What is better: A few big objects, or many small ones?

[GameplayProgger]

Hi, a question regarding the usage of modular assets, for example to create a building.
Should they be merged or left as is - which is better for performance?

In favour of merging:
Saving lightmaps for each of 1000 objects on the map,
OR having just a few lightmaps?

In favour of not merging:

A big object not culling off as its too big and almost always occupies some screen space, draining performance.
OR this object being split in parts and not visible parts being culled off?

Please explain which is better in which cases

Thank you

[ClockworkOcean]

If you’re using static lighting, there’s no difference. Because it’s a toss up between two things

1. Lots of modular objects with their own lightmaps

2. One massive object with a huge lightmap

If you’re using Lumen, then modular is better, especially when the meshes have volume ( like wall depth ) and are not just planes.

In terms of rendering, yes, a large object will always draw if any part of it is on the screen. So better again, to chop it up.

The only caveat to that is possibly Nanite. I’m not sure what does with large objects when parts of them are not visible.

[GameplayProgger]

Thanks a lot for your help, sorry for the previous misunderstanding. Text communication and cultural barriers make it harder to understand what people actually mean, as we dont see the emotions of the person. Peace

[ClockworkOcean]

Peace to you also