In tune, by design.
Set TUNE to OFF, SOFT or HARD, with key and five scales. Only the undervoice is corrected, never your lead. When pitch detection is unsure, the tonal layers step back instead of guessing.
Try free for 14 days ↗NORTHFX / CREATIVE VOCAL INSTRUMENT
NEWUndervoice: your lead gets a shadow ↓Your voice, split into light.
Turn a single voice into choirs, robots, strings, alien languages, rhythmic vowel melodies and cinematic space. Now with UNDERVOICE, a synthetic shadow voice that grows under your lead. Sing into it. Catch a moment. Play it back on your keyboard.
Turn a voice into choirs, robots, strings and cinematic space.
Freeze a vocal moment and play it on your MIDI keyboard.
A deep, gritty second voice under your lead, built automatically from one vocal.
14-day free trial. Full version, no licence key needed. Buy when you are ready.
Watch the demo ↓Windows 10 / 11 · 64-bit VST3 · Contact NORTHFX for pricing and delivery.

PRISM / VOCAL DEMO
NEW / UNDERVOICE
UNDERVOICE builds a multi-layered synthetic voice from a single vocal and places it under your lead. Deep, strongly formanted, stable in pitch, gritty and robotic. It sounds like a throat, yet stays organically unstable. Your lead stays untouched while UNDERVOICE runs in parallel.

Stays untouched. UNDERVOICE runs in parallel, underneath.
A clean octave down with a phase-locked ghost sub.
Low formants, fry, growl and a robotic amp stage.
A pitch-locked carrier through PRISM's 24-band vocoder and talkbox resonances.
Micro-grains with reverse, detune and crush.
Set TUNE to OFF, SOFT or HARD, with key and five scales. Only the undervoice is corrected, never your lead. When pitch detection is unsure, the tonal layers step back instead of guessing.
ARTICULATION crossfades between voiced and unvoiced sounds, so words stay readable under the effect.
MOTION adds slow drift on eight targets. Micro-offsets between the layers keep the voice wide and human. Everything below 150 Hz stays mono.
DEPTH, FORMANT, SYNTH, STRAIN, GRAIN, MOTION and MIX get you there fast. Need more? UV LAYERS opens 33 layer controls for deep editing.

Start from a shadow that already works, then shape it with the macros. No extra latency, and your lead is never touched.
BEYOND VOCODING
Blend, morph and animate sound sources running in parallel. Five core engines (Vocoder, Talkbox, Choir, Texture and Robot) meet fifteen creative modules and the new Undervoice engine, each with a character of its own.
Catch & Play. Freeze a sung moment with CATCH and play it chromatically on your MIDI keyboard with up to six voices. Formant Hold preserves the vowel colour. Scan the slice for evolving textures or hold it as a drone.
Voice Harp. Six tuned strings resonate with your voice. Tune them from MIDI chords or let PRISM build a chord from the note you sing. Choose nylon, steel or shimmering glass.
Throat. Add a pitch-locked subharmonic one or two octaves below your voice, or isolate a whistling overtone for throat-singing tones.
Phoneme Split. Send vowels into the engines, turn consonants into octave-up stereo glitter and give breaths their own space. Keep robotic vocals intelligible.
Alien Language. Remap syllables consistently into one of sixteen invented dialects for film, games and sci-fi.
Human Texture. Add vocal fry, growl, breath and subtle pitch jitter to bring synthetic or overly clean vocals to life.
Vowel Sequencer. Stamp tempo-synced A-E-I-O-U patterns onto your voice.
Syllable Slicer. Detect, stutter, reorder or reverse syllables on the grid without manually cutting clips.
Harmony Arp. Arpeggiate diatonic chord tones or your MIDI chord in time.
Shepard Voice. Create endlessly rising or falling voices for build-ups and transitions.
Bloom. Let the last note of a phrase grow into a lush pad that gets out of the way when you sing again.
Orbit. Move your voice around the listener with distance, Doppler pitch shift and head shadow.
Radio. Tune across AM, shortwave, telephone, digital dropout and number station, with static and whistles between stations.
Swarm. Create up to 24 drifting voices, from insects and ghost choirs to stadium crowds.
Reverse Swell. Land reverse reverb on the first word of a phrase. Optional lookahead makes it automatic, without bouncing or manual edits.
Voice Modulation. Use level, pitch, vowel brightness and consonants as modulation sources for 30 targets, including space, grit, orbit speed and radio dial.
MIDI CC Out. Send your voice as MIDI CC 20–23 to control filters and parameters on other synths.
Morph Pad + Motion Recorder. Morph between four scenes on an XY pad. Record tempo-synced motion loops of 1–8 bars, and load factory sounds into a scene with a right-click.
Mutate. Generate variations in one click. Lock the sections you love; disabled engines stay off.
A/B, favourites and presets. Compare two states, mark the sounds you love and save your own presets.
Level Match & Wet Solo. Judge the effect at equal loudness, or listen to the processed signal on its own.
Scales to your screen. Resize the interface from 60 to 150% without losing detail.
THE INTERFACE
Ten focused pages and a resizable interface with live feedback. You see a phoneme timeline, the caught waveform, vibrating harp strings, vowel steps, an orbit view, a radio dial, the new layer view and a 20-channel engine matrix.
Values read in plain language, such as “180 ms”, “C2” or “MAJOR”, instead of bare numbers.

| Format | VST3 (64-bit) |
|---|---|
| System | Windows 10 / 11, x64 |
| Inputs | Mono or stereo vocal; optional stereo sidechain carrier |
| MIDI | Input for chords, Catch & Play and harmony; optional CC output (20–23) |
| Carrier modes | Auto (follows your voice), MIDI (up to 6 notes), Sidechain |
| Latency | Zero by default; optional 350 ms host-compensated lookahead for Reverse Swell. Undervoice adds no latency. |
| Sample rates | 44.1 / 48 / 96 kHz |
| Interface | 1280 × 940, scalable 60–150%, ten pages |
| Content | 21 parallel sound sources · 46 factory sounds · 191 automatable parameters |
MORE FROM NORTHFX

Every last word, thrown on time.
Explore FJORD 14 days free→
Find your way through the mix.
Explore NORTHSTAR 14 days free→
One voice. A whole session of takes.
Explore TWIN 14 days free→
Underneath it. Above it. Around it.
Explore VANTA 14 days free→