Generating files that overlap with the next file in the sequence #

Recent deployments with SM4s have been generating files that overlap. For example, if there are two files, 20171201_060000.wav and 20171201_120000.wav, the first file will have a duration of 3617 seconds – that is 17 seconds past the start of the next file. Visual inspections of the neighbouring files reveal that the end and start of the overlapping files DO NOT match and are not duplicate data. In each of these cases nearly every file in a deployment overlaps!

Our current theory is that the sensors maybe oversampling (sampling at, for example, ~22075Hz rather than 22050Hz) and stopping once the “correct” number of sample have been collected, earlier than the file should have stopped. This theory explains why, despite a logical overlap, the collected data does not overlap.

Status #

Major Problem

Status with vendor #

Not yet contacted WA

Effects of the problem on common tools #

Acoustics Workbench (Ecosounds, A2O): #

• The website rejects files that overlap more than 12 seconds. Due to the distributed nature of harvesting protocols some files in a harvest a harvested and some are not – leading to a lengthy process required for deleting the files from Ecosounds, trimming them, and re-harvesting.

Workarounds: #

Overlap report script: https://github.com/QutBioacoustics/Ecoacoustics/blob/master/Scripts/generate_overlap_report.ps1

Trimming script: https://github.com/QutBioacoustics/Ecoacoustics/blob/master/Scripts/trimEndOffAudioFiles.sh

Example files