Split Audio into Stems Online Free

Audio stems are individual components of a mixed song, separated into distinct tracks. A typical stem split produces:

Stem separation opens up a wide range of creative and practical applications:

• Remixing: Use individual stems to create remixes, mashups, and edits of existing songs.
• Music production: Sample individual elements — a drum loop, a bass line, or a vocal hook — for your own productions.
• DJ sets: Use acapella vocals or drum stems for creative DJ transitions and live remixing.
• Music education: Isolate individual instruments to study arrangements, transcribe parts, or learn by ear.
• Karaoke: Use the vocal stem for pitch correction practice or the instrumental for karaoke.
• Film and video: Extract specific elements for use in video productions.

Our stem splitter uses STFT spectral masking — a frequency-domain approach to source separation:

1. STFT Analysis: The audio is converted to a time-frequency representation using Short-Time Fourier Transform.
2. Spectral Masking: Frequency-domain masks are applied to separate energy into vocal, drum, bass, and other components.
3. Harmonic/Percussive Separation: Drums are separated using percussive transient detection; bass is isolated using low-frequency masking.
4. Reconstruction: Each masked spectrum is converted back to audio using inverse STFT.

All processing happens locally in your browser — no audio is ever uploaded to any server.

• Use high-quality source files: WAV or 320 kbps MP3 produces cleaner stems than low-bitrate files.
• Simpler productions work better: Songs with fewer overlapping elements (e.g., acoustic recordings) separate more cleanly than dense electronic productions.
• Centered vocals: Lead vocals panned to center separate most cleanly from the instrumental.
• Post-process your stems: Use our Noise Reducer to clean up any bleed between stems after separation.