Audio, explained · 23 / 36

Fun Voice Changer

Turn your voice into a character with presets you can tweak.

Visual walkthrough36 configuration controlsSynced audio walkthrough

The tool always delays the voice by 21 ms, whichever preset is active, and outputs the same voice on both channels. A safety limiter keeps peaks below −1 dBFS. It is a fun effect, not voice anonymization: people may still recognize you.

01 / Understand

The idea, before the settings.

Pick a preset such as Pitch Correct, Deep, Robot or Radio, then adjust any of its settings. Pitch correction pulls your voice to the notes of a scale; pitch and formant shifting change how high and how big the voice sounds; ring modulation, bit crushing, saturation, whisper, tone, modulation and space add character.

Microphone chainFun Voice ChangerCable to Discord

Use it in AudioRouter

Add Fun Voice Changer after your microphone tools and before the output Discord uses as its microphone.
Choose a preset, press Play and talk; the inspector shows the note it hears while pitch stages run.
Adjust the preset's sliders, or open the sections below for every setting. Reset to preset restores the recipe.

What success looks like

Your voice comes out as the chosen character, at about the same loudness as your normal voice, 21 ms later.

If it does not work

If the effect sounds broken in Discord, turn Discord's noise suppression off for the cable input. If pitch correction warbles, lower the noise gate or add Denoise before the tool.

Select the node to open Properties. Click its name to rename it. While playing, supported parameter-only edits save after a short pause. For structural edits, use Save when the canvas marks a change. Enabled and Bypass are distinct live controls; see the in-app help for their behavior.

02 / Explore

Make the behavior visible.

A synthetic sung voice is regenerated with the chosen pitch and formant, then given the ring buzz. It teaches what each control changes, not the engine's exact sound.

Pitch-5 stFormant-3 stRobot buzz0 %

The example repeats until you press Stop, and control changes are heard immediately. The picture shows the tool's shape rather than time; follow the timing in Follow the sound below. This is not a DSP measurement or control for your running AudioRouter session.

03 / Configure

Every control, with a reason.

Preset

preset

The starting voice. Choosing one writes all of its settings at once.

Range / choicespitchCorrectSubtle · pitchCorrectHard · deep · chipmunk · monster · child · robot · alien · radio · megaphone · telephone · whisper · cave · customDefaultcustom

↓ Lower / alternativeChoose Custom to start with every effect off.

↑ Higher / alternativePick a character such as Robot or Deep.

Try thisStart from Pitch Correct (hard) for the tuned-vocal sound.

Keep in mind After you change a setting the preset shows as edited; Reset to preset restores it.

Noise gate

gateDb

Mutes sound quieter than this before the effects.

Range / choices-90 dBFS → -20 dBFSDefault-90 dBFS

↓ Lower / alternativeLets quieter sounds through; −90 dB is off.

↑ Higher / alternativeIgnores more room and keyboard noise.

Try thisTry −55 dB with a headset microphone.

Keep in mind Too high cuts the start and end of words.

Pitch correction

correctEnabled

Pulls the voice to the nearest note of the scale.

Range / choicesOn / offDefaultfalse

↓ Lower / alternativeOff: the voice keeps its own pitch.

↑ Higher / alternativeOn: notes snap toward the scale.

Try thisTurn it on and sing a scale slowly.

Keep in mind Needs a clear, voiced sound; noise and whispers are left alone.

Key

key

The key of the correction scale.

Range / choicesc · cs · d · ds · e · f · fs · g · gs · a · as · bDefaultc

↓ Lower / alternativeChromatic ignores the key.

↑ Higher / alternativeChoose the key of the song you sing along to.

Try thisC major fits many simple melodies.

Keep in mind With Monotone the key is the one note held.

Scale

scale

Which notes correction may use.

Range / choiceschromatic · major · minor · majorPentatonic · minorPentatonic · blues · monotoneDefaultchromatic

↓ Lower / alternativeChromatic allows every note.

↑ Higher / alternativeNarrower scales jump between fewer notes.

Try thisMonotone with retune 0 ms gives a flat robot voice.

Keep in mind A scale that does not match the song pulls notes to wrong pitches.

Retune speed

retuneMs

How quickly a note is pulled into tune.

Range / choices0 ms → 400 msDefault50 ms

↓ Lower / alternativeFaster correction; 0 ms snaps instantly.

↑ Higher / alternativeSlower, more natural correction.

Try this0 ms for the hard tuned-vocal effect; 80 ms for subtle help.

Keep in mind Very slow retune barely changes short notes.

Humanize

humanizePercent

Loosens correction on held notes.

Range / choices0 % → 100 %Default0 %

↓ Lower / alternativeHeld notes stay locked to the note.

↑ Higher / alternativeLong notes keep more natural vibrato.

Try thisTry 40 % for subtle correction.

Keep in mind Has no effect while retune speed is 0 ms.

Pitch and formant

shiftEnabled

Turns the pitch and formant shift on.

Range / choicesOn / offDefaultfalse

↓ Lower / alternativeOff: the voice keeps its pitch and size.

↑ Higher / alternativeOn: Pitch and Formant apply.

Try thisTurn it on for Deep, Chipmunk or Monster.

Keep in mind Large shifts can sound buzzy; that is part of the fun.

Pitch

pitchSemitones

Moves the voice up or down in semitones.

Range / choices-12 st → 12 stDefault0 st

↓ Lower / alternativeLower voice; −12 is one octave down.

↑ Higher / alternativeHigher voice; 12 is one octave up.

Try this−5 for a deep voice, +7 for a chipmunk.

Keep in mind The words keep their timing; only the note changes.

Fine pitch

pitchCents

Small pitch steps added to Pitch.

Range / choices-100 cents → 100 centsDefault0 cents

↓ Lower / alternativeSlightly lower.

↑ Higher / alternativeSlightly higher.

Try this100 cents is one semitone.

Keep in mind Useful for detuned, wobbly characters.

Formant

formantSemitones

Changes the size of the voice without changing its note.

Range / choices-12 st → 12 stDefault0 st

↓ Lower / alternativeBigger, darker voice.

↑ Higher / alternativeSmaller, younger-sounding voice.

Try thisCombine −3 formant with −5 pitch for Deep.

Keep in mind Moving formants far from the pitch shift sounds unnatural.

Ring modulator

ringEnabled

Multiplies the voice by a tone.

Range / choicesOn / offDefaultfalse

↓ Lower / alternativeOff: no ring modulation.

↑ Higher / alternativeOn: metallic, robotic or alien sound.

Try thisRobot uses 50 Hz at 70 % depth.

Keep in mind High frequencies make words harder to understand.

Ring frequency

ringHz

Frequency of the modulating tone.

Range / choices20 Hz → 2000 HzDefault50 Hz

↓ Lower / alternativeLow buzz.

↑ Higher / alternativeBell-like, alien sound.

Try this30–80 Hz for robots, 300 Hz for aliens.

Keep in mind Around 100–200 Hz the voice can sound out of tune.

Ring depth

ringDepthPercent

How much ring-modulated sound replaces the voice.

Range / choices0 % → 100 %Default100 %

↓ Lower / alternativeMore of the normal voice.

↑ Higher / alternativeStronger effect.

Try thisStart at 50 %.

Keep in mind 100 % replaces the voice entirely.

Bit crusher

crushEnabled

Lowers audio quality on purpose.

Range / choicesOn / offDefaultfalse

↓ Lower / alternativeOff: clean sound.

↑ Higher / alternativeOn: grainy, digital sound.

Try thisTelephone uses 12 bits and downsample 6.

Keep in mind Strong settings add harsh noise.

Bits

crushBits

Resolution of the crushed sound.

Range / choices4 → 16Default16

↓ Lower / alternativeGrainier distortion.

↑ Higher / alternativeCleaner; 16 is clean.

Try thisTry 8 bits for an old game console.

Keep in mind Below 6 bits words become hard to follow.

Downsample

crushDownsample

Holds each sample for longer.

Range / choices1 → 12Default1

↓ Lower / alternativeCloser to the original; 1 is off.

↑ Higher / alternativeLo-fi, muffled and aliased.

Try this6 gives about an 8 kHz phone sound.

Keep in mind Adds metallic aliasing tones.

Saturation

driveEnabled

Distorts the voice.

Range / choicesOn / offDefaultfalse

↓ Lower / alternativeOff: no distortion.

↑ Higher / alternativeOn: grit, radio or megaphone sound.

Try thisMegaphone uses 70 % drive.

Keep in mind Distortion adds loudness; check the output level.

Drive

drivePercent

How hard the voice is pushed into distortion.

Range / choices0 % → 100 %Default0 %

↓ Lower / alternativeCleaner.

↑ Higher / alternativeMore distortion.

Try this30 % for a radio, 70 % for a megaphone.

Keep in mind Output is levelled automatically, but very high drive still sounds louder.

Whisper

whisperPercent

Replaces the voiced tone with breath.

Range / choices0 % → 100 %Default0 %

↓ Lower / alternativeMore normal voice; 0 % is off.

↑ Higher / alternativeMore breathy whisper.

Try this100 % for a full whisper.

Keep in mind Discord's noise suppression may remove whispered sound.

Tone

toneEnabled

Turns the tone filters on.

Range / choicesOn / offDefaultfalse

↓ Lower / alternativeOff: full range.

↑ Higher / alternativeOn: low cut, high cut and peak apply.

Try thisRadio uses 300–3,000 Hz.

Keep in mind Narrow ranges sound thin; that is the point for radio and phone.

Low cut

lowCutHz

Removes bass below this frequency.

Range / choices20 Hz → 2000 HzDefault20 Hz

↓ Lower / alternativeMore bass.

↑ Higher / alternativeThinner sound.

Try this300 Hz for radio.

Keep in mind Has no effect while Tone is off.

High cut

highCutHz

Removes treble above this frequency.

Range / choices1000 Hz → 20000 HzDefault20000 Hz

↓ Lower / alternativeDuller, more muffled sound.

↑ Higher / alternativeBrighter sound.

Try this3,000 Hz for radio.

Keep in mind Has no effect while Tone is off.

Peak frequency

peakHz

Where the tone peak acts.

Range / choices200 Hz → 8000 HzDefault1000 Hz

↓ Lower / alternativeWarmer, boxier character.

↑ Higher / alternativeBrighter, thinner character.

Try this2 kHz for a honky megaphone.

Keep in mind Has no effect while Peak is 0 dB.

Peak

peakDb

Boost or cut at the peak frequency.

Range / choices-12 dB → 12 dBDefault0 dB

↓ Lower / alternativeCuts that frequency.

↑ Higher / alternativeStronger honk at that frequency.

Try this+6 dB at 2 kHz for a megaphone.

Keep in mind Large boosts can sound harsh.

Modulation

modEnabled

Turns chorus or flanger on.

Range / choicesOn / offDefaultfalse

↓ Lower / alternativeOff: no modulation.

↑ Higher / alternativeOn: doubled or sweeping sound.

Try thisAlien uses chorus; Robot uses flanger.

Keep in mind Modulation smears speech slightly.

Modulation type

modType

Chorus doubles the voice; flanger sweeps it.

Range / choiceschorus · flangerDefaultchorus

↓ Lower / alternativeChorus for a thick, doubled voice.

↑ Higher / alternativeFlanger for a jet-like sweep.

Try thisFlanger with feedback for robots.

Keep in mind Switching type fades the effect out and in.

Rate

modRateHz

How fast the modulation sweeps.

Range / choices0.05 Hz → 5 HzDefault0.5 Hz

↓ Lower / alternativeSlower sweep.

↑ Higher / alternativeFaster wobble.

Try this0.2 Hz for a slow robot sweep.

Keep in mind Fast rates sound seasick.

Depth

modDepthPercent

How wide the modulation sweeps.

Range / choices0 % → 100 %Default50 %

↓ Lower / alternativeSubtler effect.

↑ Higher / alternativeStronger effect.

Try this60 % for an alien shimmer.

Keep in mind Has no effect while Modulation is off.

Feedback

modFeedbackPercent

Flanger only: feeds the sweep back.

Range / choices-90 % → 90 %Default0 %

↓ Lower / alternativeNegative values invert the ring.

↑ Higher / alternativeRinging, metallic sweeps.

Try this50 % for a robot.

Keep in mind Ignored by chorus.

Space

spaceEnabled

Adds a room or cave echo.

Range / choicesOn / offDefaultfalse

↓ Lower / alternativeOff: dry voice.

↑ Higher / alternativeOn: echo around the voice.

Try thisCave uses a large, long space.

Keep in mind Echo makes speech harder to follow in fast talk.

Room size

spaceSizePercent

Size of the simulated room.

Range / choices0 % → 100 %Default50 %

↓ Lower / alternativeSmaller, closer room.

↑ Higher / alternativeBigger, more distant space.

Try this90 % for a cave.

Keep in mind Has no effect while Space is off.

Echo length

spaceDecayMs

How long the echo takes to fade.

Range / choices200 ms → 4000 msDefault1000 ms

↓ Lower / alternativeShorter echo.

↑ Higher / alternativeLonger echo.

Try this2,500 ms for a cave.

Keep in mind Long echoes build up during continuous speech.

Echo level

spaceLevelPercent

How loud the echo is.

Range / choices0 % → 100 %Default30 %

↓ Lower / alternativeLess echo.

↑ Higher / alternativeMore echo.

Try this50 % for a cave.

Keep in mind Has no effect while Space is off.

Mix

mixPercent

Blend between the changed and the normal voice.

Range / choices0 % → 100 %Default100 %

↓ Lower / alternativeMore of your normal voice; 0 % is unchanged.

↑ Higher / alternativeOnly the changed voice at 100 %.

Try thisUse 0 % to compare without the timing jump of Bypass.

Keep in mind Mixing a shifted voice with the original sounds like two voices.

Output

outputDb

Level after the effects.

Range / choices-24 dB → 12 dBDefault0 dB

↓ Lower / alternativeQuieter.

↑ Higher / alternativeLouder.

Try thisPresets set this so each voice matches your normal level.

Keep in mind A safety limiter keeps peaks below −1 dBFS.

04 / Put it to work

Start with a useful example.

Example 1

Sound like a tuned vocal

Choose Pitch Correct (hard).
Set Key and Scale to the song, or leave Chromatic.
Sing slowly; raise Retune speed for a more natural sound.

Example 2

Make a robot

Choose Robot.
Lower Ring frequency for a deeper buzz.
Reduce Bits for a rougher, older machine.

Example 3

Use it in Discord

Put Fun Voice Changer between your microphone chain and the cable Discord uses as its microphone.
In Discord, turn noise suppression off or to Standard for that input.
Ask a friend whether the level matches your normal voice; adjust Output.

These are starting points. Compare at a comfortable monitoring level and adjust for your own source.

05 / Get unstuck

Small details that matter.

Discord's noise suppression can remove whisper and robot buzz; turn it off for the cable input.
Monitoring your own changed voice while talking is distracting because it arrives late.
Bypass skips the 21 ms delay, so the timing jumps; compare with Mix at 0 % instead.