0

I am making a procedurally generated cave in UE. I have a question regarding what's the most optimised way to do it. I already have a function that creates the verticies and triangles for a cave section from point a A to point B (The final cave consisting of multiple cave sections).

My question is, should the entire cave consist of:

  • A single procedural mesh component that contains all the cave sections in a single Mesh Section
  • A single procedural mesh component that has a different Mesh Section for every cave section
  • Multiple procedural mesh components, each for a cave section
  • Multiple child actors, each representing a cave section

Thank you in advance.

glad_hd
  • 27
  • 3

1 Answers1

1

I would advise against using a single large mesh for the following reasons:

  • Entire mesh needs to reside in GPU memory any time part of it is visible.
  • UV, texture, normal and shadow mapping resolution is reduced.
  • LOD is impractical.

I would split each cave section into its own actor depending on poly count and visibility.