Scene helpers
NameFrame expresses spatial intent as actors you place in the level. They are visible in the editor, they can be moved with the normal transform tools, and they are what the compiler reads when it turns a scenario into a plan.
The five helpers#
| Helper | Actor class | What it declares |
|---|---|---|
| Spawn Zone | NameFrameSpawnZone |
Where the spawner may place objects. |
| Target Zone | NameFrameTargetZone |
The region of interest a camera should be looking at. |
| Exclusion Zone | NameFrameExclusionZone |
Space placement must avoid, according to the adopted placement contract. |
| Camera Zone | NameFrameCameraZone |
Where a sampled camera may stand. |
| Flight Path | NameFrameFlightPath |
A path for path-based camera and scene authoring. |
All five can be created from the Studio panel, and all five are ordinary level actors afterwards: select one in the outliner, move it, and the next dry run reflects the change.
Why they are separate concepts#
It is tempting to collapse them. Resist it, because they answer different questions and a scene usually needs them to disagree.
A Spawn Zone says where objects may be. A Target Zone says where the camera should look. They are frequently different volumes: you want a camera aimed at the loading bay while crates are scattered across the whole yard, and if one volume served both purposes you could not express that.
A Camera Zone is a third thing again: where the camera may stand. Zone shapes are what make a viewpoint distribution something you authored rather than something that happened.
Preview actors are not dataset positives#
Studio can materialise preview actors so you can see what a rule will place before committing to a run. Those previews are temporary and are cleared from the capture target set before generation.
This matters more than it sounds: a preview that survived into a capture would appear in the identity buffer and become labelled objects that nobody intended to spawn. The clearing step exists so that looking at your scene cannot change your dataset.
Next#
Helpers describe space. Classes and labels describe what the objects in that space mean, and the two together are the minimum a capture needs.