🎚️ Voice Effects¶
Voices can run through a Source Effect Chain: bitcrusher, filter, EQ, ring modulation, the pitch shifter EPVC ships with, whatever you assemble. That's how you get a radio crunch, a helmet filter, a possessed-monster voice, or a phone call.
It lives on the ProximityVoiceConfig asset, so it follows the same rules as everything else: author presets, assign, swap at runtime.
1. Build the chain¶
In the Content Browser:
- Right-click → Audio → Effects and create one or more Source Effect Presets (Bit Crusher, Filter, EQ, Ring Modulation, Waveshaper...).
- Tune each preset by double-clicking it.
- Right-click → Audio → Effects → Source Effect Preset Chain and add your presets to its
Chainarray, in the order you want them applied.
Warning
Voice is mono. Use effects that support 1 channel. Stereo-only effects (stereo delay, panner, stereo widening) will not behave as expected on a voice source.
2. Assign it¶
In your ProximityVoiceConfig asset, under Playback:
| Property | Description |
|---|---|
Source Effect Chain |
Source effects applied to the voice itself, per speaker and per emitter. |
Note
The chain is picked up on the next talk burst. Swapping it while someone is mid-sentence won't change that sentence, the next one will use it.
3. Clean on the body, crunchy on the radio¶
The chain is applied per emitter, not per speaker. Since a receiver can override the config with its own Output Configuration, the very same voice can come out clean from the player's mouth and processed from a radio, at the same time, on the same machine.
The usual setup:
| Asset | Used by | Chain |
|---|---|---|
DA_NormalVoice |
The speaker's Default Configuration |
None, or a light one |
DA_RadioReceiver |
The receiver's Output Configuration |
Bitcrusher + band-pass, the radio crunch |
See Radios & Receivers for the receiver side.
4. A radio crunch that works¶
A convincing radio is mostly bandwidth, not distortion:
- Filter (band-pass): cut everything below ~300 Hz and above ~3 kHz. That alone already sounds like a speaker grille.
- Bit Crusher: drop the sample rate / bit depth a little for digital grit. Go easy, too much and speech becomes unintelligible.
- Optionally a touch of Waveshaper for saturation.
Tip
Just need a muffled voice, no chain at all? Base Low Pass Frequency on the same config gives you an always-on low-pass for free. See Configuration.
5. Pitch shift the voice¶
Unreal ships no pitch shift source effect, so EPVC includes one: SourceEffectVoicePitchShift.
Right-click → Audio → Effects → SourceEffectVoicePitchShift, tune it, then add it to a chain like any other preset.
| Property | Description |
|---|---|
Pitch Shift Semitones |
Pitch offset, -24 to 24. Negative is deeper, positive is higher, 12 is one octave. |
Random Semitone Range |
Extra offset drawn once per source, somewhere inside ± this many semitones. One preset then gives every speaker a different voice. 0 = everyone shifted identically. |
Window Length Ms |
Window the shifter reads over, 10 to 100. Short warbles on held vowels, long smears consonants. 25 to 40 suits speech. |
Wet Level |
Gain of the shifted signal. |
Dry Level |
Gain of the untouched signal. Leave some in and the detuned copy doubles the voice. |
Tip
Random Semitone Range around 2 is the cheapest way to stop a crowd sounding cloned. Each speaker draws their own offset when their voice starts, so a single preset covers your whole server.
Note
Why not just a pitch multiplier on the audio component? Because voice is a live stream. A multiplier changes how fast the source is consumed, so playback drifts against the jitter buffer and latency creeps up over a long sentence. This effect moves the pitch and leaves the clock alone.
Retune it from Blueprint, with the preset asset as the target:
Set Pitch Shift Semitones (Target = SEP_VoiceDeep, In Semitones = -7.0)
Warning
A preset is a shared asset. Changing it at runtime affects every source currently running it, not one player. To pitch a single player, give them their own config with its own chain and swap it with Set Voice Configuration.
6. Swap effects at runtime¶
The chain is part of the config asset, so Set Voice Configuration swaps it like everything else: put on a gas mask, switch to DA_MaskedVoice, take it off, switch back.
Set Voice Configuration (New Configuration = DA_MaskedVoice)
Warning
The active config is replicated. Set it once on the server (authority) so everyone hears the same thing. See Configuration.