An annotated laboratory cutaway¶

A complete 3.0 example combining a detailed 3D scene with vector annotations, a dimension, a projected transfer route, two detail views and an analytic response plot. All geometry is original and uses no downloaded assets.

What the figure contains¶

The scene has 265 objects: 235 meshes, ten curve objects, fourteen landmark empties, three cameras and three lights. It includes two reactors, a feed pump, control display, a rack of 24 sample vials, a flow cell and detector, gas bottles, a recirculator, overhead extraction and cable services, benches and cabinetry. The open walls and roof allow the overview camera to show the equipment.

Twelve callouts use two ordered label columns. Their leaders and target markers are vector geometry; text remains editable in SVG and embedded in PDF. Solid leaders identify visible targets; dashed leaders identify obscured ones, based on the saved depth pass. The custom column placement is part of this example's authoring code, not an automatic scene-label placement API.

The 8 m footprint dimension uses dimension3d(). The amber transfer route uses path3d(hidden='dash') so hidden sections remain legible. Two orthographic detail views show the process bench and the analysis island. The response curves are analytic illustrations, not simulated or measured outputs of the rendered apparatus. This is a conceptual laboratory, not a construction plan.

Build it¶

From the release checkout with the rendering extras, Blender and Poppler installed:

python examples/v3_complex_scene.py --quality final

The output directory is out/v3-complex. Open figure.html for the review; the folder also contains PNG, SVG, PDF, the editable laboratory.blend, its creation log, a scene inventory and export manifests. The first build creates the geometry, then renders three camera views using the GPU/CPU policy from render jobs. Later builds reuse the scene and cached pixels.

Use --quality preview for faster review, --blender /path/to/blender to select an installation, or --rebuild to replace the generated scene. Changing only the annotations does not require a fresh Blender render.

Figure source · Scene authoring source · Reusable templates · Gallery

Reading the depth annotations¶

Visibility is sampled at each target point or along the route centreline. Glass, silhouettes and transparent regions have the same depth-pass limitations as other scene annotations. The target dots mark authored landmark positions; they do not certify a surface measurement. The underlying scene remains a raster image inside SVG/PDF, while the labels, routes, dimensions and plot remain vector.