If I wanted to build say a cliff face by using lots of smaller static mesh rocks jumbled together what happens in terms of performance to the “backside” of the SM’s that are not viewed when within the game? Would I effectively still be rendering each individual SM rock even if only a small portion may be viewed game side?
If this is the case would I be better building cliffs using the SM rocks in my modelling software (Modo) and then chopping out the “backside sections” that would never be viewed?
Am I missing any more established techniques for this kind of thing?
Thanks in advance for any pointers.
We usually delete backfaces that are never viewed in-game, regardless of the engine being used for said game.
so you would delete the backfaces in modelling software?
Correct, you would do that. You can’t select individual faces in UE4 and delete them, so you have to do it in your app (i.e. Maya, etc.)
In my personal workflow, I will build out and mesh a level first, once my lead is ok with everything and I’ve completed changes based on feedback, I will go in and do an optimization pass. This is when I will start chopping stuff the player never sees. Do this at the end of your process, once you’re done meshing out in order to avoid headaches (i.e. you chop off the back of a rock, your map lead decided the hill is being moved so now the rock is exposed and you have to revert your work- not good )