Dear reader,
I use the wavelet decomposition in the spectral features DSP and I suspect that the decomposition is either incorrectly depicted or calculated. See below an example of a second level decomposition. I would not expect the signals to flair out at the beginning and the end of the window. High- or low-pass filtering the signal does not change much. Is there an issue? Can I assume the calculated features are correct?
Hello @elitac,
FYI, we’ve asked our DSP team to have a look.
They’ll come back to you as soon as possible.
Best,
Louis
Hi @elitac ,
Thanks for reporting this. This is indeed a visualization bug. The feature calculation is not affected.
We have a fix, which will be rolled out soon.
Best,
Yan
1 Like
Thank you very much.