BYTETOOLS

Add Background Music to Audio

Mix background music under a voiceover, podcast or narration, with automatic ducking so the speech stays clear. Both files stay inside your browser.

1. Main track (voice, podcast, narration)This one sets the length of the finished mix.
2. Background musicSits underneath at the volume you choose.

Both files are read straight from your device and mixed by a WebAssembly build of ffmpeg inside this tab. Nothing is uploaded. Mixing a long episode takes a while in the browser, so keep the tab open until the download appears.

What is the Add Background Music to Audio?

The ByteTools Background Music Mixer lays a music bed underneath a spoken track. Drop in your main recording and the music, set how loud the music should sit, and the two are mixed into a single MP3.

  • Mixes a music bed under any spoken track, with the voice setting the final length
  • Automatic ducking via sidechain compression so speech always stays on top
  • Adjustable music volume and fade-in so the bed enters gently
  • Loops music that is shorter than the main track, without ever overrunning
  • Both inputs resampled to a common format so mismatched files still mix
  • Runs entirely in your browser - neither file is uploaded

How to use the Add Background Music to Audio

  1. 1

    Drop your main track - the voiceover, podcast or narration - onto the first box.

  2. 2

    Drop the background music onto the second box.

  3. 3

    Set the music volume, usually 20 to 30 percent, and a fade-in length.

  4. 4

    Leave ducking on so the music dips under speech, and looping on if the music is short.

  5. 5

    Click Add Background Music, listen back, and download the mixed MP3.

About the Add Background Music to Audio

The ByteTools Background Music Mixer lays a music bed underneath a spoken track. Drop in your main recording and the music, set how loud the music should sit, and the two are mixed into a single MP3. The main track sets the length, so the music is trimmed or looped to fit rather than running past the end.

Automatic ducking is the feature that makes it sound professional. With it on, a sidechain compressor pulls the music down the moment someone speaks and lets it rise again in the gaps, which is exactly what a radio producer does by hand. A fade-in eases the music up at the start instead of dropping it in at full volume.

Both files are read straight from your device and mixed by a WebAssembly build of ffmpeg inside the tab. Nothing is uploaded, so unreleased episodes and client work stay private. Mixing a long episode takes a while in the browser, so keep the tab open until the download appears.

Frequently asked questions

How loud should background music be under a voiceover?

Usually 20 to 30 percent of the voice level, which is roughly 12 to 18 dB down. The music should be noticeable when you listen for it and invisible when you are following the words. Turning ducking on lets you sit it a little louder safely.

What does the ducking option actually do?

It runs a sidechain compressor with your voice as the trigger, so the music level drops automatically whenever someone is speaking and comes back up in the pauses. It is the same technique radio stations use to keep a presenter clear over a music bed.

What happens if the music is shorter than my recording?

With looping on, the music repeats until the main track ends. With it off, the music simply stops and the rest of your recording plays dry. Either way the mix always ends when the main track does - it never runs on past it.

Can I use any song as background music?

Only if you have the rights to it. Commercial music in a podcast or video will usually trigger a copyright claim regardless of how quiet you mix it. Royalty-free libraries and Creative Commons tracks exist precisely for this, and many are free.

Why is my mixed file taking so long?

The audio engine runs as WebAssembly in your browser rather than as native code, and it has to decode both files, mix them and encode the result. A short clip is quick, but a full hour-long episode can take several minutes. Leave the tab open and it will finish.

Related tools