Code review comment for lp:~mixxxcontributors/mixxx/features_vamp

Hi Tobisas,

thanks for all the help!

About BeatMatrix::getBpm() :
Actually it is true that one should need the full lenght of the song to compute correct bpm. Anyway, this is not the case for tempo and beat tracker from QM. If you use them, the first beat coincides (more or less) with the first nonzero samples, while the last one is always at the end.
I did some test and found in any case that temptracker:0 calculated bpm (got from BeatMatrix->getBpm() ) is more or less the same from legacy mixxx bpm detection from soundtouch.
I know that this is not a proof, but at least it is a suggestion :)

About average of all local BPMs vs BeatMatrix::getBpm() :
In this case, I think their values coincides. "I have discovered a truly marvelous proof of this, which this margin is too narrow to contain" (cit. P. de Fermat) :p

About local/global BPM and sync:
I did some ugly change in bpmcontrol::slotControlBeatSync to let it work by mean of "local" BPM. That is by using BeatMatrix::getBpmRange. If "local" BPM varies too much, sync will be lost in few beats.