PsyVOX
A synthesizer that sings with your voice.
PsyVOX is a voco-synth. A vocoder borrows your voice to shape another synth; PsyVOX makes a new voice from yours, using the synthesis that gives PsyOp its talking, singing tone. Speak or sing into a microphone, play the keys, and it answers in a synthetic voice that keeps every word, from the lowest bass note to the top of the keyboard. No microphone to hand? It sings along to sequences of its own, in time with your song.
Any sound can be its voice.
PsyVOX is a synthesizer you play from a keyboard, like any other. What makes it a voco-synth is its second input: an audio feed. Whatever you send it becomes the voice the synthesizer speaks with, while your keys decide the notes.
Sing it once, play it as chords
A sung or spoken line becomes a choir, a robot or a talking lead, following whatever you play.
Play it live
Speak into a mic while you play and PsyVOX answers in real time, on stage or at the desk.
Rhythm with a tune
Feed it a drum loop and every hit shapes the synth: percussive, pitched chords that groove.
Pads, guitars, field recordings
If a sound changes over time, PsyVOX can sing it. No input at all? Switch to Seq and it sings its own.
Hear what it does.
Three recordings from the beta, each with a different voice going in. Press play and they run in order, then round again. More are on the way.
- --:--
- --:--
- --:--
More examples are on the way.
Your voice in. A new voice out.
A classic vocoder lays the shape of your voice over a synth’s sound. PsyVOX starts from the voice instead: it takes the formant synthesis behind PsyOp and shapes a new voice from yours, so what you play speaks as clearly as what you sang.
- 01
A sound comes in
Your voice or any audio on the plug-in’s input, or one of PsyVOX’s own spectral sequences.
- 02
The analyzer splits it into bands
And measures the energy in every band, every few milliseconds.
- 03
The voice becomes a sound
The chosen voice turns what the analyzer hears into the synthesizer’s tone, and keeps up as your vowels change.
- 04
You play it
Chords, lines and bass notes all carry the words, clear at every pitch from the bottom of the keyboard to the top.
- 05
The rack finishes it
Filter, then the same eleven-module effects rack as PsyOp.
Two choices shape every patch.
Two selectors under the band display. Change the analyzer and your voice choice stays put: every analyzer ships with every voice.
How your sound is divided
How many bands, where they sit, how wide they are and how much they overlap. Three families: Modern perceptual scales such as Bark, ERB and Mel; Classic band plans in the manner of vintage hardware, from 11 to 22 bands; and Character plans like Robot, Telephone and a five-band Formant.
Controls the diction
What each band sounds like
- PreciseSmooth and clear
- BandEach band as a filtered slice
- HarmonicEvenly spread harmonics
- StepsHard-edged, square and saw
- ChirpA falling zap in every band
- BurstShaped noise, breathy
- BuzzReedy and vocal
- ResonateRinging, bell-like
Controls the timbre
A voice of its own, in time with your song.
Switch the Vocoder section from Live to Seq and PsyVOX plays a spectral sequence instead of listening to its input. A sequence stores the spectrum, not the bands, so every one plays through every analyzer, exactly as if it were coming in live.
Trim, loop, ping-pong
A spectrogram of the sequence with trim and loop markers you drag while it plays. Save just the clip you want.
Hold a vowel
Hold the current spectrum and keep playing on it after you stop singing. Snap it to save it as a one-frame sequence.
Build a vowel pattern
Line up snapshots on a step strip to make a stepped sequence, with rests where you want silence.
Four knobs for the carrier.
Every setting is matched to the same loudness, so turning a knob changes the tone and leaves the level alone.
±18 semitones
Moves every formant up or down together, for a smaller or larger voice.
Split it in two
Splits the voice into two layers up to 19 semitones apart. A little thickens, a lot turns metallic.
Wave · Voice · Pulse
From the textbook waveform, through the carrier that best carries formants, to a bright, flat pulse train.
Twin · Saw · Square
From a two-pulse comb through a plain saw to square and narrow pulse, with the pulse width built in.
Made to be played.
The vowel is shared across the chord, as with a classic vocoder. Pitch, vibrato and the amp envelope belong to each note.
Per-note bend and pressure
Slide one note without moving the others. Pressure swells a note’s sustain and can bring in vibrato.
Change the size
Move the formants up to twelve bands either way, in fractions of a band.
Up to eight slots
Unison or a slow-moving ensemble, always stereo. Start the slots in phase for a sweep on every attack.
Ready for a microphone
An input meter, gain and a low cut, then Emphasis to bring out consonants and a gate for room noise.
The beta, in numbers.
PsyVOX is in public beta. Details may change before release.
| Type | Voco-synth: a synthesizer voiced by your microphone or its own sequences |
|---|---|
| Beta build | 0.1.0 beta, VST3, Windows 10 or later, 64-bit |
| Analyzers | 18 band plans: Modern, Classic and Character |
| Voices | 8, built for every analyzer |
| Sequences | 26 factory, plus your own from recordings and imported audio |
| Tone | Formant ±18 st, split to 19.2 st, Tone and Color |
| Expression | MPE per-note bend and pressure, aftertouch, mod wheel |
| Ensemble | Unison or ensemble, up to 8 slots, stereo |
| Effects | The PsyOp rack: 11 modules, reorderable |
| Plug-in codes | Manufacturer mPsy · plug-in pv01 |
| License | The beta runs as a 60-day full trial. Offline license file at release |
Try it now.
Windows installer, VST3
The public beta is the whole voco-synth, with its factory voices and sequences, and runs in full for sixty days. Requires Windows 10 or later, 64-bit and a VST3 host. Mac versions are on the way: get one email when they are ready.
- File
PsyVOX-0.1.0-beta-20261008-1641.exe- SHA-256
21de8eb7dc7aca1abc385bd7cee18bdce9f613b80f4e0a6b9ec8fd09baf6aa80- Signed by
- MadPsy key
f226 9962 5b36 34e4
Pre-release installers are not yet code-signed with a Windows certificate, so SmartScreen may say it “prevented an unrecognized app from starting”. Choose More info, then Run anyway. Release builds will be signed.
In PowerShell, Get-FileHash .\PsyVOX-0.1.0-beta-20261008-1641.exe should print the SHA-256 above. The signature file carries the same hash, signed with the MadPsy key that signs every plug-in’s content.
Bugs, crashes and ideas all help: say which host you use and what you were doing. Contact us, or join the feedback list and we will ask now and then.
Bundle owners get PsyVOX at no charge when it is released.
Help us finish it.
The beta is the whole voco-synth, and it runs in full for sixty days. Each new beta version starts a fresh sixty days, so you can keep testing as it changes; the released PsyVOX has the standard thirty-day trial. All we ask in return is that you tell us what you find.
Tell us what you find.
Crashes, sounds that seem off, controls that confuse you, things you wish it did. Tell us the beta version, your system and your DAW, and a person reads every word. Happy to be asked now and then? Join the feedback list.
Send feedbackMade something with it?
Send us a demo track, presets or samples made with the beta. If we approve them, the plug-ins you used are yours, with your name on the work wherever it appears.
Submit to the ResidencyTwo things to connect.
PsyVOX loads as an instrument. It needs notes, like any synth, and an audio feed into its input, called Modulator in hosts that show input names. Most DAWs list that input as a sidechain. Mono or stereo both work.
Ableton Live11.1 and later
- Drop PsyVOX on a MIDI track. Arm it, or set Monitor to In, so your keyboard plays it.
- Put the sound you want on an audio track: a vocal clip, a drum loop, or a microphone input with Monitor set to In.
- On the PsyVOX device, open the Sidechain section, switch it on, and set Audio From to that audio track. Choose the Pre FX tap.
- To hear only PsyVOX, turn the source track’s volume all the way down. The Pre FX tap still feeds PsyVOX.
Reaper6 and later
- Insert PsyVOX on a new track (Insert → Virtual instrument on new track). Arm it with your MIDI keyboard as the input and monitoring on.
- On the track with your vocal, drums or microphone, click Route and add a send to the PsyVOX track, audio 1/2 to 1/2.
- Untick Master send on the source track if you only want to hear PsyVOX.
- Or skip the send: put the audio item on the PsyVOX track itself, and PsyVOX hears it directly.
Bitwig Studio4 and later
- Add PsyVOX to an instrument track and arm it for your keyboard.
- In front of PsyVOX in the device chain, add an Audio Receiver and set its source to the track with your sound.
- Turn the source track’s output off, or down, if you only want to hear PsyVOX.
Cubase12 and later
- Add an instrument track with PsyVOX and record-enable it for your keyboard.
- In the PsyVOX window, click Activate Side-Chain in the title bar.
- On the source audio track, add a send and choose PsyVOX’s side-chain input as its destination. Bring the send up to 0 dB.
- Make the send pre-fader and pull the source fader down if you only want to hear PsyVOX.
Studio One6 and later
- Add PsyVOX on an instrument track and arm it for your keyboard.
- In the PsyVOX window’s header, click the sidechain button and tick the track with your sound. Or, on that track, add a send to PsyVOX (Sidechain).
- Lower the source channel, keeping the send pre-fader, if you only want to hear PsyVOX.
FL Studio21 and later
- Route the source channel (vocal, drums or microphone) to a mixer track, and load PsyVOX into an effect slot on that mixer track.
- In PsyVOX’s plugin wrapper settings, give it a MIDI input port, for example 1.
- Add a MIDI Out channel in the Channel rack, set to the same port. Its notes, from the piano roll or your keyboard, now play PsyVOX.
Logic Pro needs the macOS Audio Unit build, which is not in the beta yet.
If something doesn’t sound right.
Two things tell you most of what you need. The Input meter at the left of the Vocoder section shows what is coming in. The light under the Live/Seq switch turns red and reads No input after two seconds of silence.
I play notes and hear nothing
Check the light under Live/Seq. If it reads No input, PsyVOX is not hearing any audio: follow your DAW’s steps above, and make sure the source is actually playing, or monitored if it is a microphone. To check that the synth itself works, switch to Seq; you should hear the selected sequence as you play.
The input meter moves, but the keys do nothing
PsyVOX is hearing audio but getting no notes. Arm or monitor PsyVOX’s own track for your MIDI keyboard. In FL Studio, check the MIDI Out channel and PsyVOX use the same port.
I hear my dry voice or drums as well
The source track is still going to your main output. Turn it down, send it to nothing, or untick its master send; PsyVOX still receives it through the send or sidechain. Also check the Monitor knob, which blends the dry input into PsyVOX’s own output, is at zero.
The input meter lights red
The feed is too hot. Lower Gain under the meter until the red lights only flicker on the loudest peaks.
The words are hard to make out
Keep Low Cut on for microphones, and raise Emphasis to bring out consonants. Analyzers with more bands, such as Modern/Bark 24 or Modern/Mel 48, speak more clearly than the Character ones. Shorter Attack and Release follow fast speech more closely.
Room noise or breaths make sound between phrases
Raise Gate until the gaps go quiet, and no further, or quiet phrase endings will cut off.
There is a delay when I sing live
That is your audio interface’s buffer. Set it to 256 samples or lower while playing live, and listen to PsyVOX rather than to your interface’s direct monitoring of the microphone.
A sequence doesn’t keep time with the song
Switch Sync on and start the DAW’s transport; synced sequences follow the song position. With Sync off, Speed sets the rate freely.
PsyVOX doesn’t appear in my DAW
The beta is a 64-bit VST3 for Windows. Rescan your plug-ins, and check C:\Program Files\Common Files\VST3\ is in your DAW’s VST3 paths. More on the support page.