Modelling - One Part or several and UV Mapping

Questions & Answers
1

Hey,

so I got two quick questions. They both are related to Modelling and one is also to UV Mapping.

The first question is.
Would it be more wise to model something with many parts being separate? So for example if you were to model a house. Would you do it in one go, like use the function called extrude to just extend from for example the bottom. Or would it be better to just Model the walls separate from the bottom, which are both separate from the roof and all those things.

There are two main reasons for why I am asking this.
The first is my seconds question. UV Mapping is easier when doing it from a smaller object so a part from something like a house mentioned above. If walls were separate it would be easer creating textures for those.

However the other reason is more of a concern. When making many different objects for one big object again a house for example. Wouldnt that take more resources from the computer because not only does it have to render more than on object to get one final object but also it has to load more textures which can also increase the power a computer needs to display all of it.

In Short (and sorry if the things above sound weird, its hard to explain that question) what are the ups and downs of modelling either one big model (example: one complete house where everyhting sticks together) or modelling (example: house) many different parts (walls, bottom floor. top floor, roof, all seperate) to form one big model/object.

Thanks a lot for all the help.
Kind Regards
Cribbel

2

I’m going to answer with regards to performance, as that is I believe what you are asking.

The real answer: It depends…

  • Are the meshes static? (will they be moving during runtime?)
  • Are the meshes small in size?
  • Will you potentially see tons of meshes at once?
  • Is your game a first person view?
  • Are you planning on using Occlusion Culling?
    etc,etc

Now in general, I would say it is better to have more, separated meshes, that share the same material, the same texture(s) and the same uv map(s).

These are the benefits:

  1. You can move them around in Unity without reexporting;
  2. The meshes out of view can be culled (frustum culling)
  3. The meshes to all be in a single draw call (if marked as static or dynamic batching is on) BIG OPTIMIZATION HERE
  4. Occlusion culling is now possible.