Krate Audio / Vorago

Vorago — User Manual

Dark-Ambient Drone Instrument

Introduction

Vorago is a dark-ambient drone instrument. Each voice is not an oscillator but a small environment: a 64-partial harmonic cloud and a living noise organism, coloured by a network of wandering resonant peaks, sustained by six coupled feedback loops, and spoken through two blended acoustic bodies. Underneath every voice runs an ecosystem of agents exchanging energy and a scheduler of slow events that fire every 20 to 90 seconds. Nothing in it is scripted or looped: every movement comes from bounded stochastic processes and from the agents feeding on each other. The instrument sinks rather than floats: a subharmonic engine, a spectral fog stage and a cavern-scale space are part of the signal path, not effects appended to it.

Vorago is the dark sibling of Seraphis. Where Seraphis is weightless, upper partials and shimmer, Vorago is subterranean: subharmonics, stone and steel bodies, pressure, and geological time scales.

Vorago is slow by design. Hold the note. With the default envelope the first sound surfaces a few seconds after the key goes down and the drone keeps swelling for about two and a half minutes (the four envelope stages add up to 155 seconds). A newcomer who taps a key and hears nothing for two seconds has not found a fault; they have found the instrument. Thirty-nine of the forty-two factory presets carry that default; Sudden Chasm opens within about ten seconds, Growth Ring in a minute, Slow Bloom in two.

Formats: VST3 (Windows, macOS, Linux) and Audio Unit (macOS). One MIDI input, one stereo output.

Signal Flow

MIDI in ──────────── notes, CC64 sustain, channel aftertouch
  │
  ▼
Identity layer ───── per voice, control-rate: the ecosystem (up to 48 agents),
  │                  two slow-event schedulers, breathing and tide
  ▼
Harmonic Cloud ───── 64 additive partials + bloom children spawned from the strongest peaks
  + Noise Organism ─ four noise sources, each through resonator → comb → wandering filter
  │
  ▼
Voice Envelope ───── gates the cloud + noise only; nothing downstream is gated
  │
  ▼
Resonance Network ── 12 resonant peaks wandering in frequency, Q, gain and pan
  │
  ▼
Feedback Ecology ─── 6 coupled micro-loops (filter → delay → resonator → DC blocker)
  │
  ▼
Acoustic Bodies ──── two continuous bodies (stone, steel, hull, …) blended per sample
  │
  ▼
Voice sum ────────── + ghost tap: granular memories of the drone, an octave down
  │
  ▼
Subharmonic Engine ─ ÷2, ÷4 and fifth-below, tracking the lowest sounding voice
  │
  ▼
Spectral Smear ───── STFT fog: per-bin magnitude memory + phase decoherence
  │
  ▼
Cavern ───────────── stone early reflections → diffusion → FDN → spectral damping
  │                  → moving dampers → mix (the space is in the path)
  ▼
Output ───────────── master gain → tape saturator → true-peak limiter (always on)

The twelve macros reach across every stage of this chain at once, and the identity layer runs underneath it for as long as the plugin is alive. Both have their own sections below. The interactive diagram on the landing page shows every connection.

Interface Overview

The editor is a fixed 1100 × 760 window built around the life of the sound.

Every control carries a tooltip — hover over anything for a one-line description.

Macros

The twelve macros are the performance surface of Vorago. Each sweeps a curated set of targets across the whole engine — voice, global chain and cavern at once. Two things are worth knowing:

  1. The deep parameters set the origin the macros travel from. Moving a page parameter moves the base a macro’s sweep starts at; the macro adds its travel on top, and each target setter clamps the sum. The two never fight over a value.
  2. All macros default to neutral (0 for all, centre for Gravity). At neutral a macro contributes exactly nothing — the page parameters alone describe the sound.

Channel aftertouch is added to Pressure, so a pressure-sensitive keyboard leans into the instrument without touching the knob.

The travel columns below start from the factory defaults; a different base travels the same distance from wherever you set it.

Darkness — “Darkens the cloud tilt, the smear and the cavern”

Target Travel at full Darkness
Cloud Tilt −4 → −10 dB/oct
Space Darkness 80 % → 100 %
Smear Tilt 0 → −0.5 (the fog leans low)

Age — “Wear: duller cloud, more body damping, a shorter cavern decay”

Target Travel at full Age
Body Damping 25 % → 80 %
Space Decay 20 s → 2 s
Cloud Tilt −4 → −8 dB/oct

Density — “Richer cloud, louder and more reactive noise, deeper bloom”

Target Travel at full Density
Cloud Richness 70 % → 98 %
Noise Wake 35 % → 100 %
Noise Level −18 → −12 dB
Bloom Depth 60 % → 100 %

Movement — “Drift, wander, breathing and damper motion”

Target Travel at full Movement
Cloud Drift 8 → 50 cents
Resonance Wander Rate 0.03 → 1 Hz
Noise Wander Rate 0.03 → 1 Hz
Breathing Depth 30 % → 100 %
Space Damper Depth 35 % → 80 %

Gravity — “Pulls the resonances toward the harmonic grid” (bipolar)

Gravity is the one bipolar macro: its centre is neutral. Turning it up is stone — the twelve resonant peaks are pulled onto the played note’s harmonic grid and locked to octaves, heavy and settled. Turning it down is air — the peaks are released to float free of the note.

Target Air (full down) Stone (full up)
Resonance Gravity −1 +1
Resonance Octave Lock off on

Entropy — “Mutation, inharmonicity, decoherence and irregular breathing”

Target Travel at full Entropy
Cloud Mutation 15 % → 70 %
Cloud Inharmonicity 0.015 → 0.07
Smear Decoherence 20 % → 80 %
Breathing Irregularity 30 % → 90 %
Cloud Drift 8 → 20 cents

Pressure — “Ecology feedback, output drive and saturation”

Target Travel at full Pressure
Ecology Mix 15 % → 50 %
Ecology Loop Gain 80 % → 98 %
Output Saturation 12 % → 100 %
Output Drive 0 → +18 dB
Sub Level 0 → −12 dB (the floor pulls tight)

Weight — “Sub level, body blend and damping, a heavier spectral tilt”

Target Travel at full Weight
Sub Level 0 → +9 dB
Body Blend 35 % → 80 % (toward material B)
Body Damping 25 % → 50 %
Cloud Tilt −4 → −8 dB/oct

Fog — “Smear, ghosts, blur, tides and a foggier, darker cavern”

Target Travel at full Fog
Smear 20 % → 90 %
Smear Decoherence 20 % → 60 %
Ghost Level +40 % on the base
Ghost Blur 85 % → 100 %
Space Fog 30 % → 100 %
Space Darkness 80 % → 95 %
Tidal Depth 40 % → 80 %

Life — “Ecosystem depth, event rate and bloom spawning”

Target Travel at full Life
Eco Depth 85 % → 100 %
Event Rate 1× → 10× (fast events every 2–9 s)
Bloom Rate one every 4 minutes → one every 40 s

Depth — “Cavern mix, size, decay and width”

Target Travel at full Depth
Space Size 50 % → 95 %
Space Decay 20 s → 45 s
Space Mix, Width already at maximum by default; travel from a lower base

Mass — “Body and resonator level, body resonance and sub tracking”

Target Travel at full Mass
Body Resonance 70 % → 98 %
Resonance Mix 45 % → 15 % (the network recedes into the body)
Sub Level 0 → +6 dB
Body Mix, Sub Tracking already at maximum by default; travel from a lower base

Age and Depth share the cavern decay: Age shortens it, Depth lengthens it. Weight, Pressure and Mass all touch the sub level, and their offsets sum.

Header Controls

Control Range Default Notes
Gain 0 – 200 % unity (centre) Master output level of the whole instrument, smoothed
Voices 1 – 6 4 Maximum simultaneous voices. 5 and 6 are offered deliberately as an opt-in cost: a Vorago voice is deep and expensive (see Performance Notes)
Seed Seed 1 – Seed 16 Seed 1 Reseeds every stochastic process — the cloud, noise, resonance, ecology, ecosystem and events of every voice, the ghost tap and the cavern. Changing it re-rolls the texture without moving a parameter
Saturation 0 – 100 % 12 % Soft tape saturation in front of the final limiter. The Pressure macro rides on top of this base. The true-peak limiter always stays on
Presets — — Opens the preset browser

The Cloud Page

The harmonic cloud is the additive core — 64 partials per voice — and the bloom engine that grows new partials into it.

Parameter Range Default Description
Richness 0 – 100 % 70 % Number of active partials and how fast they roll off
Tilt −12 – +12 dB/oct −4 Spectral slope: dark below zero, bright above
Mutation 0 – 100 % 15 % How quickly the partial amplitudes slowly mutate
Inharm 0 – 0.1 0.015 Stretches partials off the harmonic grid, toward bell and metal
Drift 0 – 50 ct 8 ct Depth of the slow per-partial pitch drift
Spread 0 – 100 % 45 % Stereo spread of the partial cloud
Gravity −1 – +1 +0.1 Pulls partial energy toward the low (negative) or high (positive) end
Bloom 0 – 100 % 60 % Depth of the bloom partials
Bloom Rate 0 – 0.05 Hz 1/240 Hz How often new bloom partials spawn (default: one every four minutes)

Bloom is the minutes-scale partial generator: it reads the strongest current peaks, spawns child partials next to them, fades each in over tens of seconds, holds it, and lets it wither. Up to sixteen children live at once. At the default rate a new one buds every four minutes; the Life macro and the Bloom Rate knob bring that down to one every 20 seconds. The ecosystem’s partial agent drives bloom depth, and a scheduled event can trigger a bloom outright.

Tip: Bloom needs sounding parents — with Richness below about 50 % there are few peaks to bud from. Slow Bloom keeps Richness at 70 % for exactly that reason.

The Noise Page

The noise organism is the second sound source: four semi-independent noise sources, each feeding its own resonator → comb → wandering-filter chain. It is summed into the cloud before the envelope, so it is gated with the tone and coloured by everything downstream.

Parameter Range Default Description
Noise Lvl −96 – +12 dB −18 dB Level of the noise layer
Wake 0 – 100 % 35 % How strongly the noise follows the energy of the drone (the resting wake of every source)
Wander 0.01 – 100 Hz (log) 0.03 Hz Rate of the slow wandering of the noise colour

Each of the four slots then has its own column:

Parameter Range Default Description
Model Direct / Filtered Wind / Granular Dust / Metallic Hiss Wind, Dust, Direct, Hiss The organism the slot grows: raw colour, a wind through a wandering filter, grains of dust, or a shifted metallic hiss
Type White / Pink / Tape Hiss / Vinyl Crackle / Asperity / Brown / Blue / Violet / Grey / Velvet / Vinyl Rumble / Radio Static Brown The noise colour feeding the slot
Freq 20 – 19845 Hz (log) 60 Hz Comb fundamental
Spread 0 – 100 % 35 % Comb tooth spread
Fdbk 0 – 90 % 55 % (75 % for Metallic Hiss) Comb feedback

The ecosystem’s four noise agents wake and sleep the four sources individually, and the Density macro raises both the resting wake and the level. Four factory presets show one model each: Dead Air (Direct), Abyssal Wind (Filtered Wind), Spore Drift (Granular Dust) and Steam Vent (Metallic Hiss).

The Resonance Page

Two stages share this page: the resonance drift network that colours the excitation, and the feedback ecology that sustains it.

Resonance Drift Network — twelve resonant peaks whose frequency, Q, gain and stereo position wander on slow, bounded, individually seeded lanes.

Parameter Range Default Description
Res Grav −1 – +1 0 In Hybrid mode, pulls each peak from its free anchor onto the played note’s grid (positive) or mirrors it away from the note (negative)
Res Mix 0 – 100 % 45 % Level of the resonator bank in the mix
Res Wander 0.002 – 1 Hz (log) 0.03 Hz Rate at which the resonators wander in pitch
Anchor Free / Keyed / Hybrid Hybrid Free peaks ignore the note; Keyed peaks are retuned to the played pitch; Hybrid sits between the two, where Res Grav decides how far

Feedback Ecology — six coupled micro-loops. Each is a filter → delay → resonator → DC-blocker chain fed back into itself and, far more faintly, into its neighbours. An RMS governor holds the population below self-oscillation at every setting, so the loops sustain without ever running away.

Parameter Range Default Description
Eco Mix 0 – 100 % 15 % Level of the ecology feedback loops
Loop Gain 0 – 90 % 72 % Feedback gain of the loops
Loop 1–6 Filter Lowpass / Bandpass / Highpass Lowpass The filter in each loop

The ecosystem’s resonator agents wake individual peaks (and lift their level and wander), its feedback agents wake individual loops, and the slow events do the same in sudden swells. The Pressure macro pushes the ecology forward and tightens the loop gain.

Tip: Different filters in different loops are what turn the ecology from a hum into a population. Feedback Mire runs Lowpass, Bandpass and Highpass twice each.

The Body Page

Every voice speaks through two continuous bodies at once — physical resonators driven by a sustained stream rather than struck — blended per sample. Both bodies run at every blend value, so there is never a switch or a click, only a weighting.

Parameter Range Default Description
Material A Glass / Strings / Metal Plate / Chamber / Ice / Stone Chamber / Steel Tank / Wooden Hull / Cathedral Column / Cavern Wall / Glass Sphere Stone Chamber First body material
Material B same list Steel Tank Second body material
Blend 0 – 100 % 35 % Blend between material A (0) and B (100)
Damping 0 – 100 % 25 % Damps the bodies’ high modes, darker and shorter
Resonance 0 – 100 % 70 % How sharply and how long the bodies ring
Body Mix 0 – 100 % 100 % Level of the bodies in the mix (down for the raw network)

The six darker materials — Stone Chamber, Steel Tank, Wooden Hull, Cathedral Column, Cavern Wall and Glass Sphere — exist for Vorago; the first five are shared with Seraphis. The Weight macro pulls the blend toward the heavier material B; Mass sharpens the resonance of both.

The voice envelope also lives on this page. It gates only the excitation (cloud + noise): resonance, ecology, bodies, ghosts and the cavern all ring past note-off, and the voice retires on its own level detector once it has fallen silent.

Parameter Range Default Description
Env Mode Standard / Growth Standard Four timed stages, or one slow logistic swell
Stage 1 – 4 0 – 120 s each (log) 20 s, 30 s, 45 s, 60 s The four standard stages: the attack, then three plateaus the sound climbs through. 155 s to full by default
Release 0 – 120 s (log) 45 s Release after note-off
Growth 1 – 120 s (log) 120 s Length of the Growth-mode rise

Growth mode replaces the staged attack with a single logistic swell — the note arrives. It is the engine behind Growth Ring. Velocity scales the whole envelope.

The Sub / Smear Page

Two global stages, after the voices are summed.

Subharmonic Engine — three synchronous dividers tracking the lowest sounding voice (÷2 one octave down, ÷4 two octaves down, and a fifth below the fundamental), filtered, saturated and DC-blocked, then added to the untouched dry path. When no voice sounds the last pitch is held, so the floor never jumps.

Parameter Range Default Description
Sub Level −24 – +24 dB 0 dB Level offset of the whole sub layer (where Weight, Mass and Pressure act)
Tracking 0 – 100 % 100 % How closely the sub follows the played pitch
Oct Down −60 – +6 dB −18 dB Level of the sub one octave below
2 Oct Down −60 – +6 dB −24 dB Level of the sub two octaves below
Fifth Down −60 – +6 dB −30 dB Level of the sub a fifth below

Spectral Smear — the fog stage: an STFT in which each frequency bin remembers its recent magnitude (lows for longer than highs) and has its phase decorrelated.

Parameter Range Default Description
Smear 0 – 100 % 20 % Amount of spectral smearing
Decohere 0 – 100 % 20 % Phase decoherence of the smeared spectrum
Smear Tilt −1 – +1 0 Weights the smear toward low (negative) or high (positive) bands

The tide of every voice folds into the smear amount, so the fog thickens and thins on a minutes-long cycle even with the knob still. The Fog macro is mostly this stage plus the cavern’s own fog.

The Space Page

The cavern: an enormous underground bunker, in the signal path. Sparse stone-flavoured early reflections feed a diffusion network and an FDN reverb with spectral damping, whose damping filters wander under Brownian drift — the “moving dampers” that keep a sixty-second tail from ever sounding like a static room.

Parameter Range Default Description
Size 0 – 100 % 50 % Size of the reverb space
Darkness 0 – 100 % 80 % High-frequency damping of the space
Decay 0.5 – 60 s (log) 20 s Reverb decay time
Fog 0 – 100 % 30 % Diffusion haze inside the reverb
Damp Depth 0 – 100 % 35 % Depth of the moving dampers
Damp Rate 0 – 100 % 15 % Rate of the moving dampers
Mix 0 – 100 % 100 % Wet level of the space (equal-power against the dry)
Width 0 – 100 % 100 % Stereo width of the late field
Density 0 – 100 % 75 % Echo density of the reverb
Dimension 0 – 100 % 50 % Dimensionality of the reverb network
Breath 0 – 100 % 50 % Slow breathing of the reverb size
Early Size 80 – 300 ms (log) 220 ms Spread of the twelve early reflections
Early Lvl 0 – 100 % 80 % Level of the early reflections
Absorb 0 – 100 % 60 % Stone absorption of the early reflections (each tap has its own)
Early Send 0 – 100 % 70 % How much of the early reflections feeds the tail
Freeze on/off off Hold the reverb tail indefinitely, conserving its energy

Note that Mix defaults to 100 % — by default you hear the instrument through the cavern. The space is excited by the early reflections only, so there is a deliberate short gap before the late field blooms. Freeze is not saved into presets as on; it is a gesture you make once the drone has bloomed (Cathedral Void is written for it).

The Life Page

The identity layer: the ecosystem, the slow events, the ghosts and the life modulators.

Ecosystem — up to 48 agents of five kinds (partial, resonator, noise, feedback, ghost) live on a toroidal habitat, holding energy and exchanging it with each other, with a resource field and with a pool; nothing is ever created or destroyed. Each agent publishes one wake value that the voice routes into its own destination: a bloom slot, a resonant peak, a noise source, a feedback loop or the ghost tap. The ecosystem view in the concept band shows this colony live.

Parameter Range Default Description
Event Rate 0.1 – 10× (log) 1× Rate multiplier for the slow events: the fast scheduler fires every 20–90 s and the slow one every 3–10 minutes at 1×
Eco Depth 0 – 100 % 85 % How strongly the ecosystem agents shape the sound
Eco Sync 0 – 0.5 0 How strongly the colony’s agents fall into step
Self Affin −2 – +2 −1 How much each agent kind attracts (positive) or repels (negative) its own kind

Slow events are the “something happens” clock: a seeded scheduler draws a wait of tens of seconds to minutes, then plays one smooth attack / hold / release swell into one destination — a bloom trigger, a noise wake, a peak wake, a loop wake or a ghost burst. Two run per voice, a fast one and a slow one. Raise Event Rate or the Life macro and the drone visibly moves more often.

Ghosts — a global granular tap that records the raw voice sum and plays back twelve-second memories of it an octave below, blurred and wide. A ghost is an event, not a wash: its level is driven by the ecosystem’s ghost agent and by ghost-burst events, so with the knob up you hear them surface and sink rather than sit there.

Parameter Range Default Description
Ghost Lvl 0 – 100 % 0 % Peak level of the ghost echoes (off by default; the Fog macro raises it)
Ghost Blur 0 – 100 % 85 % How blurred the ghost echoes are
Reverse 0 – 100 % 0 % Probability that a ghost plays reversed
Ghost Events on/off off Let spontaneous events trigger ghosts on top of the base density

Life modulators — the slow autonomous processes that animate the voice.

Parameter Range Default Description
Breathing 0 – 100 % 30 % Depth of the breathing modulation (it also breathes the sub layer)
Irregular 0 – 100 % 30 % Irregularity of the breathing
Tidal 0 – 100 % 40 % Depth of the slow tidal modulation (it folds into the smear)

Presets

Click Presets in the header to open the browser. Categories appear as tabs; double-click a preset to load it. Loading restores every parameter; Save writes the complete current state into your user preset folder:

Loading a preset never reseeds an unchanged seed, so the drone you are holding is not re-rolled under you.

Factory Library

42 presets ship with the plugin across seven categories. Every one is rendered and verified during the build: each is proven to make sound, stay bounded, be measurably distinct from the others, and to make its headline feature audible. Each loads with its own note in the browser; all but three need the same patience — hold the note; the sound surfaces after a few seconds and keeps swelling for about two and a half minutes.

Drones

Preset Character
Crowded Dark A thin drone that fills to bursting, partials and embers packing the dark shoulder to shoulder
Drifting Strata Layers of drone sliding over one another like slow geological strata under a rolling tide
Monolith A single vast mass of sound, body and sub fused into one unmoving block of stone
Slow Bloom A sparse drone that takes two minutes to open, budding new partials as it breathes
Strung Abyss Vast slack strings stretched across a chasm, humming a low, dark, sustained chord
Chamber Drone A warm, closed wooden chamber filled by one breathing, slowly blooming drone
Hull Ark The creaking wooden hull of an ark rolling on a dark sea, its timbers groaning with the drone

Abyss

Preset Character
Tectonic Floor A fifth below the floor: slow plates of sub grind under a dark, close cloud
Lightless A bright cloud pressed down into lightless black, every overtone smothered as it sinks
Stone Gravity Resonant peaks dragged down and locked to octaves, heavy as stone settling in the deep
Weighted Deep Two octaves under the drone a sub swells up, pulling the whole cavern down with it
Sudden Chasm The ground gives way at once: a dark, full drone that opens beneath you within about ten seconds
Abyssal Wind Wind howling up out of a bottomless pit over a deep sub and slowly wandering peaks

Caverns

Preset Character
Resonant Shaft A deep vertical shaft whose walls sing back the played notes in slow, drifting peaks
Cathedral Void A vast, slow nave around a wide cloud. Engage Freeze once the drone has bloomed to hold the space
Endless Descent A drone falling into a cavern that keeps opening beneath it, each echo farther down than the last
Glass Well A ringing glass shaft: a bright, barely damped body singing into a pale cavern
Column Hymn Tall stone columns in a sunken cathedral, each one singing its own grave, sustained hymn
Cavern Wall Rough, wet rock walls close around a heavy drone, thudding back a dull, massive resonance

Organisms

Preset Character
Colony Pulse A restless colony of small voices, quickening and feeding on each other in the dark
Teeming A dense colony teeming in the dark, many small lives waking slowly and feeding on the drone
Bloom Colony A sparse drone where a hidden colony swells new partials into bloom and lets them wither
Singing Colony Free-floating resonant peaks that a colony wakes into song, one voice rising as another fades
Growth Ring One slow organic growth, a minute long, that keeps budding new partials as it breathes
Spore Drift Clouds of spores sifting through the dark, their grains drifting backwards into humming loops

Machines

Preset Character
Feedback Mire Choked feedback loops churning in a sump of rust, each filtered to its own grinding band
Hull Resonance The inside of a vast steel and stone hull, every plate ringing under the drone
Entropic Hum A clean machine hum slowly coming apart, its partials skewing and fraying into noise
Pressure Front A compressed wall of pressure: loops churning and the floor pulled tight into a dense, saturated hum
Feeding Loops Feedback loops that wake and feed on one another, flaring into a grinding chorus and sinking back
Iron Plate A great iron plate struck by the drone, ringing with clanging, inharmonic overtones
Steam Vent Metallic steam hissing from rusted vents over a grinding deep sub and whistling loops

Textures

Preset Character
Wind Through Basalt Raw wind and hiss pouring through cracks in black stone over a thin, dark tone
Smeared Horizon A drone dissolved into a wide, blurred haze where every partial bleeds into the next
Erosion Weathered stone crumbling to dust: crackle and grit wearing a ringing drone down to a dull husk
Swarm Breath Hiss and glassy air stirred by a swarm, flaring and settling as the colony breathes
Ice Shelf A glittering shelf of ice, bright shards of tone cracking and shimmering in the cold

Ghosts

Preset Character
Choir of Absence Voices that are not there: reversed fragments of the drone surfacing and sinking away
Fogbound Static and rumble drifting through a thick fog, the drone heard only as a ghost of itself
Haunted Colony A dark room where a colony calls up ghosts of the drone in sudden, blurred bursts
Glass Sphere A hollow sphere of glass drifting in fog, ringing faintly with the ghost of a far-off drone
Dead Air A dead channel hissing in an empty room, tape hiss and static where a voice should be

Performance Notes

Cross-Platform

Platform Formats Notes
Windows 10+ VST3 64-bit
macOS 12+ VST3, Audio Unit Universal (Intel + Apple Silicon)
Linux VST3 x86-64

State saved on any platform loads identically on the others.

Tips & Tricks

Vorago is free and open-source software from Krate Audio, released under the MIT license.

Back to Vorago | Krate Audio