Current Species List

[stevedee]

My current Species List (based upon Location & Date) contains 110 species, but I get common swift detections even though this is not on the list.
Is this a problem or am I misunderstanding how this works?

[max-mauermann]

In theory your understanding of how this works is correct. You can generate a species list in the species-tab and then select the generated file as a custom species list in the analysis-tab and all species not on the list will be filtered out from the predictions.

So from what i understand, you created a species list using the species-tab with your location(lat+lon) and week and this list does not contain the common swift.
And when using this species list in the analysis-tab you still get detections for the common swift.

If this is the case i would convert this into an issue and investigate further.
Could you also provide an example recording and the lat/lon/week settings used to generate the list?

[stevedee]

Just to clarify; I used the Species List Tester on BirdNet-Pi (Nachtzuster) just to check what was currently being used as a Species List on BirdNet-Pi.

I thought I'd posted this in the BirdNet-Pi forum, so not sure how it showed up here. Does this affect your reply?

[max-mauermann]

I have to admit that i am not too familiar with how BirdNET-Pi works, so i can't tell what could be causing this problem.

If BirdNET-Pi is not directly using the BirdNET-Analyzer, including the postprocessing of filtering out according to the species list, i doubt this is an issue in the Analyzer.

Still we might add some tests to ensure the filtering works as intended.