Since BirdNET filters if given latitude and longitude coordinates, a user providing a sound from beyond what we have hardcoded for lat/long in this script might have the non-detection hidden from them without understanding why. Moreover, in using this script for random tests on birds from across the globe, we'd likely prefer that it by default include all the classes in the model.
birdnetlib does warn that lat/longs are specified in its terminal output but that doesn't feel like quite enough, particularly given it doesn't explicitly state the ramifications of that for the classification.
Since BirdNET filters if given latitude and longitude coordinates, a user providing a sound from beyond what we have hardcoded for lat/long in this script might have the non-detection hidden from them without understanding why. Moreover, in using this script for random tests on birds from across the globe, we'd likely prefer that it by default include all the classes in the model.
birdnetlib does warn that lat/longs are specified in its terminal output but that doesn't feel like quite enough, particularly given it doesn't explicitly state the ramifications of that for the classification.