Nodes
Every node R3NE ships with, organized by category. The graph is type-checked: ports of compatible types snap together, the rest don't.
Every node has the same anatomy: a set of typed input ports on the left, a set of output ports on the right, and (where it makes sense) an inspector with sliders, color pickers and file pickers. Wiring a value into an input port replaces its inspector value live.
Pick a category to dive in:
Constant values, time, color, and the FFGL host bridge. The smallest building blocks, every reactive graph starts here.
Arithmetic, range remapping, conditional selection and combiners. Wire these between sources and parameters to drive everything.
Procedural texture sources, Perlin and fractal noise, Voronoi, plasma, gradients, grids, stars, Truchet tiles, Hilbert curves and more.
Picture / video / live camera input, Apple spatial video, primitive shapes, transform, render-to-texture and projection mapping.
Audio file playback and live capture with FFT / band / volume outputs, a sound-driven modifier, a 7-band splitter and onset / beat detection.
Per-pixel effects: blur, glow, color correction, displacement, kaleidoscope, glitch, lens flare, grain, scan lines, halftone, and many more. Chain freely.
Channel and UV manipulation: shuffle, sampler, distance fields, normal map generation, MatCap, recolour, gradient remap.
Single-float transforms: range remap, sine wave, smoothing, accumulator, quantise, delay, limiter, curve remap, trigger envelopes.
The Scene3D composer node and how its post-process pipeline (bloom, SSAO, fog, DoF, exposure) works.
Cube, sphere, plane, torus, cylinder and cone primitives, mesh import (OBJ / USDZ / PLY), 3D text, Gaussian splat clouds and the depth point cloud.
Directional, point and spot lights, plus a perspective / orthographic camera node.
Live MSL PBR Material, six baked presets (Chrome, Gold, Steel, Glass, Velvet, Neon) and an HDRI environment map.
Vertex modifiers (Noise Displace, Twist, Wave), Instancer cloning, GPU particle emitter and audio analyser.
Apple Vision / CoreML analysis: monocular depth estimation, face detection, body pose tracking, optical flow and depth keying — overlay textures plus float outputs.
SVG vector graphics: import, effector (translate / rotate / jitter / wave), repeat, random color, trim-on, rasterize to texture and extrude to a 3D mesh.
Send the final composite to other apps via NDI 5 or Syphon. The Root node already composites; these expose it.
Type system
Connections in R3NE are typed. A few cross-type conversions are allowed , most aren't.
| Type | Compatible with | Used by |
|---|---|---|
float | float | parameters, time, math |
color / float4 | interchangeable with each other | tints, RGBA, vec4 |
texture | texture | shader output, picture, video |
geometry | geometry | 3D meshes, modifier outputs |
light | light | directional, point, spot lights |
camera | camera | perspective / orthographic camera |
material | material | PBR Material, baked presets |
particles | particles | GPU particle emitter |
cycleDetected error rather than silently looping forever.