Checkout and protected delivery are not open yet. Check the product page for current availability.

ZPS Performance Glitch 1.0.0

Performance Glitch captures incoming audio and rearranges it into bursts, repeats, reversed fragments and other controlled disruptions. The artwork, sound and live displays in this release are the accepted Performance Glitch implementation. Nothing generates sound without an input signal.

Quick start

Load ZPS Performance Glitch as a VST3 effect on an audio track. Begin at a comfortable monitoring level. Let a little audio pass through to populate the history, then turn Engage on. Original mode is the default. To explore the expanded engine, click MODE in the lower screen to select Shaped, then choose a Character and Pattern.

Click RESTORE ORIGINAL in the lower screen to recover the exact legacy engine. Activity, Damage, Dry/Wet, Engage, Bypass and the saved shaped settings are retained. Changing MODE back to Shaped recovers those settings. A Character or Pattern named Original inside Shaped mode is not a substitute for MODE: Original.

Main controls

Drag knobs vertically. Hold Shift for finer changes. Wheel and arrow keys adjust values; double-click or Home returns the control to its default. Dry/Wet defaults to 100%; its ordinary wheel/arrow step is 1%, or 0.1% with Shift.

Engage is the approved square on-screen toggle. Existing host/controller momentary mappings continue to address the same Engage parameter. Engage off, Bypass on or Activity zero recovers exact dry input within the existing 64-sample transition. Neither bypass nor a dry mix adds an output gain stage. Dry/Wet at zero keeps the engine capturing/processing so it can be blended back in.

Slice, Burst, Variation, Character and Pattern shape the expanded engine when MODE is Shaped. Structural changes take effect at the next event; the screen shows pending status. Damage and Dry/Wet remain live.

Character and Pattern

Original Character retains sample hold and truncating quantization. Dark uses two low-pass stages. Digital provides stronger sample hold and bit reduction. Tape provides saturation and darkening without added noise or makeup gain. Damage zero removes Character coloration, while Pattern can still move audio.

Tape Pattern creates forward splices with fractional-speed playback that slows across a burst. Tape Character and Tape Pattern are separate treatments, applied once each. This is a creative tape character, not a claim of reproducing a particular physical machine.

Sequential cycles through Repeat, Reverse, Jump and Crush. Stutter Fade repeats a fragment with a falling envelope. Random increases randomized decisions as Variation rises. Shaped Original uses legacy-style family selection with the expanded length controls. Select MODE: Original for exact legacy operation.

FAMILY inside the lower screen can force Repeat, Reverse, Jump or Crush instead of Mixed. A forced family naturally reduces differences between patterns that otherwise select families.

Timing and recall

Free uses Activity-based gaps between bursts. Sync follows valid playing host tempo and beat position at 1/4, 1/8, 1/16 or 1/32-note grids. It displays Free fallback when the host is stopped or valid musical timing is unavailable. First Engage can begin immediately once enough source audio exists; subsequent events follow eligible grid crossings. Events do not stack into a backlog.

LOCK repeats the stored sequence of decisions every eight event starts while drawing from fresh audio. NEW SEED changes the saved seed. Recall restarts the sequence; it does not save the live capture buffer or the current event phase. Original mode ignores the additional shaping controls.

Existing plugin identity, parameter IDs, automation and MIDI mappings are preserved. Legacy pre-0.2 states migrate to Original mode; 0.2.x states retain their saved mode and settings. Saved Dry/Wet is retained where present. Versions 0.2.0, 0.2.1 and 1.0.0 share schema 3. Older pre-0.2 binaries cannot read schema-3 states, so preserve original projects when testing a rollback.

Reading the live displays

IN and OUT show actual stereo min/max envelopes over four seconds at a fixed ±1 full-scale range. OUT is measured after Dry/Wet. Slice marks actual source history and read position. Stutter/repeat marks, burst spans and variation measurements come from the engine's actual events. The lower screen shows past event activity, seed, lock, timing/fallback and pending state.

Rapid text reports real sample/event identifiers, source/read positions, direction, lengths, repeat count and host format. It is diagnostic data, not fictional commands. Idle, silence and unavailable data are represented truthfully; event history may remain after audio stops. DROPS reports skipped visualization records if the bounded display queue fills. Audio does not wait for the display.

L/R output meters show post-blend sample peaks with decay and peak hold. They are not true-peak or loudness meters. The plugin does not normalize or automatically raise output volume. SIZE cycles 50%, 75% and 100% while preserving the 3:2 composition. Default size is 1152 × 768 at 75%. Display updates stop while the editor is closed, hidden or minimized.

Support

Contact [email protected] with product version, operating system, host/version, sample rate, buffer size and steps to reproduce a problem. Do not include passwords or account tokens. The plugin itself does not require a network login, subscription or online activation. Screen telemetry stays inside the plugin; it is not transmitted analytics.