Audio separation is a preparation step for translation or dubbing. It can provide the original speech for script review and, with the appropriate scene, separate music and effects for editing. Music Remover does not transcribe, translate, generate a new voice, clone a speaker, or synchronize lip movements. Those steps happen in your chosen external tools or recording workflow.
What does audio preparation do for a dubbing project?
Start with a video you have permission to adapt and an audible source-language dialogue or narration. If the original project already has separate dialogue, music, and effects tracks, work with those directly. Separation is useful when you only have a mixed soundtrack, but the estimated tracks may contain remnants of one another.
Decide what the finished edit needs before processing. Translated subtitles can keep the original speech. A full dub normally replaces that speech with a new recording while retaining selected non-dialogue sounds. A translated narration over a product demo may need different pacing from the original. These are different editing choices; selecting a separation scene does not make them for you.
Voice separation keeps speech or singing according to the input. It does not select a target language or isolate one speaker from a conversation. Do not treat the downloaded voice as a completed dub: it still contains the source-language vocal content.
How do you separate the source tracks?
- Keep the source video as your timing reference. Note the start of important lines, speaker changes, and effects that should remain in the finished version.
- Upload a supported file up to 100 MiB, or submit a direct public HTTPS media-file link. Sign in to start processing. Private share pages and streaming watch pages are not direct media links.
- Choose the scene for the source. General audio · Audio produces voice and background tracks for dialogue and narration. Music and other non-vocal sounds are combined in the background result.
- Choose an audio format accepted by your editor or next tool. AAC, MP3, WAV, M4A, FLAC are available. An export format does not determine the language or generate speech.
- Wait for the upload to finish and the task to be queued. Processing then continues in the background. Return to My processed files to check the result, and save downloads while their temporary links remain available.
- Listen to every track you intend to use. Check whether original words remain in music or effects, and whether important effects were lost or changed. Keep an alternate source ready for passages that do not separate acceptably.
If your source is a song, sung vocals are different from spoken dialogue. A song's singing and accompaniment can be separated with the Music scene, but preparing a translated song also requires work beyond this tool.
Which files should you keep for the new mix?
| Purpose | File to keep |
|---|---|
Check the source script or send speech to another tool | Separated original voice audio |
| Rebuild selected non-dialogue sounds | The background track, after listening for speech leakage; music and effects are combined |
Work with the original visuals and exact edit context | Original source video |
Make an intermediate version with the source-language voice | Voice-only MP4; this is not the translated final video |
For Audio, the background result combines non-vocal sounds; it is not an independent music-and-effects pair. A background label does not guarantee a perfectly dialogue-free track. If unwanted source words remain, replace that section or use other material you are entitled to use.
How do you finish the translated or dubbed video?
Prepare and check the source script using your transcription tool or a manual transcript. Translate it in your chosen workflow and review names, meaning, and terminology. Record or generate the target-language speech in a separate tool, using voices and recordings you are authorized to use. The transcription and subtitle preparation guide explains how to verify the source words.
Bring the original video and new voice into your editor. Mute the original mixed soundtrack for a replacement dub, then align the new recording to the relevant scene or speaker. Adjust phrasing, cuts, and pauses where necessary; another language can require different timing. Add the selected music and effects, then listen for doubled speech, source-language remnants, and transitions that no longer fit the picture.
Review the complete export with someone able to check the target language, and inspect the video in its intended player. Audio separation does not verify translation quality, deliver lip sync, or guarantee a lossless copy of the video. Preserve the original and the editable project so corrections remain possible.
Open the music-removal tool to prepare the source tracks. Read the
voice-only video export guide
to understand the intermediate MP4, and the privacy policy before submitting sensitive recordings.