I have tried reading through both aubio and librosa papers but I'm not very familiar with the terms,
I'm writing a code that reads through a batch of audio files in .wav and .mp3 to create a metadata. The code is already retrieving the title, artist, length and other info from the file itself. Previously I used Mixmeister program to embed the BPM into .mp3 files but I cannot do so with .wav files.
So, I am trying to find a way that calculate the BPM of each audio file whether its .mp3 or .wav.
update
I ended up using the average tempo from the two method How does librosa estimate tempo?
The average of both results, returns a closer estimate to Mixmeister, but it there is a lot of octave-errors (some results needs to be divided by half to equal actual BPM, for example actual BPM 59 the method returns BPM 188) How can we improve tempo detection accuracy in Librosa?
Is there away to detect which songs needs to have its result divided in half?
Can you share with me the current method that you find most accurate in dealing with BPM or beats per minute of an audio file?