Ingesting Discord Voice Data

Hi there, I'm currently investigating using Membrane to record voice data from Discord to mp3/ogg files uploaded to S3.

I want to create a separate audio files per line of speech from each active user.

My guess is that I'm going to need to create a custom Membrane Source that provides the opus packets and then run that through the Opus decoder?

Then have that run through the ogg/mp3 encoder and then finally to the S3 upload sink

I'm looking for help just validating my theory and potentially providing information as to which encoding or formats to use.

The overall project will require being able to playback each line of audio individually so I need to keep each input and output distinct from each other

16 responses