How loud should background music be under your voice?

Background music should sit roughly 15 to 20 decibels below your voice, so it's felt more than heard. In practice: set your voice first so it's clear and full, bring the music in until you can just hear it under the speech, then turn it down a little more. Use your editor's ducking feature so the music dips while you talk and comes back up in the pauses.

Start with the voice

Get the speech right before adding anything:

• Level: speech should be the loudest thing in the mix, peaking a few decibels below the maximum.
• Consistency: a compressor evens out loud and soft words so the music doesn't swallow the quiet ones.
• Clean-up: remove hum and room noise first. Music doesn't hide noise, it adds to it.

Platforms like YouTube and Spotify turn loud uploads down to a standard loudness, so making everything louder doesn't help. Making the voice clearer does.

Then set the music

• 15 to 20 dB below the voice while you're speaking is a good starting point. For calm, spoken content, go quieter.
• Back up to near full level in intros, outros and B-roll without speech.
• Listen on phone speakers and cheap earbuds, not just good headphones. That's where most viewers are, and it's where music drowns speech first.

Duck it automatically

"Ducking" lowers the music whenever there's speech. Most editors can do it for you:

• Premiere Pro: in the Essential Sound panel, tag the voice as Dialogue and the track as Music, then turn on Ducking.
• DaVinci Resolve: use the ducking or sidechain controls in the Fairlight page, with the voice as the trigger.
• iMovie: select the voice clip and use "Lower volume of other clips".
• Other editors usually have an "auto ducking" or "auto volume" option in the audio settings.

Set the dip to about 10 to 15 dB, with a short fade in and out, so the music moves smoothly instead of pumping.

The EQ trick

Speech is easiest to understand in the middle frequencies, roughly 1 to 4 kHz. If your editor has an equaliser, cut the music a few decibels in that range. The voice gets clearer without the music getting quieter.

Pick music that leaves room

The track matters as much as the level:

• No lyrics. Sung words fight spoken words directly.
• Few instruments in the voice range. Sparse piano, bass and soft drums leave space; busy melodies don't.
• Steady dynamics. A sudden loud part will need its own fix.

Lofi is popular under speech for exactly these reasons: soft, steady, instrumental. The Study Session room and lofi jazz are good starting points, and every track is free to use with a credit line.

Quick checklist

1. Voice clear and even.
2. Music about 15 to 20 dB under the voice.
3. Ducking on, around 10 to 15 dB, with smooth fades.
4. A small cut in the music around 1 to 4 kHz.
5. Checked on a phone speaker.
6. Credit line in the description. Formats here.

For Shorts and Reels, where the whole thing is 30 seconds, see picking the right 30 seconds.

Written by Suman Kumar with AI help and checked before publishing. Numbers about our own music come from our Spotify for Artists and DistroKid exports; other sources are linked in the post.