Mirelo and the research lab Kyutai have released MuScriptor, an audio-to-MIDI model that transcribes a finished recording directly into separate instrument tracks—without first splitting the song into stems. The system takes a complete mix as input and outputs distinct MIDI tracks for voice, drums, bass, keyboards and other instruments, effectively letting producers convert any finished song back into an editable score.
Continue reading
The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in✓ Signed in — this article isn’t included in your current plan.