How to Mix Background Music Under Voiceovers and Spoken Audio
Mixing is addition. Two waveforms are summed sample by sample, each scaled by the level you chose, and the result is a single track containing both. The whole craft is in the levels, because a bed that is too loud makes speech tiring to follow and one that is too quiet is a waste of everyone's attention. The default here is music at 0.2, roughly a fifth of the voice, and it is a good starting point rather than a rule. Speech occupies the middle of the frequency range and so does most music, so they compete directly for the same space. The reliable test is not whether you can hear the words while concentrating, it is whether you can follow them while distracted, which is how anybody actually listens. If you find yourself paying attention to hear the speech, the music is too loud, and this is more pronounced on phone speakers and in cars than on the headphones you are mixing with. Looping handles the common case where a thirty-second music bed has to sit under a twenty-minute episode. The music simply repeats until the voice ends. Whether that is acceptable depends entirely on the bed: material written for the purpose loops without drawing attention, whereas a song with a recognisable structure announces every repeat. The output always takes the length of the voice, so music longer than the recording is cut off rather than extending past the end. Two things to sort out before mixing. The music you use has to be music you may use, and background beds are exactly what stock and Creative Commons libraries exist for. And the peaks are caught by a soft limiter as the two are summed, which prevents clipping but is not a substitute for choosing sensible levels: if the limiter is working constantly, the mix is too hot and the voice will sound squashed.
How to Use the Audio369 Online Add Background Music (Step-by-Step)
The main track goes in the workspace and determines the length of the finished mix.
A second file, decoded as it arrives. Use something you have the right to use; this is precisely what royalty-free libraries are for.
Music from 0 to 1 and voice from 0.2 to 1.5. Starting at 0.2 for the bed is sensible, and the right answer is usually quieter than you expect.
Leave looping on to fill a long recording with a short bed. The mix ends when the voice does, whatever the music was doing.
Technical Architecture & Audio Engine Specifications
Choosing a music level
| Music level | Effect | Suits |
|---|---|---|
| 0.05 to 0.1 | ✓ Barely there, felt more than heard | Long-form speech and audiobooks |
| 0.2 | ✓ Clearly present, speech still effortless | The default, and right for most podcasts |
| 0.3 to 0.4 | ✓ Prominent; speech needs attention | Short promos and adverts |
| 0.5 and above | ✓ Competes with the voice | Music-led pieces with occasional speech |
| 1.0 | ✓ Equal to the voice | Almost never what you want under speech |
Who Uses Audio369 Add Background Music? (Practical Creative Workflows)
A bed under a podcast introduction
Music under the opening thirty seconds sets a tone and marks the start of the show, then gets out of the way for the content.
Voiceover for a video
Narration over a quiet bed sounds finished, where dry narration on its own can feel like a placeholder.
Adverts and promos
Short commercial pieces carry more music than long-form, because they have seconds to make an impression rather than minutes.
Meditation and exercise audio
Continuous music under instruction is expected in this material, and the bed is doing real work rather than decorating.
Covering an inconsistent room
A quiet bed masks small changes in background noise between takes, which is a legitimate trick and cheaper than re-recording.
Frequently Asked Questions (FAQ)
Does the music automatically get quieter when someone speaks? +
No. That is ducking, driven by a sidechain, and this tool applies fixed levels for the whole file. For an intro where music plays alone and then drops under the voice, mix the opening separately and join the pieces.
What level should the music be? +
Start at 0.2 and judge it while distracted rather than while concentrating. If you have to attend to the speech to follow it, the bed is too loud, and it will be worse on a phone speaker.
What happens if the music is shorter than the voice? +
It loops, if looping is enabled, repeating until the recording ends. Purpose-made beds loop invisibly; songs with a clear structure make every repeat obvious.
What if the music is longer? +
It is cut off when the voice ends, because the output always takes the length of the main recording. Fade the end afterwards if the cut is abrupt.
Can I use any song as background music? +
Only music you have the right to use. Commercial recordings under a podcast are a licensing problem, not a technical one, and stock and Creative Commons libraries exist for exactly this.
How loud should background music be relative to spoken dialogue? +
A general audio mixing rule is to set background music volume between 15% and 25% of the voice level (roughly -18 dB to -24 dB relative to speech) so vocals remain clear and articulate.
How does auto-looping work if the music track is shorter than the voice? +
The engine repeats the backing soundtrack smoothly until the spoken audio finishes, applying subtle boundary crossfades so listeners do not detect abrupt musical restarts.
Will adding background music increase the risk of digital clipping? +
Summing two audio signals increases total amplitude. Our mixer calculates combined peak headroom and applies a transparent soft-knee limiter to prevent clipping during loud passages.