AETHER//CORE MK2

IN DEVELOPMENT — Features described here include the current design and planned extensions. Final specifications and release details will be confirmed before launch.

VUHWEX AUDIO / SYNTH VST / IN DEVELOPMENT

A universe of sound.
A core you can play.

AETHER//CORE MK2 is a growing synthesis laboratory built around a fast Play page, visual sound shaping, expressive macros, and controlled instability. Start with a playable sound. Follow the interesting part.

Start with three lanes. Find the interaction.

The current build offers a single active layer with three Waveform / virtual-analog lanes. Shape and tune each lane, blend the sources, and explore pulse width, sync, sub, noise, ring modulation, and unison controls. Lane levels and panning keep the components within reach.

See the sound take shape.

Wave, Filter, Envelope, Modulation, and Scope views share the top display. Follow the waveform history, inspect the filter response, compare amplitude and filter contours, or watch the output signal while editing. A Home button and breadcrumbs keep navigation anchored.

Eight handles on a much bigger idea.

MASS · MOTION · GLASS · GRIT · DRIFT · SPACE · CHAOS · DAMAGE. Macro assignments combine multiple destinations with amounts, curves, and smoothing. Shape the performance around broad gestures, then return to the individual controls when you want to go deeper.

Safe. Weird. Wreck It.

Choose a randomization style, roll a new direction, and replay a seed when you find something worth revisiting. Section locks let you keep the oscillator, mix, filter, envelopes, modulation, shaping, effects, or macros out of the experiment. Undo, redo, and A/B controls support comparison.

Explore the current build

Waveform history and lane readouts

Filter response and shaping readouts

Amplitude and filter envelope view

Current LFO, performance-controller, and drift controls

Output scope and level display

Inside AETHER: engines, layers, and signal architecture

The sections below expand the complete development specification.

The architecture: up to twelve synthesis lanes

• Planned architecture: four independently configurable layers, each containing three engine lanes, a mixer, filters / Former processing, modulation, and effects. Each lane can select its own synthesis family, allowing hybrid combinations rather than a single engine choice for the entire patch.
• Layer performance modes: stack, key split, velocity split, round-robin, and random-robin, with controllable assignment priority, cycle reset, and random seed behavior.
• Voice design: global voice allocation, configurable polyphony, graceful voice stealing, unison, detune, stereo spread, per-note pan, glide, and mono / legato / poly modes. Complex patches require a CPU budget; twelve lanes do not imply unlimited simultaneous voices.
• Current engine: one active layer, three Waveform / virtual-analog lanes, mixer, unison, filtering, basic modulation, macros, randomization, and chorus / delay / reverb controls. Additional engines and deeper pages are explicitly marked as later phases in its interface.

Wavetable engine — planned

• Band-limited single-cycle tables, interpolated playback, smooth table-position scanning, phase, coarse / fine tuning, sync, unison, warp / morph, and spectral shaping.
• Original harmonic, vocal / formant, metallic, noisy, and evolving tables, plus optional sub and noise sources.
• User-audio import and analysis into tables where practical, preview before import, and independently saved table building blocks.
• Evolve extension: adjustable partial / operator / noise components define wavetable waypoints, with modulation of those recipes. This is an original development direction, not a claim to reproduce proprietary Waldorf Seed internals.

Waveform / virtual-analog engine

• The current lane interface shows waveform selection / morphing, pulse width, coarse and fine tuning, level, pan, and lane enable controls.
• The waveform engine uses band-limited sine → triangle → saw → pulse morphing, PWM, hard sync to Lane 1, sub, noise, and ring modulation.
• The mixer exposes sub level, noise level, ring 1×2, and sub waveform selection. The unison section exposes voice count, detune, and spread.
• Planned additions include phase behavior, drift / character, additional original forms, and stable FM where practical. These are design targets rather than independently verified specifications.

Particle engine — sample and granular roadmap

• Single-sample playback, mapped multisamples, transient playback, and true granular playback as distinct modes.
• Start / end / loop markers, loop crossfade, reverse, root note, key tracking, sample gain, and stereo-aware playback.
• Grain size, density, position, spray, window, pitch scatter, envelope, and per-note variation.
• User sample drag-and-drop, waveform display, missing-file resolution, and safe long-file streaming. Particle is not presented as available in the current waveform-only build.

Resonator engine — planned

• Exciter plus modal, string / comb, bar / plate, tube, and diffuse / noisy structures.
• Excitation from noise, impulses, an oscillator, sample transients, or another lane through the routing architecture.
• Fundamental pitch, decay, inharmonicity, damping, brightness, excitation, stereo width, and modulation controls.
• The intended result is a tunable resonating body within a hybrid patch—not simply an EQ preset labeled as physical modeling.

Kernel FM engine — planned

• Up to six operators per lane, with audio-rate FM, AM, ring / phase interactions, and bounded feedback where stable.
• Graphical routing, readable algorithm presets, free connections, per-operator ratio, fine tuning, level, envelope, and output assignment.
• Original algorithms and sound content. Compatibility imports are conditional on practical format support; proprietary factory banks are not promised.

Cross-engine routing — planned

• Lane-to-lane audio and envelope routing: FM / AM / ring modulation, wavetable position, granular position / density, resonator excitation, and filter / Former input where technically meaningful.
• Audio-rate and slower modulation paths, with bandwidth and stability safeguards. Arbitrary recursive feedback is not implied.
• Serial, parallel, split, and per-layer paths, with explicit lane level / pan and gain staging.
• Example design possibilities: an oscillator excites a resonator; a granular source feeds a shaped filter path; wavetable and FM lanes contribute different transient and sustained components within one layer. These illustrate the planned architecture.

Modulation and expression

• Current modulation: one LFO with sine, triangle, saw, square, and sample-and-hold choices; sync and retrigger; pitch, cutoff, wave, pulse-width, and amplitude amounts.
• Current performance-controller controls include mod-wheel vibrato / cutoff, aftertouch cutoff / vibrato, and Variation / Drift amount and rate.
• Planned expansion: six or more LFOs, six envelopes, an editable multi-stage complex modulator, and a target of at least 40 matrix slots per layer where feasible.
• Matrix design: source, destination, depth, polarity, curve, smoothing, and per-voice / global distinction. Sources include velocity, key tracking, pressure, CC, macros, random-per-note / trigger, note age, sequencer lanes, and envelope followers.
• Modulation inspector: show every route affecting a selected parameter, with searchable sources / destinations and direct route creation.

Per-note locks and performance — planned

• Select or record a MIDI note, then attach up to 16 parameter overrides for engine selection, tuning, oscillator position, filters, FX, or macros.
• Piano-roll indication, copy / paste, clearing, lock amount, audition, and a bypassable lock system.
• XY performance pad, MIDI learn, aftertouch / MPE mapping where supported, momentary / latch behavior, chord capture, and on-screen notes / pads.
• Polyphonic arp / sequencer with step pitch, gate, velocity, probability, ratchet, parameter locks, and host sync. MIDI output remains dependent on host support.
• Microtuning direction: .scl / .kbm import where practical, per-patch tuning, reference pitch, and a reset to 12-TET.

Sample Lab and portability — planned

• Batch import, draggable key / velocity zones, root-note detection, preview keyboard, overlap / fade controls, batch transpose / normalize, and naming tools.
• Auto-map by note names in filenames, with manual confirmation and velocity-layer naming support where practical.
• Waveform zoom, audition, trim, crossfade loops, gain, reverse, transient markers, and undoable editing that preserves the source file.
• Autosampling direction: record a note / velocity range, trim, normalize, detect silence, and build a playable map. Host-plugin capture is conditional; reliable input capture is the minimum intended fallback.
• Portable preset bundles collect referenced samples, report dependencies and size, check missing files, and support relinking. Library paths, tags, favorites, and recents support ongoing work.

Filters, Former, and effects

• Current Play page exposes filter cutoff, resonance, envelope amount, drive, amp envelope, variation, and chorus / delay / reverb controls. The filter display also shows Former-related readouts.
• Planned architecture: two stereo digital filters per layer / voice with low-pass, high-pass, band-pass, notch, comb, and state-variable-style options; 12 / 24 dB slopes where appropriate, key tracking, stereo controls, and modulation.
• Former roadmap: saturation, bit-depth / sample-rate reduction, ring modulation, wavefolding, frequency shift, and comb / metallic processing.
• Expanded reorderable FX roadmap: chorus / dimension, phaser, flanger, tempo delay, reverb, EQ, compression, distortion / saturation, limiting, glitch / gate, pitch-shifting delay, and spectral smear / freeze where feasible.
• External-audio processing and per-layer FX routing are planned; they are not advertised as verified features of the current build.

Macros, randomization, and patch workflow

• MASS, MOTION, GLASS, GRIT, DRIFT, SPACE, CHAOS, and DAMAGE. The current macro editor shows multiple destinations, signed amounts, response curves, smoothing, factory mapping, and clear controls.
• Safe / Weird / Wreck It randomization, seed replay, and section locks are visible in the interface. Broader targets include per-control locks and patch / layer / lane / engine / modulation / sample / FX / sequence / macro randomization.
• The development direction includes Save / Save As, dirty-state indication, undo / redo, A/B comparison, previous-patch recall, autosave recovery, searchable tagged libraries, and missing-sample warnings.
• Factory content target: roughly 100 useful original presets if development time allows, prioritizing meaningful sounds and macro mappings over a count. No completed or auditioned bank is asserted.
• System formats, sample-rate support, automation coverage, performance limits, and host compatibility remain release-validation items, not guarantees from the development design.

AETHER//CORE MK2 is an original VUHwex Audio instrument inspired by broad multi-engine synthesis concepts associated with the Waldorf Iridium family. It is not affiliated with or endorsed by Waldorf, and exact hardware emulation is not claimed.