Processes

For step-by-step notes on selected processes, see the Process Guides. The Utilities pages cover the browser-based companion tools for Automation Score breakpoint lanes, image scores, acoustic imprints, and Spatial Score automation JSON.

The Package Browser discovers user-facing tools and controllers from their script metadata and groups them by workflow. The inventory below follows those same groups. Helper libraries and support dialogs remain hidden from the browser, while included JSFX engines and effects live under Effects/s3g.

Utils

Automation Score: opens the browser utility for composing generic breakpoint lanes and section markers that can later be assigned to track or FX automation. An optional Max bridge in Scripts/s3g-mc/utilities/automation-score-max-bridge can play the same JSON at control rate in Max.
Displacement Score: opens the browser utility for designing 24-point 3OAFX virtual speaker displacement JSON. An optional Jitter bridge in Scripts/s3g-mc/utilities/displacement-score-jitter-bridge can display the exported geometry in Max with points, polygon lines, and a projected heatmap.
Image Score: opens the browser utility for composing PNG scores used by 3OAFX Image Sonogram Field.
Imprint Sketch: designs seeded directional responses and geometry-derived echo paths for connected 3D regions, including wall, overhead, and beneath-listener branches with shaped portals. It exports compatible JSON for 3OAFX Synthetic Ambisonic IR Bank and .s3gimprint files for s3g Ambi Imprint 64 in s3g-dsp.
Ray Sketch: designs musical moving-source fields from the same connected geometry, materials, portals, and echo paths. It exports .s3gray maps for s3g Ambi Ray Encoder in s3g-dsp.
Load Automation Score JSON: opens an ImGui target-assignment window for mapping Automation Score lanes to selected track volume envelopes or named FX parameter envelopes, with append or replace write modes and optional project marker writing.
Load Spatial Score JSON: loads exported Spatial Score JSON into REAPER, writing AED automation to a focused/touched FX with Azimuth, Elevation, and Distance parameters, falling back to X/Y/Z when AED is unavailable, or alternatively creating the 3OA encoder bus/child tracks. It can be used with plugins such as s3g Ambi Point Encoder, SPARTA AmbiENC, IEM MultiEncoder, and ICST AmbiEncoder, depending on exposed parameters.
Spatial Score: opens the browser utility for composing spatial motion JSON outside REAPER.
s3g-mc Package Browser: browses and launches the installed package scripts.

Channel Mixing / Automation

128ch Automation Mixer: faders, mute/solo, channel groups, meters, and plugin pin remapping for high-channel-count tracks.
128ch Ambisonic Node Track Mixer: selected ambisonic source tracks become movable nodes, with whole ACN/SN3D streams weighted and summed channel-for-channel without decode, encode, or speaker remapping.
128ch Node Track Mixer: selected source tracks become movable channel-shape nodes, so stereo, ring, cube, double-ring, or dome source layouts can be mixed as spatial objects or aligned stacked shapes. The controller includes track-name node labels, routing overview, top/side camera views, cursor falloff/gate controls, and automation lane helpers.
Focused FX Automation Capture: locks the currently focused or touched track FX, filters/selects parameters, and writes the current plugin settings as editable automation points at the edit cursor.
MC to Stereo Autogain: multichannel fold-down with layout modes, width, rotation, weighting, 3D projection attenuation, autogain, and output trim. The JSFX declares only stereo outputs; in multichannel listening chains, set its plugin I/O to zero unmapped output channels.
Patch Routing View: ReaImGui bottom-up patch-canvas view of REAPER track routing, with draggable track nodes, folder/send/master patch cords, channel counts, FX state, wire selection, send editing, per-folder collapse, minimap navigation, and a node inspector.
Snapshot Surface: captures track and FX states as named Voronoi regions, stores them in the REAPER project, generates randomized variants from the captured target set, and interpolates between snapshots with a manual cursor or automatable s3g Snapshot Surface Cursor JSFX.
Transaural Crosstalk Canceller: stereo loudspeaker playback processor with feedforward and matrix-inverse crosstalk cancellation modes, speaker geometry, low protection, stereo preservation, and safety gain controls.
Vertical Timeline Navigator: ReaImGui companion view that draws time top-to-bottom and tracks left-to-right for navigating items, markers, regions, edit cursor, linked transport, minimap overview, time selection, item moves with optional snapping, marker/region add-edit, all/selected track filtering, muted REAPER-derived item colors, channel subdivisions, mouse scroll, and mouse zoom.

MIDI Composition

Form Learner: NumPy-backed composer that analyzes selected MIDI items and expands their rhythm, pitch, velocity, duration, and channel traits into a longer editable MIDI form.
Lattice Drums: creates an editable drum MIDI item from a layered table-scanning gesture, with each layer assigned to a drum voice and cells shaping velocity, duration, and density with Superior-style or GM note maps.
Lattice Tables: creates an editable MIDI item from a table-scanning gesture model, with scale selection, voicing, duration variation, table or single-channel output, and visible translation layers.
Musical Space: creates an editable MIDI item from a path through melodic and harmonic spaces, with expanded scale choices, selectable rhythm models, density, voicing, note-length variation, MIDI-channel focus, and geometric preview.
Polymetric Drum States: creates an editable drum MIDI item from changing polymeter configurations, using Superior-style or GM drum maps, integer or fractional state lengths, beat-grid snapping, and jump or glide transitions between states.
Polymetric Pitch Lanes: creates multiple Euclidean lanes with preset banks, independent lengths, pulses, rotations, pitch degrees, note-length variation, and lane or single-channel MIDI output.
Spectral Trace MIDI: NumPy-backed audio-to-MIDI analyzer that traces spectral peaks, centroid motion, or partial stacks from selected WAV items, including multichannel sources up to 16 channels, into editable MIDI with scale quantization and audio-channel-to-MIDI-channel mapping.
Terrain Form: NumPy-backed song-duration MIDI generator with section maps, terrain-shaped density/register/channel fields, expanded scales, and optional project markers.

Procedural Synthesis

These tools use included JSFX synth engines in two ways: render actions create new media items offline, while MIDI controllers configure the same engines for timeline-driven realtime use.

Carto Synth Render: renders multichannel dust, pulse-packet, logic/fractal-drone, byte-mask, and spline-drift materials.
Carto Synth MIDI Controller: loads the Carto engine on the selected track and maps MIDI notes, velocity, gate, and MIDI channel focus to synth behavior.
Lattice Synth MIDI Controller: loads a table-driven resonator synth on the selected track, mapping generated MIDI to gesture paths, excitation, resonance, divider shadow, and MIDI-channel focus.
Lattice Synth Render: creates a temporary Lattice MIDI score and renders the Lattice Synth engine to a multichannel media item.
Spectra Synth Render: renders slower spectral-mass and resonant materials: partial clouds, comb strata, formant bands, impulse resonators, and noise spectra.
Spectra Synth MIDI Controller: loads the Spectra engine on the selected track with the same MIDI response layer.

The render actions use breakpoint curves for parameter motion and channel behavior, then write a new multichannel media item.

Offline Synthesis / IR

These NumPy-backed renderers handle processes that are easier to do offline with Python than in Lua or JSFX. Breakpoint envelopes shape the render over time. Density means event or peak admission before synthesis, not gain modulation afterward.

Dense Grain Cloud: source-item grains scattered into a multichannel field.
EVP Field: eSpeak-backed speech/formant field for EVP, chant, choir, possession, and damaged-broadcast textures. Voice treatments reshape the spoken source, time expansion can fill the requested duration, and an optional spectral shaper can apply preset masks, typed tone-list resonances, or a selected-item profile imprint. Optional Speex codec damage can be added before fixed 3OA ACN/SN3D or multichannel ring, double-ring, cube, and dome output. eSpeak NG is required for this process; Speex is only required when Codec damage is enabled.
Fata Morgana Resynth: hybrid oscillator resynthesis from 2-16 selected source items, recombining timing, pitch, amplitude, and spatial traits.
IR Toolkit: reshapes a selected impulse response item with silence trim, tail fade, normalization, early reflections, and channel decorrelation.
Karplus Field: Karplus-Strong plucked resonator events distributed across a multichannel bed.
Mass Partial Field: additive partial events with drift and channel motion.
Modal Terrain: source-free modal/resonator-bank synthesis with struck, dust, swell, and strata excitation models, rendered as 3OA ACN/SN3D or an even-channel ring.
Partial Trace Resynth: STFT peak tracing rendered as a multichannel oscillator field, with linked, point, smear, and frozen trace modes.
Resonant Terrain: struck resonator banks for metallic, synthetic-IR, and resonant-dust materials.
Subharmonic Bank: subharmonic divider-bank synthesis with masks, instability, pulse/sine blend, folded articulation, and spatial spread.

Spatial Panners

12ch Dodeca Panner: AED-native panning for up to 8 mono sources across a 12-channel dodecahedron loudspeaker layout, with a dodecahedron controller view.
17ch Cube XYZ Panner: 3D DBAP-style Cartesian amplitude panning for up to 8 mono sources across a 17-speaker cube layout, with native XYZ controls and a mirrored AED editing view. Small spread values can isolate a source to one speaker.
Layout Panner: a general-purpose panner for quad, octophonic ring, 8ch cube, 12ch ring, 16ch ring, 16ch double ring, 20ch double ring, and 24ch dome without overhead. Speaker numbering starts near the stereo-right position and proceeds clockwise.
25ch Cosine Dome Panner: soft angular-focus panning for up to 8 mono sources across the 25-speaker dome.
25ch DBAP Dome Panner: distance-weighted amplitude panning for up to 8 mono sources across the 25-speaker dome.
25ch LBAP Dome Panner: layer-based amplitude panning for up to 8 mono sources across the 25-speaker dome.
25ch Region Dome Panner: region-locked panning that constrains up to 8 mono sources to speaker-defined rings, arcs, ribs, triangles, and caps.
25ch Vector Morph Dome Panner: four-scene spatial snapshot morphing for up to 8 mono sources, driven by two automatable controls.
25ch VBAP Dome Panner: nearest-region vector-style panning for up to 8 mono sources across the 25-speaker dome.
Spatial Automation Composer: previews path, graph, and topology-informed AED or XYZ motion with source relationship modes, then writes editable automation lanes for supported 8-source s3g panners.

The included panners are intended for use with loudspeaker arrays available in the RISD Studio for Research in Sound & Technology (SRST). They include a 12-channel dodecahedron layout, a 17-channel cube layout, and the shared 25-channel dome layout. The 25ch dome panners offer several approaches to the same array, giving composers room to compare and work with different spatial behaviors.

3OAFX

Ambisonic Stereo Decoder: package-native JSFX stereo decoder for ACN/SN3D 1OA, 2OA, or 3OA. It decodes to an internal virtual speaker field, then derives a practical loudspeaker stereo fold-down using stereo pickup models such as XY, MS, Blumlein, ORTF-style, and spaced omni.
6ch Ambisonic Decoder Router: package-native JSFX monitor decoder/router for ACN/SN3D 1OA, 2OA, or 3OA into a compact 4-speaker bed plus 2 elevated side speakers, with direct 6-channel routing mode.
3OAFX AED Granulator: NumPy-backed object-grain renderer that takes mono, stereo, or multichannel WAV-backed media, reads grains from source channels, assigns generated AED positions, and encodes directly to 3OA ACN/SN3D output.
3OAFX Ambiance Extractor: extracts room tone, noise bed, or other profile-like material from an ambisonic source while preserving the directional decode/re-encode workflow.
3OAFX Ambisonic Kernel Collage: creative ambisonic cross-convolution that treats selected ambisonic recordings as spatial/spectral kernels for another ambisonic source, with mixed-order kernel adaptation.
3OAFX Image Sonogram Field: PNG image-score synthesis where X is time, Y is frequency, color data determines AED positioning, and amplitude comes from alpha, edge contrast, or a separate image. Includes additive, filtered-noise, hybrid, and selected-source granular engines, plus transposed image reading, for 3OA ACN/SN3D or multichannel ring output.
3OAFX Offline Renderer: NumPy-backed ambisonic decode/process/re-encode action for 1OA, 2OA, and 3OA media, with moving AED focus, dry attenuation, and included regional effects such as gain, filtering, delay, diffusion, resonator, saturation, ring modulation, tremolo, pitch shift, and spectral smear.
3OAFX Object Space: NumPy-backed object-to-space render for ambisonic or non-ambisonic source material, with resonance bloom, spatial occupation, motion counterpoint, and spatial allusion modes.
3OAFX Object / Field Split: automatic NumPy-backed split of an ambisonic source into object-like foreground and field-like spatial bed outputs.
3OAFX Particle Cloud: NumPy-backed ambisonic particle-cloud renderer for selected ambisonic or non-ambisonic WAV-backed media, with source-pool, density, asynchronicity, intermittency, scan, grain envelope, playback, and HOA yaw controls, plus a source-to-cloud diagram and time breakpoints.
3OAFX Pulsar Field: NumPy-backed ambisonic pulsar-train synthesizer with fundamental/formant curves, pulse masking, pulsaret shape, envelope shape, AED trajectory controls, a pulse-field diagram, and time breakpoints.
3OAFX Scene Navigator: multi-source ambisonic scene surface where selected soundfield files become nodes and an editable listener path renders a new perspective traversal.
3OAFX Spatial Occupation Montage: event montage from one or more source items into ACN/SN3D ambisonic space, with source-format inference and stereo sum/difference expansion for non-ambisonic stereo sources.
Stereo Expand to Ambisonic Bed: NumPy-backed mono/stereo source expansion to 1OA, 2OA, or 3OA ACN/SN3D, deriving front, side, rear, and optional height cues from left/right and mid/side material.
3OAFX Offline Ambisonic Convolve: NumPy-backed ambisonic convolution. It can convolve one ambisonic source with one same-order ambisonic IR, or use direction-accurate IR banks for P-format first order and 8-direction 2OA/3OA, with optional lower-order and sparse FOA IR adaptation.
3OAFX Send Return Controller: places a 24-channel insert lane inside an ambisonic decode/encode chain.
3OAFX Spatial Grains: spatial-grain renderer for 1OA, 2OA, and 3OA media. Grain position, envelope, duration, rate, and overlap are shared across all encoded channels, with source-time navigation and optional HOA yaw/order transforms.
3OAFX Spatial Freeze Trace: NumPy-backed ambisonic spectral freeze and trace renderer. A selected spectral imprint or moving trace is applied coherently across encoded channels, with optional HOA yaw drift and order weighting.
3OAFX Spectral Hole Maker: carves profile-shaped spectral space from an ambisonic source per direction.
3OAFX Spectral Profile Match: steers the source spectrum toward a reference profile per direction without mixing the reference into the output.
3OAFX Spectral Profile Subtract: ambisonic noise/material-profile subtraction. Select a source and a profile item; both are decoded to the same directional layer before cleaned or residue output is re-encoded.
3OAFX Spectral Residue Extractor: renders the removed spectral material as its own ambisonic item for creative or diagnostic use.
3OAFX Synthetic Ambisonic IR Bank: designs encoded ambisonic IR banks for 3OAFX Offline Ambisonic Convolve, with room size, material absorption, scattering, early reflections, late-field controls, optional Imprint Sketch JSON import, and separate or stacked output formats. Higher-order banks use 8 directions: stacked 2OA is 72 channels and stacked 3OA is 128 channels.

Spectral / Convolution

Convolve selected items: convolution of two selected media items, including mono, stereo, multichannel pairing, and summed matrix modes.
Cross Synthesis: offline STFT cross-synthesis for two WAV-backed media items. The first selected item keeps phase and timing while its spectral magnitudes are blended toward the second item’s magnitudes.
Chaotic Resonant EQ: protected multichannel resonant filter-bank process with chaotic detuning, drive, and channel-to-channel feedback.
Render MC Impulse Field: procedural multichannel impulse fields for convolution.
Spectral Accumulate: spectral sustain where each frequency band holds until stronger energy replaces it.
Spectral Ambiance Extractor: extracts source material that resembles a room-tone or noise-bed profile while preserving the source channel layout.
Spectral Blur: offline magnitude blur across neighboring STFT frames, with safe envelope mode and optional time expansion.
Spectral Freeze: imposes one selected spectral frame across the item while preserving phase/timing motion, with safe envelope mode, envelope floor, and optional time expansion.
Spectral Hole Maker: carves profile-shaped spectral space from a source item without ambisonic decoding.
Spectral Morph: live or frozen spectral morph between two WAV-backed media items.
Spectral Profile Subtract: removes a learned spectral profile directly across ordinary mono, stereo, or multichannel media while preserving the source channel count.
Spectral Residue Extractor: renders the material removed by a profile subtraction as its own multichannel item.
Spectral Shaper: offline spectral envelope transfer for two WAV-backed media items, with an alternate formant-vocode algorithm. The first selected item is the carrier/tune/timing source; the second supplies the spectral envelope or broad formant contour.
Spectral Spatializer: distributes frequency bins across even output channel counts from 2 to 64.
Spectral Step Drunk Freeze: stepped freeze or random-walk freeze through spectral frames.
Spectral Trace: partial tracing with modes to keep loudest partials, suppress loudest partials, threshold, or thin randomly.

Multichannel Texture / Montage

Brownian Walk: short fragments follow a bounded random walk through source time and output channels.
Cascade spatial echo: equal segments print decaying echoes through space.
Channel orbit delay: whole-item delay repeats orbit around output channels.
Channel Smear: slices duplicate to neighboring channels with gain compensation.
Crumble spatial groups: slices move through progressively smaller channel groups.
Flutter Gate: moving active-channel groups create flutter patterns.
Fracture: ordered slices from one source channel disperse across a channel path, with jitter, drop, and spread voices.
Frame Gate: rotating active-channel groups print gate patterns.
Frame Shift: channel-frame rotation, mirror, odd/even split, pair interleave, or half-swap render.
Loop Drift: NumPy-backed seamless source loops distributed across a multichannel bed, with overlap-add seam crossfades and per-channel playback-rate spread, source-pool routing, quantization, direction, gain, and motion variation.
Loop Rift: NumPy-backed loop sections with graceful dropouts, overlap-add fades, source-pool routing, grouped channel openings, and unstable playback rates.
Marker spatial montage: project markers or active-take markers inside the selected item define chunks for ordered or shuffled montage.
Mono Fill: one source channel fills every output channel with optional gain compensation and slice rotation.
Scatter Slices: multiple selected items are sliced by equal divisions, project markers, or active-take markers, then randomly arranged across a target multichannel duration with scatter, ordered-walk, stutter, repeater, channel-smear, channel-motion, shape, and breakpoint-density variations.
Shred / Slice: equal, project-marker, or active-take-marker slices with ordered mono spread, random mono scatter, and multichannel reorientation modes.
Spatial Repeater: one source channel repeats around clockwise, ping-pong, or random channel paths.
Spatial Stutter: repeated short slices advance through a spatial path.
Stereo Spin: sliced stereo images rotate around a multichannel output field.
Texture Clouds: dense short fragments from one source channel scattered across an output field.
Zigzag channel walker: equal slices walk back and forth across output channels, with optional reverse source-slice order.

Track Building / Routing

Build multichannel stem from selected tracks: routes selected tracks to consecutive channels on a new multichannel destination, then renders a bounded stem.
Cycle mono tracks into multichannel stem: selected mono tracks become a multichannel stem, with repeat or grouped downmix behavior when the requested output count differs from the source count.
Route selected tracks to multichannel folder bus: creates a new parent folder bus and assigns each child track to consecutive bus channels.