exact same settings for Gain, and everything else, with only the Amplitude Threshold been varied over a range, across the 8 devices.
For each night the total size of the files on the mSD card were compared to the device that was recording continually, to allow comparison by percentage.
Results
Overall by far the the biggest effect on file size is background noise. A windy night in trees, or heavy rain will mean the Audiomoth records continually whatever the setting of amplitude threshold.
It was also noted that even on quiet nights there can be smaller variances between individual Audiomoths. This may be due to variability in internal noise of the device.
To allow comparisons between nights, the file sizes for a setting of Amplitude threshold are compared as a percentage of a similarly configured AudioMoth recording continually. The results are best summarised in the Graph below. Click to enlarge.
Download the entire dataset (in Excel format)
Discussion of results
The results show that the internal noise of the Audiomoth will result in continual recording at low settings of amplitude threshold, until a point is reached where small changes in amplitude threshold rapidly decrease the file size until about 10-15% of the size of continual recording. After this, the effect of further increases in amplitude threshold diminishes.
The setting of gain also has a huge effect on the “optimal” setting of Amplitude threshold. Or put another way, a amplitude threshold that produces tiny files at lower gain, will result in continual recording at higher gains.
Background
The ability to configure a Amplitude Threshold for recordings as added in the 1.4.0+ firmware released on 4th June 2020. To quote from the paper:
“amplitude threshold to be set such that only those audio segments (each32 KB in length) which contain a sample value greater or equal to the threshold will be written to the SDcard. This results in much smaller WAV files as silent periods are automatically removed. The lengths and positions of these silent periods are also encoded into the resulting WAV file so that its original length, and the relative timings of events within it, can be restored using an additional feature of the configuration app.”
“choosing the threshold value requires some care and can be done on a trial basis or through analysisof previously collected recordings. If the threshold is set too low the resulting recordings will likely contain toomuch background noise. If it is too high the recordings will likely miss some quiet bat calls. The AudioMoth red LED will only flash when it is actually writing data to the SD card and can be used to judge whether or not background noise is triggering the recording. A value of 512 works well in many cases. A future release of an online planning tool will assist in choosing the threshold value by allowing conventional files to be uploaded, and the effect of various filter and threshold settings experimented with before deployment.”
Further Work
Finish the dataset with a High gain recording (rain stopped play).
Repeat process in locations with more background noise.