Low-quality audio makes detection worse
doreen08 opened this issue · 1 comments
Hey, thanks for your sharing. It is a great work.
I found that when I use 16000Hz audio I get worse results than 22050Hz.（audio from the same music）
Inputs are all automatically resampled to 22050 Hz.
How can I do better when I only have low quality audio of 16000Hz.
Hi, Thanks for your interest!
In general, audio with a lower sampling rate (leading to lower Nyquist frequency) contains less information than audio with a higher frequency. It can automatically lead to somehow lower performance. However, In order to make the system work properly with the custom frequencies, the estimator sampling rate parameter should be set manually. To do that, make sure you change the sampling rate value here (line 56 of the BeatNet script to 16000. It will be added as an init input argument in the next release.