Sampling rates for bats (UK)

Hi all

I'd be really grateful for any advice - is there a reason why more bat calls would be identifiable from recordings using a 250 kHz sampling rate compared to a 384 kHz sampling rate?

A lot of the literature recommends using 384 kHz for bats, however, I came across this from Somerset bat group: https://somersetbat.group/advice/which-bat-detector/audiomoth/testing/sample-rate/

Having observed some similar findings, I'm curious as to why this might be?

Many thanks.

4 comments

4 Comments

Alex Rogers

Jun 24

At 250kHz there is possibly less interference from SD card noise writes than 394kHz since they occur less frequently. Both 250kHz and 384kHz use a single ADC sample per WAV file sample. At 192kHz the AudioMoth will oversample, using two ADC samples per WAV file sample, so that improves the signal to noise ratio a little. There might also be differences in the classification algorithm and the generation of the spectrogram that lead to some differences.

Edited

s_per93

Jun 25

Replying to

Thank you for your reply Alex, really interesting with regard to the SD card noise. Would you expect the noise from card writes to also vary between different SD cards - SanDisk Ultra vs Extreme for example? I've used the Ultra cards at both 250kHz and 384kHz, but appear to get more classifiable calls at 250kHz.

Alex Rogers

Jun 25

Replying to

Yes, SanDisk Ultra are definitely noisier than SanDisk Extreme (red and gold). We only generally use SanDisk Extreme cards as they are fast, quiet and have low energy consumption.

Stazione Teriologica Piemontese

Jun 24

If in your study area, you have species that emit at frequencies above 110 kHz, if you use 250 kHz sampling rate you will not be able to record them. In the UK Rhinolophus hipposideros emits at around 110 kHz and there are no other species that exceed this value.

Ciao. Paolo