WonderTexture Engine

Understand the relationship between Blender, the WonderTexture add-on, and the native GPU-powered processing engine.

// ARCHITECTURE

Blender and the Engine Work Together

WonderTexture is split into two parts that work together:

Blender Add-on

Provides the node editor integration, graph authoring tools, panels, asset management, Blender images, materials, and the user-facing workflow.

WonderTexture Engine

Runs the native texture-processing runtime that evaluates WonderTexture graphs and produces texture results.

You author the graph in Blender; the native engine does the processing. As you edit, Blender keeps the project data while WonderTexture sends the relevant graph state to the engine for evaluation.

Runtime State Can Be Rebuilt

Synchronization and preview state on the engine side belong to the current runtime session. After a restart or reconnect, WonderTexture clears that session-specific state and synchronizes graphs again as needed.

That’s why a graph can be perfectly fine in Blender while its runtime status says Unsynced. It simply means the engine does not currently have a usable synchronized copy of that graph.

What Happens Where?

GRAPH TYPEDESCRIPTION

Graph SerializationAdd-onAutomaticWonderTexture translates Blender graph state into data the native engine can understand.

Graph EvaluationEngineGPUThe native runtime evaluates texture-processing nodes and manages processing dependencies.

Node / 2D PreviewsEngine + BlenderInteractiveThe engine produces preview results while the add-on displays them inside Blender.

Bake to BlenderEngine + BlenderManualThe engine produces enabled outputs and the add-on creates or updates their Blender image data.

Blender MaterialBlenderPersistentThe add-on builds or updates a Blender material that consumes baked WonderTexture outputs.

NextGraph Lifecycle & Sync →