drumscript.audio_processor.stem_splitter

This module uses the demucs library () to extract stems from multi-layer audio files. It also contains functionality for re-mixing stems to create drumless backing tracks for user export. Running: python3 -m drumscript.audio_processor.stem_splitter path_to_audio_file, <output_path> It supports generating ‘drumless’ tracks, isolating specific instruments, and format conversion on demand.

Functions

extract_drum_stem	Legacy wrapper for the transcription pipeline.
mix_stems	Uses soundfile + numpy so the WAV output path is ffmpeg-free.
separate_audio	Separates a full audio track using Demucs and processes the outputs based on user-input args (optional)