Home Technical Talk

Ground and buildings pipeline/workflow/approaches

I was surprised not to find a separate section dedicated to game development, so I'm asking here. While I learned a lot about making game assets recently, there are two topics that absolute mystery to me.

The first one is how to make a ground for a game. Imagine having a road, a sidewalk, a gravel/dirt whatever. E.g. here is from Escape from Duckov:

Or The Ascent:

Or Mutant Year Zero:

Should I use tile approach and create fixed size tiles for every surface as well as transitioning types, create a large non-square areas filled with tileable texture, or something else?

And the second question about architecture. I want to make buildings with navigable interior for the game. I can do the following:
  • Either create a unique mesh completely in Blender and paint it in the Substance and then make an asset in Unity
  • Or create smaller blocks (like bricks or wall tiles) and build a prefab inside the Unity
  • Or there is another approach? (As usual, a combination of the both)
Like again in MYZ:

This example less relevant to my case since I'm planning to have intact buildings (the world is just abandoned), thus this approach might be less efficient. In the other hand, I could use these bricks as debris as well as building blocks. And not paint another building which has the same material.

Any ideas?

Replies

  • poopipe
    Offline / Send Message
    poopipe grand marshal polycounter
    The whole forum is about game development, we don't need a section :)

    The problem with these sort of questions is that the answer depends on what you're making and what your engine and toolchain can cope with. 
    and
    on top of that - the answer will often be "all of the above" 

    Your best bet on the terrain / road stuff is to explain what you're trying to achieve (ideally with some concept art or reference photos) and which engine you want to do it. People will have many ideas and you'll learn terms to search for. 
    Your suggested approach could be the best solution in some situations while also being the worst possible solution in others - we need more context.


    as far as the building goes.. 
    One way to think about this is to consider your range of view distances. 
    2m - 200m - you're probably best off using modular pieces cos you get more flexibility and you don't need to worry about what happens at a distance
    2m - 5000m - you're probably best off making buildings as one model so you can lod them properly and get less entities overall.
  • maaboo
    Thanks. Well, as for references, I already provided them. It's something similar to any of the ref with more/less same view angle and distance, i.e. it will be top-down action with kind of isometric view (actually perspective but it's very subtle) and the engine is Unity (I mentioned it, but you're right, I should clearly indicate it in the very beginning). LODs are not needed in my case.

    I'd prefer the approach that saves me a lot of time (since I'm making it alone) yet providing a good optimisation on resource loading since I don't see that polycount/VRAM is an issue.
Sign In or Register to comment.