Downsampling

[filipandrei87]

How could I downsample the input? This works great on music, but if you are only interested in human voice, you only need to go with 16000hz.

In this case, this should also optimize overall speed.

[worldveil]

This would be cool to have. I think that downsampling (decimation) might be useful for some users, though you are the first to mention it.

For those reading, here's the basics.

I'd welcome a working PR here, though I'm not sure it's mission critical.

[filipandrei87]

This will be especially useful for those that want to ingest content coming from TV or similar. Anything that is speech heavy. In this scenario, the overall performance should be increased significantly.

Unfortunately, I am not familiar with Python so can't be of much help to you guys.

[lacatus]

The best option to downsample an audio file is by using FFMPEG first and afterwards pass it to Dejavu