mattpointblank’s avatarmattpointblank’s Twitter Archive—№ 21,554

  1. …in reply to @FlynnMcLean
    @FlynnMcLean @hiattb I wonder what size training dataset you'd need for this to be feasible... eg Get Back had hundreds of hours of audio alone (besides all the wider Beatles catalogue). Or perhaps separating guitars from vocals is universal enough for most music?
    On twitter.com Twitter logo mattpointblank’s avatar Mood -2 🙁