Skip to main content

Concepts

Overview​

A Concept is a place to generate, edit, and iterate on images for a creative subject. Use it to explore a direction, compare variations, and choose an image to carry into other StudioTwin tools.

ConceptLab is the visual workspace for this process. Character and Object are the first Concept Types you can create and refine. The resulting images are prepared for use with downstream 3D mesh generation tools. Character rigging is Coming soon.

Concept Types​

Every Concept has a Concept Type that describes the subject and connects it to relevant workflows.

Concept Types:

  • Object for props and other discrete subjects.
  • Character for character development.
  • (Coming soon) Environment Map for image-based environments and lighting.
  • (Coming soon)World for world and scene development.

Develop a direction​

Each generated, uploaded, or edited image stays with the Concept, so you can:

  • review earlier versions without losing your latest work;
  • compare two images to see how a design changed;
  • choose the image that represents your current direction; and
  • continue editing from a previous image when you want to explore another path.

The image you select becomes the starting point for the next edit.

Work with Concepts across StudioTwin​

ConceptLab​

Use ConceptLab to create a Concept, generate or upload a starting image, make visual changes, compare versions, and download the image you choose. ConceptLab is available in Preview.

MCP​

You can also ask an agent to work with Concepts through StudioTwin Web MCP or Unreal Engine MCP. The Web MCP path works without a DCC application; the Unreal Engine path combines StudioTwin capabilities with work in an open UE project.

Example: develop a game character​

  1. Create a Character Concept and describe the character's silhouette, clothing, and visual style.
  2. Generate a starting image, then edit details such as colors, materials, or accessories.
  3. Compare the variations and select the image that best represents the character.
  4. Use the selected image as input for downstream 3D mesh generation.
  5. (Coming soon) Rig the character.

Review the output at each stage before continuing. Generated results can vary.