Mixing & Buses

Every voice routes through a bus, a mix group with its own volume and mute. Buses are how you give players a Music slider that doesn’t touch gunfire, or duck the world under dialogue.

The buses

Bus	For
Master	Everything. Feeds the output device.
Music	Score and licensed tracks.
SFX	World and gameplay sound. The default for new sources.
UI	Menu clicks, notifications, anything non-diegetic.
Voice	Dialogue and VO.
Ambient	Beds, room tone, weather.

The five named buses all feed Master, so Master’s volume scales the whole mix. Pick a bus per source with its Bus property, per play call with the play params’ Bus field, or move a live voice with SetBus.

Volume multipliers

A bus volume is a plain multiplier, 1.0 is unchanged and 0.0 is silent. The starting values ship in the audio settings, and anything can change them at runtime.

GAudioContext->SetBusVolume(EAudioBus::Music, 0.4f);
GAudioContext->SetBusMuted(EAudioBus::Voice, true);

Sound.SetBusVolume(EAudioBus.Music, 0.4f);
Sound.SetBusMuted(EAudioBus.Voice, true);

Volume changes are ramped rather than applied instantly, so dragging a slider doesn’t click.

This is the whole pattern. Read the current values to populate your sliders, write them back as the player drags, and persist once when they hit Apply.

The difference between the two languages is only in the last step. C# has a single SaveMixSettings() call; in C++ you copy the live volumes into CAudioSettings and ask the config system to write it, which is exactly what that C# call does internally.

// Populate.
const float Music = GAudioContext->GetBusVolume(EAudioBus::Music);
const float Sfx   = GAudioContext->GetBusVolume(EAudioBus::SFX);

// Live preview while dragging.
GAudioContext->SetBusVolume(EAudioBus::Music, MusicSlider);
GAudioContext->SetBusVolume(EAudioBus::SFX, SfxSlider);

// Apply.
CAudioSettings* Settings = GetMutableDefault<CAudioSettings>();
for (uint32 Index = 0; Index < NumAudioBuses; ++Index)
{
    const EAudioBus Bus = (EAudioBus)Index;
    Settings->SetBusVolume(Bus, GAudioContext->GetBusVolume(Bus));
}

GConfig->SaveSettings(CAudioSettings::StaticClass());

// Populate.
float Music = Sound.GetBusVolume(EAudioBus.Music);
float Sfx   = Sound.GetBusVolume(EAudioBus.SFX);

// Live preview while dragging.
Sound.SetBusVolume(EAudioBus.Music, MusicSlider);
Sound.SetBusVolume(EAudioBus.SFX, SfxSlider);

// Apply.
Sound.SaveMixSettings();

Either way the values land in the project’s AudioSettings.json, so they come back on the next launch.

Ducking

There’s no automatic sidechain, but a duck is a few lines: drop the bus you want out of the way, restore it when the important sound finishes.

FAudioPlayParams Params;
Params.Bus = EAudioBus::Voice;

FAudioHandle Line = GAudioContext->PlayAudio(Dialogue->GetAudioData(), Params);
GAudioContext->SetBusVolume(EAudioBus::Music, 0.25f);

// When GAudioContext->IsPlaying(Line) goes false, put it back.
GAudioContext->SetBusVolume(EAudioBus::Music, 1.0f);

PlayingSound Line = Sound.PlayOnBus(Dialogue, EAudioBus.Voice);
Sound.SetBusVolume(EAudioBus.Music, 0.25f);

// When Line.IsPlaying goes false, put it back.
Sound.SetBusVolume(EAudioBus.Music, 1.0f);

Reverb

The engine has one shared reverb return. Buses send into it, so you can wet the world without touching UI or music.

C# takes the four reverb parameters as arguments; C++ fills in an FAudioReverbParams struct with the same four fields.

GAudioContext->SetBusReverbSend(EAudioBus::SFX, 0.35f);
GAudioContext->SetBusReverbSend(EAudioBus::Ambient, 0.5f);

FAudioReverbParams Reverb;
Reverb.RoomSize = 0.8f;
Reverb.Damping  = 0.4f;
Reverb.Width    = 1.0f;
Reverb.WetLevel = 0.5f;

GAudioContext->SetReverbParams(Reverb);

World.Audio.SetBusReverbSend(EAudioBus.SFX, 0.35f);
World.Audio.SetBusReverbSend(EAudioBus.Ambient, 0.5f);

// Room size, damping, stereo width, wet level.
World.Audio.SetReverb(0.8f, 0.4f, 1.0f, 0.5f);

Parameter	Does
Room Size	Decay length. 0 is a small room, 1 is a cathedral.
Damping	High-frequency absorption in the tail. Higher is darker.
Width	Stereo spread of the tail.
Wet Level	Gain of the reverb return.

Sends default to 0 and the reverb network isn’t built until something asks for it, so a project that never uses reverb pays nothing for it.

Because the parameters are live, reverb zones are just a trigger volume that sets them on enter. Starting values for the sends and the reverb shape live in the audio settings.