I have merged some static meshes together to make walls for a room. The walls are made out of pieces. The merged static mesh has 4 million triangle and 2.6 million vertices approximately.
When I to change to LOD Group to something else (LevelArchitecture in this case) the editor will say “waiting for static meshes to be ready” for a very long time and it stays that way for about 20 minutes before I give up. The UnrealEditor.exe starts using up more and more RAM but doesn’t seem to ever complete the task.
Am I giving up to quickly or is there a better way of doing this?
System Specs:
AMD Ryzen 7 3800X (8-core at 3500mhz)
32GB DDR4 Ram at 3200MHZ (Kingston)
Nvidia RTX 2070 Super 8GB (Studio Drivers)
SSD Samsung 990 Pro 2TB
SSD Seagate Firecuda 120 2TB
I will try with again with other lod group types but so far it seems to be doing the same.
Hi, even if you were using Nanite that’s probably going to be way too many triangles - not only for the CPU when generating LODs etc, but VRAM, disk footprint etc
Thanks for the your advice. I tried simplify on the mesh and got it down to under 1 million on triangles and vertices but it still won’t work. If I do any less the mesh gets messed up.
I will have to try some other options to improve performance.
Often you’ll find that if you leave the modular parts as individual meshes and instance them all, you can get slightly faster performance - it’s dependent on the number of materials (draw calls) though - if most of the modular parts are sharing materials especially.
If you’re wanting built buildings for destruction etc you can swap in geometry collections when needed.
Thanks for the information I will look into instanced static meshes. However I feel its too late for that as I have already merged lots of rooms together and would have to do a lot of work to rebuild them.
Tonight I will leave my pc on and see if I can build the LOD’s using edit property matrix. Just curious to see if it would complete the task given enough time.