Object Set Use Camera Lock Parent

Let's be honest up front: you will probably never need Object Set Use Camera Lock Parent, and that's fine. It's part of Avatar Graph - Avatech's pack for rigging a flat 2D character image into an animatable, real-time avatar using an embedded Blender (bpy) - and it's one of the auto-generated ObjectSet_ nodes, the family that sets a single property on a Blender object and passes it along. This particular one just happens to wrap a Blender flag whose use case is vanishingly rare in avatar work.

What it actually does

Blender's use_camera_lock_parent flag comes from motion tracking. When you track a camera into a 3D scene, Blender can use the tracking data to move the camera - and because the object is parented to something in the scene, that movement can drag the object around relative to where you wanted it. Setting use_camera_lock_parent to true locks the parent relationship relative to the tracked camera, so the object doesn't get yanked along when the camera's tracking data shifts.

For Avatar Graph's pipeline - segment a character image, build a mesh layer, rig it, animate it - there's no tracked camera and no motion-tracking solve, so this flag has nothing to act on. It defaults to false and that default is correct. It's here purely because the pack scrapes every simple attribute Blender's object type exposes and wraps each one in a node (check blender/ops_mesh.py for the full list). Some of those attributes are useful for avatars. Some, like this one, are just... present.

The inputs and outputs that matter

• BPY_OBJ - the Blender object to set the flag on.
• value - boolean, default false. true enables the camera-lock-parent behavior.

Output is a single BPY_OBJ - the same object, flag applied. Wire it back into the next Blender node in your graph.

How to install it

It ships inside the Avatar Graph pack; there's no standalone version. Install via ComfyUI Manager by searching "Avatar Graph", or manually:

cd ComfyUI/custom_nodes
git clone https://github.com/avatechai/avatar-graph-comfyui
cd avatar-graph-comfyui
pip install -r requirements.txt

then restart ComfyUI. The usual Avatar Graph snag applies: it pins bpy==3.6.0, which only has wheels for Python 3.10.x, so the install fails on any newer Python. Dedicated python=3.10 conda env on Linux/macOS, or Avatech's prebuilt Python-3.10 ComfyUI zip on Windows. The pack also pulls SAM and MediaPipe models on first launch, so the first start is slower than the rest.

Gotchas

The honest troubleshooting answer here is: if you flip this to true and nothing changes, that's expected - nothing in the avatar pipeline tracks a camera. It's a no-op unless you've built your own motion-tracking setup around the output. Don't spend time debugging it.