Mesh Format v2 #10

Open
opened 2025-03-14 22:31:29 +00:00 by Quaternions · 2 comments
Owner
  • Only triangle lists allowed instead of polygon lists - exact face normals
  • Change vertex precision to P32 - face normals fit in full precision & reduces final calculation precision requirements
  • Change rotation precision to P32 - reduces final calculation precision requirements
  • Translation precision remains P64 - place models in the full range of world coordinates
- Only triangle lists allowed instead of polygon lists - exact face normals - Change vertex precision to P32 - face normals fit in full precision & reduces final calculation precision requirements - Change rotation precision to P32 - reduces final calculation precision requirements - Translation precision remains P64 - place models in the full range of world coordinates
Author
Owner

An idea that crossed my mind was to use a fractional-only Fixed<32,31> for vertices and a integer-only Fixed<32,0> for transforms, but this is not obvious to use, and has the same output range and precision as simply using balanced fix points. It's also possible to double the range by making vertices unsigned, but this is a lot of engineering work and is more difficult to use. The physics also expects the origin to be contained within the mesh.

An idea that crossed my mind was to use a fractional-only `Fixed<32,31>` for vertices and a integer-only `Fixed<32,0>` for transforms, but this is not obvious to use, and has the same output range and precision as simply using balanced fix points. It's also possible to double the range by making vertices unsigned, but this is a lot of engineering work and is more difficult to use. The physics also expects the origin to be contained within the mesh.
Author
Owner

WIP branch mesh-v2

WIP branch `mesh-v2`
Sign in to join this conversation.
No Label
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: StrafesNET/strafe-project#10