How to Mix Vocals and Background Music: Audio Ducking, EQ and Levels Explained

To keep a voice clear over background music, lower the music by 12 to 15 dB whenever someone speaks. This is called audio ducking, and a sidechain compressor does it automatically. A small cut in the music’s midrange and a final loudness check complete the mix.

Why Music Drowns Out a Voice

When a voice is hard to follow over music, turning the voice up rarely solves it. The problem is masking: two sounds in the same frequency range compete, and the ear follows the louder one. Speech is understood mainly through the range from about 1 to 4 kHz, where consonants sit, and that is exactly where guitars, pianos, synths and sung vocals are strongest.

Masking is a well-studied property of hearing; the University of Minnesota’s open textbook explains it in its chapter on critical bands and masking .

There are two ways to make room for the voice. You can lower the whole music track while someone is speaking (ducking), or you can lower only the part of the music that competes with speech (EQ). Most podcast and video mixes use the first, and add the second when needed.

What Is Audio Ducking?

Audio ducking means automatically lowering one sound whenever another one is present. On the radio, the music drops when the presenter starts talking and comes back up when they stop: that is ducking. The same effect is used in podcasts, YouTube videos, audiobooks with music, and announcements in shops and stations.

The tool behind it is a sidechain compressor. A normal compressor turns a track down when that track itself gets loud. A sidechain compressor listens to a different track: it watches the voice and turns down the music. Four controls shape the result:

• Threshold: how loud the voice has to be before the music is lowered. Set it so normal speech triggers the effect but breaths and quiet noises don’t.
• Reduction (depth): how far the music drops, in decibels.
• Attack: how quickly the music drops when speech starts.
• Release: how quickly the music returns when speech stops.

Sidechain compression and ducking are the same technique. In dance music it is used on purpose with fast, deep settings so the whole track “pumps” in time with the kick drum; for speech you want the opposite, a change smooth enough that nobody notices it.

The Best Audio Ducking Settings

These starting points work for most spoken content. Adjust them by ear: the music should step back without seeming to switch off.

Too short a release makes the music jump back between words, which listeners hear as pumping. Too long a release leaves the music quiet during long pauses. Half a second is a safe middle.

How to Mix Voice and Music Online, Step by Step

The Auto-Ducker takes a voice track and a music track and mixes them with automatic ducking. Both files stay in your browser.

Manual Volume Changes vs. Automatic Ducking

Before sidechain compressors, engineers “rode the faders”, pulling the music down by hand every time someone spoke. You can still do this in any editor by drawing volume changes, and it gives the most control: you can let the music swell for a dramatic pause or keep it low through a short breath. Automatic ducking is faster and more consistent, which makes it the better choice for long recordings such as a 40-minute episode. For a 30-second advert with three sentences, doing it by hand is often just as quick.

Use EQ to Make Room for the Voice

Sometimes the music should stay at a steady level, for example under a fast-paced video. In that case, lower only the range that competes with speech. Open the music track in the 3-Band EQ and reduce the Mid band by 2 to 4 dB. The music keeps its bass and sparkle, but leaves a gap where the voice sits.

On the voice itself, a Low Cut at 80 to 100 Hz removes rumble that would otherwise add up with the bass of the music. Avoid large boosts: cutting from the music is cleaner than boosting the voice. The guide to bass, mid and treble explains the ranges in more detail.

Common Mistakes When Mixing Voice and Music

• Music with lyrics. A second voice competes directly with yours. Instrumental tracks are far easier to mix under speech.
• Ducking too deep. If the music nearly disappears and then leaps back, the effect is distracting. Reduce the depth before you touch anything else.
• A release that is too short. The music pumps up between words. Lengthen the release until the level stays down through short pauses.
• Busy music. Tracks with a strong lead melody or lots of mid-range instruments mask speech even at low volume. Simple, steady beds work best.
• Checking only on headphones. Play the mix once on a phone speaker. If the voice is clear there, it will be clear everywhere.

Which Music Can You Use?

Use music you made yourself, music you have licensed, or royalty-free tracks whose license covers your use. Putting a commercial song under your voice and publishing it usually requires permission from the rights holder, and platforms may mute or remove videos that use music without it.