Listener Mode

Listener Mode lets a plugin analyze its own ambisonic field and use what it hears to influence what happens next. It turns spatial output into an internal score: direction, activity, change, and, where supported, spectral and temporal character can guide later movement, synthesis, or field shaping.

This is a shared design idea rather than one identical circuit copied into every plugin. A room effect may rebalance reflections, a procedural instrument may choose its next region, and a recurrent instrument may return decoded audio to its generating network. The individual plugin page remains the authority for its exact controls and tap point.

Signal Flow

The plugin forms an internal ACN/SN3D ambisonic field, usually before final OUT trim.
Four tetrahedral or eight cube-corner virtual pickups sample that field in several directions. Some specialized listeners align pickups to a loaded profile or let their directions adapt.
Each pickup measures recent signal or energy. Extended listeners also derive features such as fast and slow energy, novelty, roughness, spectral balance, habituation, charge, and directional gradients.
A listening mode decides which directions or pickup relationships matter.
A response law maps the analysis back to bounded synthesis, movement, register, reflection, or field-shaping controls. Plugins with an explicit FIELD RETURN may also run a separate, conditioned audio-return path.

The virtual pickups decode the HOA field mathematically inside the plugin. They do not require a loudspeaker decoder, an extra REAPER track, or monitoring feedback through the host.

The Auditory Body

The common listener uses a symmetric directional body. TETRA 4 provides four broad viewpoints, CUBE 8 adds their opposites, ICOSA 12 provides twelve evenly distributed vertices, DODECA 20 is the twenty-vertex dual of the icosahedron, and SPHERE 24 extends the body for high-order work. Plugins with fixed four- or eight-pickup listeners retain their authored bodies. In AUTO, order-adaptive Ambi Effects use twelve points for 1OA-2OA, twenty for 3OA, and twenty-four for 4OA-7OA. Processor Ambi Imprint instead follows the directions stored in its response profile, while Ambi Encoder Neural Ecology can steer each pickup within a bounded region around its tetrahedral or cubical anchor.

These pickups hear spatial relationships that a mono sum would discard. A source moving between directions changes which pickups become active; width, rotation, environment return, reflections, and ambisonic order may also change the field presented to the listener. Pickup meters are analysis displays, not additional audio outputs.

Shared Control Vocabulary

Control family	General meaning
LISTEN / FIELD LISTEN	Enables the listener or selects its directional law. `OFF` preserves the plugin's open-loop behavior.
FOLLOW	Favors recently active or strongly heard regions.
COUNTER	Applies the heard relationship toward the opposing direction.
BALANCE	Favors underexposed directions or restrains dominant ones.
LOCAL / TRACE	Associates a voice, lobe, or event with its nearest pickup.
CROSS	Reads an opposing or cross-connected pickup.
DIFFUSE	Uses a stable mixture of the complete pickup body.
ROAMING / RING	Moves a read head through the pickup relationships over time.
RESPONSE	Chooses what the heard features do. Names such as `SETTLE`, `RESONATE`, `IMPRINT`, `ENTRAIN`, and `WRITE` are plugin-specific.
AMOUNT / ECOLOGY / CAPTURE / VISCOSITY	Scales the depth of the listener's control influence. The label reflects the synthesis model.
FIELD RETURN	Mixes a directional audio return into a generating circuit when that plugin supports one. This is distinct from control feedback.
PROPAGATION	Delays and, where documented, filters an audio return to give it temporal distance.
RETURN BYPASS	Disables only the audio-return path. Analysis, meters, and control responses may remain active.

Not every listener exposes every family. In particular, FOLLOW/COUNTER/BALANCE describe a directional preference, while LOCAL/CROSS/DIFFUSE/ROAMING describe how several voices or neural lobes read an auditory body. They solve related but different routing problems.

Response Laws

The response name describes how a plugin interprets the same broad class of evidence:

Reinforcing responses increase attraction, event activity, resonance, or motion in heard directions. Names include FOLLOW, CHARGE, CHORUS, and RESONATE.
Balancing responses seek quieter regions or reduce activity where the field remains dominant. Names include BALANCE, SETTLE, and capture or viscosity controls.
Imprinting responses transfer features of the heard field into newly generated material. IMPRINT, ENTRAIN, and Wrangler's WRITE belong to this family, although each affects very different state.
Legacy or classic responses provide the baseline listener behavior, making them useful references when comparing the alternate response laws.

Response names are not interchangeable presets. For example, Water's IMPRINT affects event material and size, Pulsar's IMPRINT latches features into a new grain, and Wrangler's WRITE can influence a shift-register bit.

Control Feedback and Audio Return

Most s3g Listener Modes are control feedback. The plugin measures its field, then changes bounded control values used to create the next field. Audio samples are not recirculated. Ambi Ray, for example, changes directional reflection gains without changing the authored room geometry, and Wave Terrain changes traversal decisions without feeding decoded audio into its oscillator.

Ambi Encoder Neural Ecology, Ambi Pulsar, and Wrangler can also use an explicit audio return. Those paths are delayed, normalized, filtered, saturated, or otherwise bounded according to the generating circuit. A return bypass therefore does not necessarily turn Listener Mode off; it can leave the auditory analysis and control response running.

Final OUT trim is kept outside the listener loop where a plugin documents pre-output listening. This allows final audition level to change without retuning the ecology. Other exclusions are deliberate: Ray listens to its early and late room response but excludes the direct source, while Bilocation keeps separate listeners for its A and B spaces.

Workflow

Build or recall a stable patch with listening OFF, and use that as the open-loop reference.
Follow it with an ambisonic decoder.
Enable the listener and choose the simplest directional mode that expresses the goal: reinforce with FOLLOW or LOCAL, oppose with COUNTER or CROSS, or redistribute with BALANCE.
Raise the listener amount, ecology, capture, or viscosity gradually. The middle of the range usually makes the relationship easier to hear than an immediate maximum setting.
Where available, compare response laws without changing the underlying patch. Watch the pickup and response meters to distinguish what the field hears from what the synthesis receives.
Add FIELD RETURN only when audio recirculation is part of the intended instrument. Use RETURN BYPASS to compare that deeper loop while retaining the listener's control behavior.

Mode and amount changes are smoothed in the listeners designed for live adjustment. Saved plugin state includes the relevant listener controls; consult the plugin page for its portable preset behavior.

Plugins Using the Concept

Plugin	What listens	What changes next
Ambi Effect DJ Filter	Incoming encoded field at an upscaled Icosa12 or Dodeca20 body	Individually filtered, topology-routed, and directionally masked residual re-encoded into that field
Ambi Effect Delay	Incoming encoded field at the same order-adaptive body	Per-pickup delayed and feedback-shaped correction, with macro relationships, topology, and directional dry/wet masking
Ambi Effect Pitch	Per-pickup field energy and interval/window relationships	Dual-grain pitch correction routed and masked inside the encoded field
Ambi Effect Gain	Per-pickup field energy and effective gain relationships	Directional attenuation or boost, topology routing, and optional complementary dry suppression
Ambi Effect Resonance Print	Short overlapping spectral observations at each order-adaptive pickup	A saved frequency/amplitude/stability print that the continuing field excites as a topology-routed modal body
Ambi Effect Partial Trace	Global spectral peaks reconciled with direction evidence from the order-adaptive pickups	A live or frozen sinusoidal trace with temporal smear, returned through Local, Cross, Diffuse, or Roaming topology
Ambi Effect Response Trace	Incoming pickup signals captured and replayed through one directional convolution kernel per pickup	A latency-aligned wet response returned through Local, Cross, Diffuse, or Roaming topology
Processor Ambi Imprint	Re-encoded wet response	Directional response-profile gains
Ambi Encoder Ray and Ambi Encoder Ray Bilocation	Early and late room fields	Bounded reflection and late-line balance
Ambi Encoder Wave Terrain	Generated pre-output field	Region choice, scan orientation, activity, and traversal weighting
Ambi Encoder Water, Ambi Encoder Wind, and Ambi Encoder Insect	Generated field plus encoded environment	Directional motion and response-specific event or material behavior
Ambi Encoder Stochastic	Generated pre-output field	Voice-local negative-feedback capture and evolution rates
Ambi Encoder Modal	Internal modal HOA field at a fixed selectable Tetra 4 or Cube 8 listener	Direction-aware continuous actuation across four to eight independent resonant bodies
Ambi Encoder Wrangler	Completed generated field	Register writing or settling, with optional delayed audio return
Ambi Encoder Pulsar	Post-depth point field	Event and grain response ecology, with optional neural return
Ambi Encoder Neural Ecology	Live recurrent field	Neural equations, adaptive pickup directions, motion, and lattice decisions

What Listener Mode Is Not

Ray listener position: LISTENER X/Y/Z is a geometric point used to calculate source distance, direction, delay, and reflections. FIELD LISTEN is the room's internal directional memory. Ray can use both at once.
A decoder listening mode: Ambi Decoder Head and Ambi Decoder Stereo render an existing field for audition. They do not make the upstream instrument respond to its own output.
DAW or acoustic feedback: the listener is contained inside the plugin. It does not listen to a microphone, the master output, or sound returning through the room.
A universal modulation mapping: plugins share an auditory idea and vocabulary, but each maps the evidence into controls appropriate to its own synthesis or processing model.