A first scene takes four operators: a shape, a Camera, a Scene that collects them, and a renderer that draws the Scene. Materials and Lights come after. Place each one from the Geometry view of the Voro menu, under Render.

Build it

1. Geometry. Place a Geometry. It is a TouchDesigner Geometry COMP: put a polygonal SOP or POP inside and turn its render flag on. Geometry COMPs already in your network can join the scene too. For many copies of one shape use Place. For text or an SVG use Font, and for a model file use Asset Import.
2. Camera. Place a Camera. It is a TouchDesigner Camera COMP with a Voro page, so you pose it the TD way. A plain TD Camera COMP works too.
3. Scene. Place a Scene. On its Collection page, set Geometry to the Geometry COMP’s name and Camera to the Camera. Geometry and Lights take a name or a wildcard pattern.
4. Renderer. Place Realtime Trace to watch the scene move, or Path Trace to let a still view converge. On its Scene page, set Scene to your Scene and Scene Mode to Scene.

Now pulse Frame Scene on the Camera’s Voro page. The Camera pulls back until the object fits in frame, then stays an ordinary TD camera you can move.

A scene needs a camera. Nothing falls back to a default one, so a renderer with no Camera set, on the Scene or on the renderer itself, reports “no effective Camera selected” and draws nothing.

Look and light

• Material. Place a Material and select it on the Geometry’s Look page. An object with no Material draws in plain grey. See Material authoring to write your own look.
• Sky. Place a Sky and select it on Scene’s Collection page for an atmosphere, clouds and sun, or an HDR environment. Without a selected Sky, the renderer’s Environment page supplies the fallback sky and sun.
• Light. Place a Light and add its name to Lights on the Scene for more lighting. Plain TD Light COMPs work too.

What a renderer gives you

Realtime Trace and Path Trace publish the same ten images. beauty is the picture. The other nine are guides for compositing and denoising, such as depth, normals, albedo and motion. beauty passes through the renderer’s Display Transform, ACES by default. The guides stay linear float. The Path Trace page lists all ten.

Two renders in one picture

Depth Composite takes two renders, each as colour and depth, and keeps the nearer one per pixel. Partial foreground coverage blends as premultiplied alpha. Render both from the same camera at the same size. A mismatch is reported, not reprojected.

Adding a solver

A solver such as FLIP Liquid or Gas publishes points and fields. Its reference page lists what it takes and what it produces. A Force Field pushes on it and an SDF Collider gives it something to hit. Warp FX is where you write a solver or a deformer of your own.