NodeRenderView
, which is intensely used by the viewer
core as well as viewer extensions.
RenderViews
The nodes from the previously defined WorldTree can contain any kind of data. Some of them contain rendering related data, like meshes, lines, points. A ‘RenderView’ of a node is just it’s rendering related data from the viewer’s point of view. The viewer will use render views for all rendering related operations instead of the node itself. Across the viewer-core and viewer extensions, render views are required to be passed around for various entry points. All render viewers are accessible through theRenderTree
.
Render Tree
A render tree is just a wrapper around a part of (or the entire) world tree, with additional rendering related functionality. Here’s how you get a render tree instance for the entire model:
getRenderTree
represents the root of the render tree. So the call below will you get you the render tree highlighted in the image below:

Batching
The viewer will automatically attempt to batch together any renderables you provide within aloadObject
cycle. In order for batching to take place, the following conditions need to met:
- Objects need to be of the same GeometryType.
- Objects need to belong to the same subtree i.e loaded from within the same
loadObject
call.
- Instances are rendered as hardware instances.
- Instances of a type will not batch together with non-instances, nor with other types of instances.
- Batches are limited to a maximum of 500,000 vertices.
- Batches have a maximum object count limited by the platform’s
Max Vertex Uniform Vectors
count.