How to Separate Vocals and Background Audio
Choose a suitable source and separation scene, then review each MP3 or WAV track before using it.
Artplayer AI읽는 시간 5분
이 페이지의 내용
시작하기 전에
Sound separation can make a mixed recording easier to edit. It produces new audio tracks from the sounds in the source, but overlapping voices, music, and effects may still leak between tracks. Start with a clear file, choose the scene that fits it, and listen to every result before using it.
Prepare an audio or video source
Upload MP3, M4A, or WAV audio, or MP4, FLV, TS, AVI, MOV, WMV, or MKV video. A video needs a usable audio stream. The app accepts files up to 2 GB and 60 minutes, and checks the actual media before processing.
Use the cleanest source available and keep the original. Heavy compression, clipping, noise, and sounds that overlap closely can make separation less reliable. A video is accepted as an input, but this tool delivers audio files rather than a new video.
Match the separation scene to the recording
Choose General audio for mixed recordings or Music for a song. Both scenes deliver a voice track and a background audio track. Choose Drama or Narration when you need three categories: voice, music, and sound effects.
A scene describes the sound categories to request; it is not a guarantee that every category is present. A quiet or empty music or effects track can be a valid result when that sound is absent from the source. The tool does not identify different speakers individually.
Choose MP3 or WAV and review the quote
MP3 is compact and convenient for previewing or sharing. WAV files are much larger and often easier to use in editing workflows, but choosing WAV cannot recover detail already missing from the input. Every delivered track uses the selected format.
The credit rate is 2.8 credits per measured input minute, with a 3-credit minimum per task. Two or three tracks, the scene, and the output format use the same rate. Check the exact whole-credit quote after the file is validated and before starting.
Listen to every track before editing
Preview the voice track for missing words or audible music, and the background, music, or effects tracks for remaining speech. Listen at the start, at busy sections, and near the end. A result that sounds clean in one short passage may contain artifacts elsewhere.
Download each track separately and open the files in your editor to check timing and playback. The original is available for 24 hours after upload, while finished tracks are available for 7 days. Save the files you need before they expire.
- Confirm the chosen scene returned the expected two or three tracks.
- Check for voice leakage, missing sounds, artifacts, and timing differences.
- Keep the original beside the downloaded tracks for comparison.