Configure and process stems in Ultimate Vocal Remover

Step-by-step

1. Click Select Input and choose your audio file
2. Click Select Output to set the save location
3. Select a process architecture (Demucs, MDX-Net, VR, or Ensemble)
4. Toggle the GPU option if your computer has a dedicated graphics card
5. Click Start Process

Options

• Use MDX-Net for clean vocals with minimal artifacts
• Use Demucs for full stem separation (vocals, drums, bass, other)
• Use VR models for faster, lower-quality splits
• Use Ensemble mode to blend results from multiple models
• Use Sample Mode to process a short segment for testing

Tips

• Enable GPU processing to speed up the conversion process
• Use Sample Mode to test quality before committing to a full stem creation process