Potential Audio splicing issue

[aserfass-msft]

I was looking at this spectrogram and the clear line in the middle bothered me. https://aifororcas.azurewebsites.net/detections/detection/4aae8258-d8e8-4256-96af-476a501f8376

A sudden change in noise should be rare/unlikely. I'm interested in someone double checking my assumptions to make sure I didn't miss something that explains this. I'm concerned the audio data is being spliced incorrectly (e.g. the sudden change is due to two audio clips from different time periods being combined out of order). It is possible this sudden change in the audio could naturally happen but then we would see that change in the original data (TS files).

I looked at what I think is the source data:
rpi_orcasound_lab
14 Sep 2023 06:10:46 PDT -> epoch 1694697046
6 hours of audio data from that time

If someone can confirm this (and I didn't make a mistake) then I think there is an issue with the code generating the spectrogram in OrcaHello

[aserfass-msft]

This is the spectrogram at the link above