Saturalabs / Specific distractions
Remove background music from your audio.
Work with an audio recording where music competes with spoken content. Separate the music bed or focus on the voice, depending on which parts of the recording you need to preserve.
Start with your recordingThe track you’re looking for
The remaining recording with the requested music separated. Check soft speech and pauses.
Name one audible source. You can edit this prompt before running the separation.
Uploading opens the editor. Downloads require an account and credits. View pricing.
A focused approach
Decide how much of the original recording to keep
A podcast excerpt, recorded presentation, or rendered voiceover may contain speech over a music bed. If other sounds matter too, request the background music and listen to Background audio for the remaining mix. If the goal is a voice-only edit, try requesting the narrator or foreground speaker and keep Isolated sound.
These choices are especially useful when you have one mixed audio file rather than separate voice and music channels. A music-targeted remainder can retain more context, while a voice-targeted output may sound more isolated. Compare the options based on the next edit you need, rather than choosing whichever has the least audible background.
From mix to source
How to remove background music from audio
Use the clearest source file
Upload MP3 or WAV up to 50 MiB (52.43 MB). MP4 and MOV are supported too. Retain the original export so you can judge speech and the surrounding mix at a comparable level.
Choose a source description
Try ‘The background music’ when you want the remaining scene or recording. Try ‘The narrator’s spoken voice’ when you only need narration. Keep the prompt to one target.
Select the matching output
A music prompt places the requested bed in Isolated sound and the remainder in Background audio. A voice prompt places the requested speech in Isolated sound. Listen to both.
Review and export the chosen audio
Check sentence endings, pauses, and loud musical sections before downloading the WAV. Use your audio editor for trims, fades, and final level adjustments.
Be specific about the sound
A prompt for the result you want
Describe the source, then choose the output that matches your goal. These are starting points to try with your own recording.
| Your goal | Try this prompt | Keep this track |
|---|---|---|
| Reduce a music bedKeep the remainder and check for both speech damage and music remnants. | The background music | Background audio |
| Keep narration aloneUse when the surrounding recording is not part of the intended edit. | The narrator’s spoken voice | Isolated sound |
| Keep a conversationRequests speech as a group, rather than separate tracks for each participant. | The people speaking | Isolated sound |
Made for a real next step
Put the separated audio to work
Rework a narration mix
Try recovering narration when only a rendered voiceover mix remains. Compare against the original so sentence endings and timing stay understandable.
Prepare a podcast excerpt
Choose a focused voice output or a music-reduced remainder depending on whether you need the room and conversation context in the excerpt.
Study spoken content
Attempt to make a presentation or explanation easier to follow when a music bed competes with the speaker. Check intelligibility rather than silence alone.
Listen before you commit
Three checks worth making
Compare a quiet passage and a busy passage with the original, at a similar listening level.
Quiet syllables
Listen to softer consonants and phrase endings. Reducing music is not an improvement if important words become incomplete.
Sound between sentences
Music remnants are often easiest to hear in pauses. Decide whether they will be distracting in the next edit.
Consistency across sections
A changing music bed can affect separation differently. Check both sparse and busy moments before choosing the full output.
Know the limits
A mixed file has limits that separate channels do not
Speech and music can share frequency ranges and effects. Separation may introduce tonal changes or retain parts of the bed. If you have the original voice recording, use it as your first comparison rather than assuming extraction is an equivalent replacement.
This process does not automatically transcribe speech, remove every musical source, or repair words obscured by severe distortion. Keep the original and judge the selected output in the context where it will be heard.
Can I remove music from an MP3 and keep the voice?
You can try targeting background music and keeping Background audio. For a voice-only goal, try targeting the speaker and keeping Isolated sound. Review quiet words and loud overlaps in either result.
What is the difference between music removal and voice isolation?
Music removal targets the music and uses the remainder. Voice isolation targets speech and uses the isolated output. Choose based on whether you need surrounding sounds as well as the voice.
Can I preserve multiple people speaking?
You can request the people speaking as a group, but overlapping voices and music can limit the result. This does not automatically produce individual speaker tracks.
Will the remaining audio be silent between words?
Not necessarily. Music or ambience can remain, and reducing those sounds may also affect speech. Evaluate whether the pauses and spoken content suit your intended use.
What file do I download?
The isolated and remaining tracks are downloadable WAV audio. Downloads require an account and sufficient credits; the results screen shows the cost before downloading.
Start with a sound you know
Bring your own recording.
Describe one source. Listen to both tracks. Keep what works for your project.