Ambi Encoder Ray

s3g Ambi Encoder Ray places a mono source and listener inside a sampled room field and produces a 64-channel ACN/SN3D output bus. Motion of either point changes the direct path and early reflection directions, delays, and Doppler behavior; source motion also interpolates the sampled late-room profile. See the s3g-mc Ray Sketch guide for the field-authoring model.

Workflow

1. Insert the plugin on a 64-channel REAPER track, select the ambisonic ORDER, and route a mono source to its input.
2. Follow it with an ambisonic decoder.
3. Open the s3g-mc Ray Sketch browser utility and design, generate, or import a space.
4. Place the listener and default source, then choose the source-sampling grid and reflection-slot count.
5. Export a .s3gray Ray Field.
6. Load the field, move the source and listener in the top and side views, and balance direct, early, and late sound.

Ray Field

The built-in room makes the plugin immediately usable. LOAD accepts version 1 .s3gray files with the format identifier s3g-ambi-ray-field. A field stores room and navigation bounds, a reference listener and default source position, and 1 to 256 sampled source cells. Each cell can hold as many as 32 directional early reflections plus a late response profile. See the .s3gray format reference for the complete runtime contract.

At runtime, the encoder blends the four nearest cells as the source moves. Loaded field data is stored with the REAPER project state, so the original .s3gray file does not need to remain beside the project for recall.

Ray Atlas

ATLAS opens nineteen deterministic spaces generated through the Ray Sketch engine. The collection spans architecture, caves, caverns, tunnels, canyons, clearings, focused echo designs, and abstract connected spaces. Each entry is a complete sampled .s3gray field with stable reflection slots, listener-resolvable bounce points, and a directional late profile.

Choosing an atlas entry loads its map and authored source and reference-listener positions while preserving the order, direct/early/late balance, room shaping, Field Listen mode, movement smoothing, Doppler amount, bypass, and output gain. Atlas data is embedded into project state exactly like a user-loaded field.

Musical Interpretation

Ray Sketch resolves the designed space into a compact map of stable reflection slots and late-room profiles. The plugin owns both live positions. Its direct path is calculated continuously between source and listener, nearby source cells reshape the early and late field, and bounce-aware early paths are recalculated through their wall, floor, ceiling, portal, or echo point as either endpoint moves. Moving either point therefore does not require rebuilding or streaming a new ray trace.

The result is intentionally a musical room instrument rather than an architectural simulator. Material curves, path losses, and late estimates preserve a space's directional character and repeatable motion behavior without claiming measurement-grade acoustics.

Position and Motion

SOURCE X/Y/Z and LISTENER X/Y/Z are normalized automation parameters mapped to the loaded field's navigation bounds. The GUI reports both positions in metres. MOVEMENT applies the same trajectory smoothing time to both points.

DOPPLER controls pitch motion caused by changing path lengths. At 0%, delay heads hold while a point is moving and retarget with a short crossfade after the path settles. 50% is the musical default, 100% follows propagation timing at the speed of sound, and values through 200% exaggerate the shift. Source and listener movement both affect the direct path, and each early reflection receives its own shift from its changing source-to-bounce-to-listener distance.

The world view uses X right, Y front, and Z up. At the ambisonic boundary, world +X maps to -90 degrees on the right, world -X maps to +90 degrees on the left, world +Y maps to 0 degrees in front, and world +Z maps upward. This matches the ACN/SN3D convention used by the other s3g encoders and decoders.

The display places a top view and side view next to each other. Select SRC or LIS in the Position panel, or grab either marker directly, then drag in the top view for X/Y or the side view for X/Z. Solid green lines run from source through an exported amber bounce point to the listener. Dashed cyan events without a defensible bounce point use a stable world-space arrival anchor derived from the field's reference listener. The grey source-to-listener line is the direct path. Position automation is continuous, and loading a new field changes rooms without interrupting the audio path.

The late field remains a directional room bed interpolated from source cells. Its excitation pre-delay retargets without continuous pitch modulation, so moving the source or listener does not bend the entire reverb tail. Listener motion reshapes direct and early geometry but does not claim a second sampled late-field dimension.

Field Listen

FIELD LISTEN gives the room response a slow spatial memory. Eight virtual directional microphones analyze only the early and late field after Width; the direct source is excluded. Their envelopes apply bounded, energy-normalized gain changes to the directional early reflections and late lines. This is control feedback rather than an audio feedback loop, so samples are never recirculated and the authored ray geometry, delay, Doppler, decay, and material damping remain unchanged.

• OFF leaves the authored room response fixed.
• FOLLOW reinforces recently dominant reflection corridors and late-field directions, allowing the room's directional character to persist.
• COUNTER transfers that reinforcement toward opposing directions.
• BALANCE restrains dominant directions and favors quieter parts of the room field.

The response is most apparent with transient or broadband sources, moving endpoints, and asymmetric spaces such as tunnels, portals, caves, and focused echo networks. Symmetric rooms and sustained material naturally produce a subtler result. Mode changes are smoothed, and OFF is the default for an unmodified authored room response.

Room Controls

• DIRECT sets the distance-delayed source component.
• EARLY sets the interpolated directional reflection field.
• LATE sets the diffuse late response.
• SIZE scales propagation and room timing.
• SCATTER increases reflection variation and diffusion.
• WIDTH scales the directional spread of the room response.
• AIR adds distance-dependent high-frequency damping.
• FIELD LISTEN selects Off, Follow, Counter, or Balance room-memory behavior.
• BYPASS ROOM removes early and late energy while retaining direct spatial encoding.

Order and Routing

The direct source can use first through seventh order. The room response is represented through third order, and channels above the selected order are cleared.