First, I would like to say that this is a really nice package. I am learning about audio processing, and it has been really helpful for now.
However, I am working on a project where I would like to compute 1/3 octave band filtering over a sliding window, as it could be done for a MEL spectrogram. I think that the results will actually be really similar to those of a MEL spectrogram (correct me if I am wrong), but I am worried about spectral leakage with the sliding window. Would that be a problem for 1/3 octave filters as it is for MEL spectrograms?
I think that the answer must be "yes" if you use FFT at some point of the algorithm, but I have not been able to find this computation. It seems natural for me to think that FFT must be always calculated before both, octave band filtering and MEL, but I am not sure...
Hello,
First, I would like to say that this is a really nice package. I am learning about audio processing, and it has been really helpful for now. However, I am working on a project where I would like to compute 1/3 octave band filtering over a sliding window, as it could be done for a MEL spectrogram. I think that the results will actually be really similar to those of a MEL spectrogram (correct me if I am wrong), but I am worried about spectral leakage with the sliding window. Would that be a problem for 1/3 octave filters as it is for MEL spectrograms?
I think that the answer must be "yes" if you use FFT at some point of the algorithm, but I have not been able to find this computation. It seems natural for me to think that FFT must be always calculated before both, octave band filtering and MEL, but I am not sure...