I wonder how much we can push the poly count if we use a single material for every object.
I’m modeling an environment and all objects ( 2 human, 8 animals, 10 environment prop, and ground map ) are sharing the same material ( 128 x 128 pixel texture ). If i have them all on screen ( and duplicates ) will it be one draw call ?
Looking a mobile game as reference (archero):
It uses a pre rendered background, and the other map stuff are 3d.
I ask myself if using the entire map as 3d can be better/faster.
In my case I would be trading a single map ( 768x2048 ) for 4k tris.
But being able to create multiple variations of the the map without incresing the app size.
and what if we think that we need bigger maps later ?
As I write this i think the case is that they want use more memory usage in return to being able to focus in a good rendertime? I think that makes sense if theyr memory usage is low.
I went here wanting a simple answer and as i write i see that its a very complicated topic that changes game-by-game and we would make test with some devices to have an idea.
This post is turning into a mental conversation, so I will create the topic now
I will be glad if anyone join the discussion, or point to link to a good reading.
I may add more info later.