Audio Damage AD-202 [WiN-MAC-LiNUX]

• Product: AD-202
• Developer: Audio Damage
• Version: 1.0.0
• Format: VST3, AU, AAX, CLAP, LV2
• Requirements: Windows 10 or later, macOS 10.13 or later, Ubuntu 20 or later
• Source: audiodamage.com/collections/instruments/products/ad066-ad202

Grey-box monosynths from the early 1980s ran one oscillator into one filter and asked nothing else of the player — that’s the shape AD-202 revives, and the reason it’s the plugin under review here: a single VCO with sub oscillator and noise, feeding a 24 dB ladder filter with its own drive stage, one shared envelope, and a delayed-onset LFO. Eight-voice polyphony, MPE, and MTS-ESP microtuning sit behind a one-knob-per-function panel, while a browser demo lets you try the actual synth engine before downloading it.

Key Takeaway

AD-202’s real differentiator is the delayed LFO onset: set it slow with a few seconds of ramp and held notes drift into vibrato without an automation lane. The instrument stays a single-oscillator voice throughout, so anyone needing a second oscillator, wavetables, or a built-in sequencer is choosing the wrong grey box.

Four Sources Into a Single Clip

The VCO carries a sawtooth and a two-saw pulse from one oscillator, switchable across four footages — 16′, 8′, 4′, and 2′ — with pulse width open to manual control, the LFO, or the envelope. A flip-flop sub oscillator hangs off that same VCO, running one or two octaves down or as a two-octave pulse, staying filtered and DC-blocked the way the original divider chain did.

Saw, pulse, sub, and noise land in a shared mixer with independent levels. They sum into a soft clipper before the filter ever sees the signal. Pre-Filter Sat then drives that mixer up to eight times harder without adding makeup gain — the same overdrive a session would get by running every source wide open by hand, except reachable from one control.

A single VCO caps the raw harmonic material at whatever that one oscillator and its sub can generate; there’s no second oscillator to detune against it or cross-modulate. Anyone patching thick unison leads or layered FM textures is working against the architecture, not with it. A bass line, a lead, or a gritty mono line built from one saw and one sub is the actual job here.

No Level Penalty for the Filter Drive

The ladder filter is a resonant four-pole lowpass with soft clipping built into its own feedback path, pushed by a dedicated Drive control. Cutoff answers to the envelope, the LFO, keyboard tracking, and velocity, each carrying its own independent amount rather than a single global depth.

Gain compensation keeps output level from sagging as Drive climbs, so pushing the filter hard doesn’t also mean riding a fader to compensate. That decouples the filter’s grit from its loudness, which matters most under automation — a rising Drive sweep across a build stays at roughly the same perceived volume instead of both brightening and quietly dropping in level.

Resonance and Drive together can still get loud fast at the input stage, since four sources are already summing and clipping before the signal reaches the filter at all. Gain-staging the mixer first, then leaning on Drive, keeps the two clipping stages from fighting each other in a way that just sounds muddy rather than aggressive.

An Envelope That Remembers Its Last Position

One ADSR runs both the VCA and the filter cutoff, built from time-based exponential segments rather than linear ramps. Its retrigger behavior copies the original hardware’s quirk directly: a new attack starts from wherever the envelope currently sits, not from zero, so fast repeated notes stack rather than reset.

Three trigger modes decide what “repeated notes” even means. Gate takes the lowest held note and reuses the same voice; Gate plus Trig takes the last note played and retriggers the envelope on it; LFO Trig retriggers the envelope once per LFO cycle, producing the rolling, self-arpeggiating pattern the original grey box was known for.

There’s no per-segment curve shaping and no second envelope for independent filter and amplitude contours — one ADSR does both jobs, always. A patch that needs the filter closing while the amplitude is still open, or vice versa, needs a workaround through velocity or aftertouch routing instead of a dedicated envelope.

Vibrato Arrives After the Note Does

The LFO offers triangle, square, stepped random, and sample-and-hold shapes, running free from 0.1 Hz to 30 Hz or locked to host tempo. Its Delay control ramps LFO depth in over as much as four seconds from each new gate, so a held note starts still and only picks up motion a beat or two later.

That delay is the sleeper control on the panel. A slow triangle with a couple of seconds of ramp gives a note time to speak before vibrato blooms on its own — the kind of expressive timing that’s tedious to draw by hand in an automation lane and nearly free to play by holding a key. Optional per-note retrigger resets that ramp on every new note in a poly patch, rather than only on the first.

The tradeoff is that the LFO section owns exactly one instance per voice; there’s no second LFO for an independent pan or amplitude wobble running on its own schedule. Whatever the LFO is already doing to pitch, pulse width, or cutoff is the only modulation shape available at any given moment.

Eight Voices, Never Quite Twins

Poly mode allocates the same single-oscillator voice eight times over, spread across the stereo field by an equal-power Spread control instead of stacking every note dead-center. Portamento carries into poly playing too, with up to five seconds of glide, an Auto mode that only glides between overlapping notes, and a fixed Tune offset of plus or minus 50 cents.

Age adds fixed random drift to pitch, pulse width, cutoff, LFO rate, and mixer level, rolled fresh per voice every time a plugin instance loads. With Spread and Age both up, a poly patch stops reading as eight copies of one oscillator and starts reading as eight related but slightly mismatched ones — closer to a room of eight of the same synth than one synth doubled.

Mono mode keeps a separate identity from poly rather than acting as poly with one voice: it holds the original note-priority stack, legato behavior, and auto-glide rules a monosynth player would expect, untouched by anything Spread or Age are doing on the poly side.

Four Places for One Kind of Touch

MPE support covers per-note pitch bend and pressure, and standard aftertouch routes to four destinations independently — oscillator mod depth, filter mod depth, filter cutoff, and LFO rate — each with its own attack and release smoothing on the pressure signal. MTS-ESP microtuning retunes continuously from any compatible master, applied per note and per voice, with portamento glide preserved across a retune rather than snapping.

Aftertouch smoothing matters more here than the spec sheet implies: a fast attack smoothing setting turns pressure into an immediate filter shove, while a slow one turns the same physical squeeze into a gradual swell, and the two feel like different instruments under the hand even though nothing else in the patch changed.

None of this expressive routing does anything on a controller that can’t send MPE or channel-wide aftertouch in the first place — a standard velocity-only keyboard gets the plugin’s sound but not its touch-responsive layer.

A One-Oscillator Voice, On Purpose

A second instance layered underneath handles anything the single VCO can’t cover on its own, and Age still rolls independent drift across five parameters — pitch, pulse width, cutoff, LFO rate, and mixer level — fresh every time either copy loads.

FAQs