ROGERIAN SFX is a five-module channel strip for sound effects — de-click, transient, LFO tools, EQ, and compressor.
Its transient designer listens to a frequency range you set independently of the signal it shapes, so you choose what counts as a hit without filtering what anyone hears. Built on the same chassis as the ROGERIAN dialogue strip, with two modules swapped out for effects work. The DSP is written framework-free and compiles standalone, so the test harness exercises the exact code that ships. There is no oversampling anywhere in the chain.
Module framework is undergoing active optimization passes. Register your identity below for deployment telemetry upon final VST3 and AU stabilization.
GET_ROGERIAN_SFXTHE_CHAIN // SIGNAL_PATH
DE-CLICK → TRANSIENT → LFO TOOLS → EQ → COMPRESSOR.
The order is fixed: repair runs before anything downstream can shape an artifact, envelope work precedes modulation, tone-shaping follows both, and level control sits last so it answers to everything above it.
First in line, and silent until switched on.
First in the signal path, so nothing downstream shapes an artifact that should have been removed. Three algorithms — Low-Latency, Single-Band, and HQ (autoregressive interpolation across mid and wide gaps) — sit behind controls for strength, click size, widen, skew, and a detection high-pass. Disabled, it reports no latency at all.
- · LISTEN — Auditions only the material being removed, isolated from the repaired signal.
- · Visual Feedback — Repair events are marked on the display as they are found.
A detector you aim separately from the audio.
Six Linkwitz-Riley 4th-order bands split at 120 Hz, 400 Hz, 1 kHz, 3 kHz, and 8 kHz, telescoping so the split reconstructs exactly. Shaping is driven by dB-domain envelope differentials: ATTACK reads the onset, SUSTAIN the decay below peak, DECAY the late tail. Two draggable nodes set the high-pass and low-pass points the onset detector listens to, while the audio itself passes unfiltered.
- · SIMPLE Page — One large live envelope window. The detector's listening range is shaded, onset rings fire from the nodes, and depth stems charge on every hit.
- · Transient-Keyed Bells — Two peaking EQ bands whose gain opens with the live onset key and sits flat in between. No static coloration.
- · MULTIBAND Page — Six per-band envelope panes, per-band ATTACK and SUSTAIN offsets, band on/off, and LISTEN to solo one band's shaped slice.
Modulation that draws what you will hear.
Nine shapes — Sine, Triangle, Saw Up, Square, Random, Ramp Down, Drift, Exp Rise, Stairs — run free from 0.05 to 18 Hz or lock to host beat divisions, with phase offset and smoothing. They route to VOL, PAN, WIDTH, TONE (a swept tilt), and WARBLE (a modulated delay), all held under an AUTO LEVEL ceiling.
- · Deterministic Tracing — Random and Drift are phase-deterministic, so the on-screen scope traces exactly what plays.
- · Adaptive Mono — On a mono path, PAN and WIDTH go inert automatically.
- · Crossfaded Bypass — Engaging the module costs a roughly 20 ms crossfade, and nothing beyond it.
Seven bands, edited directly on the curve.
Each carries a shape — bell, low or high shelf, low or high cut, notch, tilt, band-pass, all-pass — plus frequency, gain, Q, and a dynamic range amount, so a band can ease off between hits rather than sit fixed. CHARACTER (Neutral, Cool, Sharp) reshapes the bell curves themselves instead of layering a fixed coloration over them.
- · Drag-First Editing — Drag for frequency and gain, command-drag for Q, option-drag for dynamics. Right-click changes shape.
- · Performance — Zero-phase and zero-latency parameters.
A constant lookahead, so the host never re-syncs.
Three voices — Clean, Warm, Punch — share threshold, ratio, attack, release, knee, makeup, mix, sidechain high-pass, and auto-release. The Peak-to-RMS detector is a continuous blend, not a mode switch. RANGE floors how far gain reduction may pull. The 3.5 ms lookahead budget is fixed, so moving the lookahead control changes how far ahead the gain reacts without ever changing the plugin's reported latency.
- · Saturation — CHARACTER saturation (Tube, Diode, Bright) routes pre- or post-compression and is bit-identical at Drive 0.
- · Metering — Gain reduction is metered live on the module's own display.
DESIGN_PHILOSOPHY // FEATURE_GRID
One Fixed Window
A 1116 × 640 focus-and-rack model. A single module holds the stage with its full controls while the other four stay live as mini-cards in the right rack.
Kelvin-Era Palette
Three themes share one token set — LIGHT, DARK (Void), and NUCLEAR (deep sage and aged-ivory drawn from the Caorso power station). Stored per user.
Bit-Exact Nulls
MIX at 0 returns the dry signal bit for bit, compressor saturation is bit-identical at Drive 0, and master bypass is latency-compensated and click-free.
Framework-Free DSP
31 of 31 automated harness checks pass across the transient and LFO routines against the exact DSP code that ships, not a stand-in build.
DSP_ARCHITECTURE
- Latency BudgetA constant 3.5 ms per instance — reported unconditionally, so it never shifts while you work. Transient, LFO, and EQ modules execute at literal zero-latency.
- OversamplingNone. Clean and framework-free processing throughout the entire chain.
- Transient Split6-band Linkwitz-Riley 4th-order telescoping crossovers (120 Hz / 400 Hz / 1 kHz / 3 kHz / 8 kHz) reconstructing exactly.
TECHNICAL_TARGETS
- FormatsVST3 · AU · Standalone
- PlatformmacOS [Native Apple Silicon & Intel Architectures]
- Presets10 Factory Profiles (PUNCH UP, TIGHT LOWS, HELICOPTER, etc.) alongside user-save states.
- Build Versionv0.7.2