Skip to main content
Version: 14 May 2024

Session Files

Use session files to re-use your current scene setup for testing or to share your setup with other members of your team. Session information includes:

  • Current scene graph layout, such as your lights or the contents of your virtual room
  • Gizmo state, a combination of the virtual room-specific gizmo state and the state of the Simulator property panels
  • Observed Geometry Cache (scanned environment state and for meshing and planes APIs)

You can save or load a session file from the Session dropdown in the Target Selector:

Session Files Menu with Example Session

By default, session files are saved and opened from the /MagicLeap/tools/zi/<version>VirtualDevice/data/VirtualRooms/ExampleRooms folder.

Save a Session File

  1. Expand the Session drop-down list.
  2. Click Save Session, and then enter the file name.
Save a Session Dialogue Window

Your session data is now saved. See the section Share a Session on how to share a session file.

Open a Session File

  1. Expand the Session drop-down list.
  2. You can open a recent session directly from the drop-down list, or click Open Session.
  3. Select the session you want to open.

Your session is now loaded into the scene view, including the virtual room, hand gestures and any scanned meshes.

When you make changes to the scene, you can save the changes to the current session or to a new session using Save As....

Share a Session File

You cannot directly share .session files as the file paths contained in them are local to the machine. You must convert them to a .room archive using the script

The script can directly be executed from the UI.

  1. Expand the Session drop-down list.
  2. Click Package Session as Room to save the currently loaded .session file to a .room archive.

Alternatively, you can also execute it from the command line.

  1. Open a terminal from the Magic Leap Hub (ML Hub).
  2. Ensure python3 can be found.
  3. Run the following command:


python3 %ML_ZI_ROOT%\scripts\ <path_to.session> <>


python3 $ML_ZI_ROOT/scripts/ <path_to.session> <>