Hot | Visual Space - Page 6

A scene is most useful when you can return to it and other people can understand what changed. Stage combines save status, local Undo history, a linked Article and Comments around the same document.

References make a rough scene easier to understand. Use a picture to show a material, a cutout to suggest a person or tree, and a note to explain the decision beside the object it affects.

An AI render turns a captured Stage view into an image guided by your brief. Use it to explore materials, atmosphere or presentation while keeping the editable 3D scene as the basis for your next decision.

Materials describe the surfaces in your scene; Environment describes the light around them. Together they help distinguish a concrete courtyard, a grassy slope and a bright interior while keeping the same underlying forms.

Landscape tools turn a flat arrangement into a place with elevation, planting and water. Start with the ground shape, then add the smaller details that help people understand scale and use.

Interactions let you describe how a place is meant to behave and compare different designs within it. They are useful for game ideas, spatial stories and presentations, provided everyone understands that **Stage records intent rather than running a simulation**.

Stage helps you develop a spatial idea into a scene that people can explore and compare. Start with the workspace, then choose a guide for the part of the project you are shaping.

Stage separates exploring a scene from saving a deliberate viewpoint. Your regular **User camera** is for working; named camera objects preserve compositions you want to return to or render.

Make a small landscape with a raised area, a planting patch and a pool in the low ground. Start with a scene you authored and decide the rough footprint before creating terrain.

Save a deliberate camera view, combine it with a mood brief, and generate an image you can revisit in Docs. You can complete the camera preparation without generating an image; the render step uses the space's AI credits. You must be the scene author to render, with space-owner authority or **Use AI credits** permission in its area. A configured Agent and a readable, nonempty credit balance are also required.