Skip to content

armgit / viz3d

Public Imported from GitHub
0

An AI agent skill that turns any subject into a 3D film or app

Brief: NLA clip order sets the exported rest pose

Nguyen Xuan Ana62ffc20882a
diff --git a/plugins/visualize-in-3d/skills/visualize-in-3d/references/brief.md b/plugins/visualize-in-3d/skills/visualize-in-3d/references/brief.md
index 17d3c00..17ab802 100644
--- a/plugins/visualize-in-3d/skills/visualize-in-3d/references/brief.md
+++ b/plugins/visualize-in-3d/skills/visualize-in-3d/references/brief.md
@@ -31,7 +31,8 @@ to a log file you read afterwards. Keep scratch files inside {{FOLDER}}, never i
Blender gotchas: --factory-startup still loads the default cube, camera and light, so start each script with
bpy.ops.wm.read_factory_settings(use_empty=True). Model each vehicle facing Blender -Y; the glTF exporter turns that
into +Z, so lookAt works in three.js. Write the facing into ASSETS.md. glTF export of clips needs
-export_animation_mode="NLA_TRACKS" with each clip pushed to an NLA strip. bmesh.ops.bevel gives new faces material
+export_animation_mode="NLA_TRACKS" with each clip pushed to an NLA strip. The exporter takes the rest pose from the last NLA
+track at its first frame, so push a clip that starts at rest last. bmesh.ops.bevel gives new faces material
index 0, so set the materials of the bevel faces again on multi-material meshes. Blender renames duplicate names to
.001, so prefix each child with its variant name (Car_Cab_Door_L, not Door_L).