First introduced in The Lord of the Rings films, PSD creates a high-dimensional space where every combination of joints produces a unique corrective shape. Old PSD was impossibly heavy. New PSD uses and sparse training —the artist only sculpts 20-30 extreme poses, and the system interpolates the 2000 intermediate poses via a compute shader. The neural inference runs at negligible cost.
At its core, a morph target is a deformed copy of a base mesh. Instead of moving bones to drive vertices, you store a second set of vertex positions. At runtime, the GPU linearly interpolates each vertex from its position towards its Target Pose position.
Historically, creating morph targets required a "round-trip" workflow: exporting a base mesh from a game engine, sculpting the target in Autodesk Maya or Blender , and re-importing the result.
First introduced in The Lord of the Rings films, PSD creates a high-dimensional space where every combination of joints produces a unique corrective shape. Old PSD was impossibly heavy. New PSD uses and sparse training —the artist only sculpts 20-30 extreme poses, and the system interpolates the 2000 intermediate poses via a compute shader. The neural inference runs at negligible cost.
At its core, a morph target is a deformed copy of a base mesh. Instead of moving bones to drive vertices, you store a second set of vertex positions. At runtime, the GPU linearly interpolates each vertex from its position towards its Target Pose position. morph target animation new
Historically, creating morph targets required a "round-trip" workflow: exporting a base mesh from a game engine, sculpting the target in Autodesk Maya or Blender , and re-importing the result. First introduced in The Lord of the Rings