I’m creating my terrain in Blender, as I feel more comfortable with it than using Unity’s built-in tool. I have this mesh now, and I’ve implemented the LOD feature as well. The next problem to fix is the collider. I want to use the mesh collider for it. But I’m worried about performances. I split already my terrain in “chunks” each with its own LOD. I need to apply to each square a mesh collider.
What do you think about it? And also do you have any optimization advise?
Extra question: should the mesh be static? Which would be the advantages to have it static?
Thank you guys for your time. To some, they might seem such stupid questions, but from someone who is starting now…Well, they still represent a big challenge.
I’d advise against doing your whole terrain as a mesh with meshcolliders. I do actually have a similar system for my caves where i model them in houdini and then have a meshcollider for every quad. But in small spaces you can heavily optimize what’s shown and which meshcolliders are active. Also if you have your terrain in chunks you’ll have to deal with the texture transition between every chunk etc. For every meshcollider you’ll also have to save a gameobject which will make loading times and scene files longer/bigger. Also editor performance tanks quite heavily with a lot of meshcolliders and gizmos on… It’s just not worth it imho. Why not just use unitys terrain system? You can convert your mesh to a terrain. I think there are some free assets for that last I checked. Also don’t forget to add unity’s terrain tools from the package manager. You’ll have to click Advanced/Show preview packages to see them. They add ALOT of features to work on your terrain.
My problem is that I feel uncomfortable with Unity’s terrain tool… With Blender, I can do things much faster and sometimes easier.
I fixed the problem of the transaction between meshes thanks to a python script in Blender.
Now my only concern is the colliders, and following what you’re saying, I realised that my idea is not worth it.