Set up ducking in OBS: lower music automatically when you speak

Music can sound comfortable during a pause yet drown out your words when you speak. Ducking lowers it automatically in response to your microphone. It is easier to configure after your voice and music are separated into different OBS sources.

Prepare a separate music source

If music, game audio, and other sounds all enter one Desktop Audio source, the compressor will lower that entire mix. First give music its own source. Check the mixer: its meter should move when music plays, and the microphone meter should move when you speak.

Add a compressor to the music

1. Open the filters for the music source and add Compressor.
2. Under Sidechain/Ducking Source, select the microphone that will trigger ducking.
3. Play your usual music and say a few sentences at your normal volume. Listen to the result in a test recording.

The compressor goes on the music, while the microphone only controls the reduction. Putting it on the microphone instead addresses a different task: changing the volume of the voice itself.

Adjust its response to speech

• Threshold sets the control-signal level at which compression begins.
• Ratio determines compression strength: a higher value limits the music more once triggered.
• Attack controls response speed, and Release controls how music returns after a pause.

Start with a moderate reduction and change one setting at a time. If music jumps noticeably between words, try a smoother return. If the beginning of a sentence is buried, check response speed and microphone level. Universal values cannot account for your voice, distance from the microphone, or the track's starting volume.

Test pauses and quiet sentences

Record calm speech, laughter, and a pause with keyboard sounds. The music should make room for your voice rather than disappear at every incidental noise. Repeat the test with your game after setting it up. A comfortable balance is audible without constantly watching the mixer.