Expandable polyphonic synthesizer

Shape motion.
Build sound.

A deep, playable synth where oscillators grow with the idea—and every route, movement and effect stays within reach.

Download the free beta Explore the engine

MaxVRS VST3 · AU 0.9.x Beta · Free

The actual MaxVRS plugin interface running on Windows

Up to 32 voice polyphony

6 oscillators per group · up to 3 layered groups

Explore

Hear MaxVRS

Press play.
Then pull it apart.

Play the full mix, then solo the bass or mute the drums to hear exactly what each MaxVRS sound adds.

Browse the stem library

AI Sound Assistant · new in 0.9.12

in a synth?
What?

MaxVRS is an AI-powered synth: describe the sound in your head, and the AI Sound Assistant builds it, or changes the patch you are playing. The sound itself stays real synthesis, open for you to tweak, and the assistant explains how it is made.

Speaks MaxVRSIt knows every oscillator, filter, route and effect, and uses them as a sound designer would.
Listens before it answersIt renders the patch and measures its level, brightness, movement and stereo width before it replies.
Learn with AIAsk how a sound is created, and it walks you through the oscillators, envelopes, routes and effects behind it.
You stay in chargeEvery request is one Undo step, and your ears have the last word.

Try it in the free beta

Works with your own Claude or ChatGPT subscription or API key, through Claude Code or OpenAI Codex on your computer.

AI Sound AssistantExample

YouA warm pad that slowly opens, with a little shimmer.
AssistantTwo Harmonic Cloud oscillators, a slow filter sweep on MOD ENV 1, Chorus and a long Reverb. Rendered: a soft attack that opens over about four seconds.
YouDarker, and put the brightness on the MOD wheel.
AssistantCutoff down an octave; the MOD wheel now opens it again.
YouHow does the slow opening work?
AssistantMOD ENV 1 rises over about four seconds and is routed to the filter Cutoff, so each chord starts dark and opens while you hold it.

Inside the engine

Make the signal
your own.

Every section is made to reveal depth when you need it and stay out of the way when you do not.

01 / Source / Oscillators

Build the voice at the source.

Start with one Wave Factory oscillator, expand to six per group, and layer up to three independently shaped groups.

Wave Factory

Morph continuously from Sine through Triangle and Saw to Square. Shape the result with Warp, Fold, Asymmetry, Skew / Width, Curve, Smooth / Edge, and Harmonics while comparing the original and modified waveforms live.

Harmonic Cloud

Choose Classic, Shimmer, Hollow, Steel, or Vintage character, then shape Voices, Detune, stereo Spread, Focus, Drift, and Tone. The voice profile and stereo-width display follow live modulation.

Noise

Choose White or Pink noise directly from the preview, then shape its spectral balance with Color and its activity with Density. Watch the resulting texture in the live display.

Wavetable

Browse 32 built-in tables across eight categories. Scan their frames with Position, reshape the waveform with Warp, and use Phase and Unison to make the result precise or expansive.

Sample / Chop

Import WAV, AIFF, FLAC, or OGG for One Shot, Loop, Ping Pong, Slice, or Granular playback. Use grid, transient, or manual slicing and edit pitch, level, pan, probability, and ratchets per slice. Keep Length preserves phrase timing as you play different notes with Keytrack enabled.

Piano

Play the built-in Electric and Felt models immediately, or download or locate the optional Salamander Grand Piano bank from the oscillator. Shape the response with Tone, Decay, and Touch, plus the shared tuning, level, and pan controls.

Sample Harmony

Add three scale-aware harmony voices with individual toggles, intervals, and levels. Balance them with Original and Blend, widen with Spread, and shape the grains with Texture. Choose a local Key and Scale or follow the arpeggiator scale.

Sample Formant

Preserve vocal character during pitch changes or shift its resonances by up to ±12 semitones. Formant processes conventional sample playback and Harmony voices; in Granular mode it applies to Harmony only.

Group Filter

Shape the combined group with low-pass, high-pass, band-pass, notch, Formant, Comb, or Ladder filtering at 12 or 24 dB/oct. Adjust Cutoff and Resonance directly on the interactive response graph.

Trance Gate

Give each group its own rhythmic gate with up to 32 steps, On / Off / Tie states, and editable step levels. Sync the pattern and shape each pulse with Depth, Attack, and Release.

Group strip: Glide enable and time, Legato behaviour, lane destination, group level, oscillator count, and the add-source control live together above the stack.
Source engines: Choose Wave Factory, Harmonic Cloud, Noise, Wavetable, Sample / Chop, or Piano for each oscillator slot. Click a source preview to browse its waves, tables, or sound styles. Every edition plays presets made with the Pro engines.
Oscillator header: The colour stripe handles enable, mute, and solo. The header also exposes actions, help, Level, Pan, Octave, and Fine tuning.
Wave Factory: Morph continuously through Sine, Triangle, Saw, and Square, then shape the result with Warp, Fold, Asymmetry, Skew or Width, Curve, Smooth or Edge, and Harmonics.
Harmonic Cloud: Build a dense bank of partials with Voices, Detune, and stereo Spread, then focus its spectrum and introduce organic movement with Drift and Tone.
Live waveform: The graph compares the computed source with the modified signal, so shaping and incoming modulation remain visible while you play.
Phase: For periodic sources, Varied uses Phase Spread to control how far starting phases differ. Fixed restarts each note at the selected Start angle for repeatable attacks.
Unison: Stack supported Wave Factory, Wavetable, Sample / Chop and, in Pro, FM Lab voices with Voices, Detune, and stereo Spread. Harmonic Cloud has its own voice controls.
Oscillator modulation: Route an oscillator into a later source using Linear FM, Exponential FM, Phase, AM, Ring Mod, Hard Sync, PWM, or Wavefold, with destination, lock, and depth controls. Polarity and Tone shape each route; Hard Sync adds Softness.
Amp Envelope: Shape each group independently with a curved ADSR editor, Raw or Smooth transient response, and a One Shot mode for percussion.
Wind Articulation: Add a filter Bite, pitch Scoop, and delayed Vibrato to a group. Set vibrato Rate and Delay for expressive attacks and sustained notes.
Oscillator actions: Duplicate, copy, paste, or save individual oscillator presets. Randomize Tone, Movement, or Everything Safe while preserving source selection, levels, routing, and sample structure.
Preset library: Search and browse Factory and User collections by category, inspect descriptions and named macros, and save complete .mvrs patches; Pro presets are marked PRO. Save individual sources as .mvrsosc presets, or return to a saved patch with Reset.
Undo and Redo: Step back through up to 100 changes of the whole patch: knobs, modulation routes, patterns, samples, and the oscillator and effect structure. ⌘Z / Ctrl+Z and ⇧⌘Z / Ctrl+Shift+Z.
AI Sound Assistant: Describe a sound or a change and the assistant builds or edits the patch, then renders and measures it before it answers. Uses your own Claude or ChatGPT subscription or API key; each request is one Undo step.
Group filter: Choose low-pass, high-pass, band-pass, notch, Formant, Comb, or Ladder. Formant shapes a vowel from A to U, Comb rings on the note like a string, and Ladder is a classic four-pole low-pass that self-oscillates and drives. Set 12 or 24 dB/oct slope, Cutoff, and Resonance against the live response graph.
FM LabPRO: Six-operator FM synthesis with the shared Unison controls, for bells, electric pianos, metallic leads, and deep basses.
Drum modulesPRO: Kick, Snare & Clap, and Hats & Metal engines with classic and modern models, shaped with Tune, Decay, Punch, Click, Drive, and Accent. DRUM ENV sets a one-shot envelope.
FormantPRO: A pulse train through five vowel formants for voices, choirs, talking basses, and croaks, with Vowel, Size, Breath, and Jitter.
ResonatorPRO: Physical models: plucked and bowed strings, tubes, and modal bodies from wood and bars to bells, glass, plates, and skins. PLUCK ENV lets each note ring down.
GrainsPRO: A cloud of short grains from a wavetable, a sample, or Live: the last second of what earlier oscillators play. Size, Density, Spray, Pitch Spray, Reverse, Scan, and Freeze.
ExcitePRO: Let any oscillator play a Resonator in place of its own pluck or bow, or feed Live Grains: noise bows a string, a formant becomes a talking drum.

02 / Motion

Movement with a visible purpose.

Add only the modulators a patch needs, inspect their shape, and manage every destination from either side of the route.

Actual MaxVRS Motion workspace showing a modulation envelope, source stack, and open influence list — Actual plugin UI · Motion workspaceCaptured from MaxVRS

Mod Envelope

Shape modulation with Attack, Decay, Sustain, and Release while editing the response directly in the visual envelope. Assign and route controls stay available in the header.

LFO

Choose from five waveforms and Retrigger, One Shot, or Free Run modes. Run freely in hertz, up to 200 Hz with FAST HZ, or lock to tempo, then use Delay and Fade to bring movement in gradually.

Step Mod

Draw a bipolar sequence with 4, 8, 16, or 32 steps. Choose Forward, Reverse, or Ping Pong travel and a free or synced rate, then use Smooth to soften transitions.

Random

Generate repeatable random motion in Hold, Glide, or Smooth shapes. Control its speed, soften changes with Slew, and regenerate the sequence whenever a new variation is needed.

MSEG

Build a custom shape from two to eight points with individual segment curves and a global curve. Loop or play once with free or synced timing, starting from Rise, Fall, Triangle, Swell, Pluck, or Pulse presets.

Mod Envelope: A curved ADSR modulation source with Attack, Decay, Sustain, Release, and a direct visual envelope editor.
LFO: Sine, Triangle, Square, rising Saw, and falling Saw shapes with Retrigger, One Shot, or Free Run modes, free or synced timing, FAST HZ up to 200 Hz for trills and buzzing, Delay, and Fade.
Step Mod: Draw 4, 8, 16, or 32 bipolar steps, choose Forward, Reverse, or Ping Pong travel, use free or synced timing, and smooth transitions between steps.
Random: Generate repeatable seeded movement with Hold, Glide, or Smooth shapes, rate or division timing, Slew, and Regenerate.
MSEG: Edit two to eight bipolar points, segment curves, and a global curve. Loop or play once with free or synced timing, or start from six shape presets.
Arpeggiator: Build 8, 16, or 32-step patterns across up to eight note rows, with ties, step velocity, Gate, and note or rhythm presets. Choose host or note sync and a musical scale.
Follow Chord: Let the arpeggiator recognise major, minor, and common extended chords. Root and scale update at step boundaries while your pattern and rhythm keep playing.
Assign: Arm a source and touch a compatible synth, lane-effect, or Motion control to create a route. Modulate supported modulator controls to build evolving movement.
Routes / Influence List: Manage up to 64 routes from the source or destination. Set depth from −200% to +200%, invert, mute, or remove each route while keeping the saved base control value.

03 / Lane Effects

Finish each layer independently.

Build ordered racks across two effect lanes with 15 effect types, six more in Pro, live visual feedback, and in-context control guides.

Actual MaxVRS Lanes workspace showing Stereo Delay, Reverb, Phaser, and EQ effect cards across two lanes — Actual plugin UI · Lanes workspaceCaptured from MaxVRS

Lane strip: Each lane has independent output routing, level, scrolling, and an ordered rack. Effect stripes provide enable, mute, and temporary solo.
Delay: Digital, Ping-Pong, Tape, and BBD modes with Time or Division, Feedback, Tone, and Mix; Advanced adds Pitch, Diffusion, Taps, Reverse, and Freeze.
Stereo Delay: Independent left and right timing, feedback, and mix, with Link, Tone, and cross-feedback from left to right or right to left.
Reverb: Room, Hall, Gentle Hall, and Plate spaces with Size, Damping, Decay, and Mix; Advanced adds Pre-delay, Diffusion, Width, Low Cut, and Modulation.
Chorus: Classic, Ensemble, and Dimension voices with Rate, Depth, Width, Voices, Tone, Feedback, and Mix plus Advanced delay, filtering, shape, and drift.
Filter: Low-pass, high-pass, band-pass, notch, Formant, Comb, or Ladder filtering with 12/24 dB slope, Cutoff, Resonance, Drive, and Mix.
Drive: Soft, Hard, and Asymmetric saturation with Drive, Tone, Level, and Mix; Advanced exposes Character and Bias.
Phaser: Two or four all-pass stages with Rate, Depth, Feedback, and Mix for focused movement or a deeper sweep.
Flanger: Free or synced Rate, Depth, Feedback, and Mix; Advanced adds Manual delay, Polarity, Stereo phase, Shape, and Tone.
Tremolo / Auto-pan: Rate or Division, Depth, Stereo Phase, and Mix, with Advanced Shape, Symmetry, Phase, and Gain Law.
Compressor: Stereo-linked dynamics with live input/output history and a transfer curve. Set Threshold, Ratio, Attack, Release, Mix, and Peak/RMS detection; Advanced adds Knee, Makeup, detector HPF, and Auto Release.
EQ: Low and High shelves plus a Mid bell, with frequency, gain, and Q controls for every band, Amount, and Output always visible. Drag graph points, adjust Q with the mouse wheel, and enable the before/after FFT view.
Bitcrusher: One to 24-bit quantisation and 1x to 64x downsampling with Tone and Mix. Compare waveforms or FFT spectra; Advanced adds Drive, Dither, Jitter, filter placement, and Output.
Stereo Imager: Control Width, Pan, Stereoize, Bass Mono, and Output with a live stereo scope and correlation meter. Drag Pan on the graph, adjust Width with the mouse wheel, and check the result with temporary Mono audition.
Shimmer: A pitch-and-feedback ambience with Size, Shimmer, Tone, and Mix; Advanced adds Decay, Pitch, Pre-delay, Width, and Low Cut.
Frequency Shifter: Move every frequency by the same number of hertz for slow beating, metallic, bell-like, and alien tones, with Feedback spirals, Stereo, and Up, Down, or Both sidebands.
TapePRO: A tape machine: Drive, Wow, Flutter, Age, and Hiss that follows the signal, for warmth and drift.
MultibandPRO: Split a lane into low, mid, and high; each later effect can work on one band, with optional OTT-style compression per band.
ResonatorPRO: Sympathetic resonance: up to six strings tuned to the held notes or a chord, or a modal body, ringing with the lane.
Grain DelayPRO: Grains of the lane’s last two seconds played back from a delay time: scattered, transposed, reversed, fed back, or frozen.
Beat RepeatPRO: Tempo-synced stutters: a captured slice repeats with Decay, Pitch, and Reverse, on the host’s bars or from each note you play, with Hold for live rolls.
Transient ShaperPRO: Bring hits forward or push them back at any level: Attack and Sustain up or down, Time, and Output.

04 / Perform

Turn a patch into an instrument.

Keep the most expressive controls on a compact rail that stays available beside the sound-design workspaces.

Actual MaxVRS Perform rail showing four macros and the modulation wheel beside Source, Motion, and Lanes — Actual plugin UI · Perform workspaceCaptured from MaxVRS

MOD wheel: A live MIDI CC1 source mirrored by the vertical on-screen wheel. It shares the same Assign and Influence List workflow, while its current position remains performance state rather than preset state.
Macros 1–4: Four independent 0–100% host-automatable performance sources, each with its own saved value, display name, assignment state, and route count. Pair related destinations or drive multiple synth and lane-effect targets with individual bipolar amounts.
Velocity: Use incoming note velocity as a performance source to make patches respond dynamically to how softly or firmly each note is played. Assign it to synth and lane-effect controls with individual bipolar amounts.
Pressure: Turn incoming MIDI pressure into expressive real-time modulation. Route it through the same Assign and Influence List workflow to shape timbre, movement, effects, or level while holding a note.

Shape motion.Build sound.

Press play.Then pull it apart.

AI in a synth?What?

Make the signalyour own.