Audio Ducking

App: apps/audio_ducking_demo/

A continuous music drone that automatically ducks (drops in level) while a “voice” cue plays (engine PLM-148 / ADR 0047) — the classic music-under-dialogue effect. Hold Space to talk; the music dips, then swells back when you stop.

Run from root

% ./plume3d audio_ducking_demo

Hold Space: the second tone (the “voice”) sounds and the music drone ducks about 14 dB; release it and the music rises back over ~400 ms. A bar shows the music’s live duck level.

What it does

• Creates a Music drone and a Voice tone, each an ordinary streaming Source routed to a Mixer.group (Music / Voice).
• Configures the sidechain once: music.duckUnder(voice, -14, 0.05, 0.4) — duck 14 dB, 50 ms attack, 400 ms release.
• Feeds the Music tone every frame (always playing) but only feeds the Voice tone while Space is held. The engine sees a playing source on the Voice bus and ducks the Music bus automatically — no per-frame ducking code in the game.
• Draws the Music group’s duckGain as a Gui.progress bar.

Key pattern

import "engine" for Mixer

var music = Mixer.group("Music")
var voice = Mixer.group("Voice")

// Configure once — the engine applies it every frame.
music.duckUnder(voice, -14, 0.05, 0.4)   // amountDb, attack s, release s

// ...later, drive the trigger by playing sound on the Voice bus.

APIs

• Mixer — MixerGroup.duckUnder, clearDuck, duckGain.
• Audio — the streaming sources on each bus.
• Gui — the duck-level bar.