Support & FAQ

A. Audio files downloaded from subscription services like Apple Music or other DRM-protected audio files (.m4p etc.) cannot be parsed or separated due to legal and technological copy protections.

Please use DRM-free audio files (.mp3, .wav, .m4a, etc.) that you own or have permission to process.

A. Audio separation requires heavy calculations. The processing speed depends heavily on your device's CPU, GPU, and Apple Neural Engine (ANE) specifications.

When running a heavy model (6-track) for the first time, or if your device has insufficient free RAM, the process might be slow or the app might crash. We recommend closing other heavy background apps before starting separation.

A. AI inference models sometimes generate computational errors or static DC offsets at the boundaries of the audio end, resulting in popping or static noise.

To eliminate this, adjust the "Ending Noise Reduction (Difference Filter)" slider (recommended range: 2.0 sec to 3.0 sec) inside the app's General settings tab. The filter will automatically scan and apply sub-millisecond mute/fade curves to clean up end noises. This setting defaults to 0.0 sec (Disabled).

A. Click "Save Settings" at the bottom of the track mixer. The current levels, 10-band EQ, and Reverb states will be saved to your device as a local preset bookmark.

To apply this to any other separated song, click "Load Settings" at the bottom of the mixer, or click the bookmark icon located next to the Track Mixer title. If you transfer settings between 4-track and 6-track models, matching stems will sync automatically, and non-existing stems will be skipped.