All red boxes use the same mesh, same material.
-Distance Cull, LOD is not used.
Type 1 was instanced by 4 groups : Part A, Part B, Part C, Part D.
Type 2 was instanced all Red Boxes at once without segmentation.
It is better to divide groups like type 1 according to the area?
Or it doesn’t have to?
If I Instance all the Red Boxes into one like Tpye2,
Even though the drawcall is small, the actor itself is very huge.
so I think the range of mesh that doesn’t come into my camera will always load as well.
Therefore, I am worried that it will adversely affect the performance.
Oh, I just thought very simple.
Type1 is data containing 16 tubes in 1 actor,
There are 64 Type2s in one actor.
However, due to the “instance,” drawcall would be integrated into 1
So I was not sure if it was right to judge that 2 was huge, so I posted this question.
When I instanced like Type2,
(In play mode) even when I’m only looking at the group A side as a character,
Since I instanced with 1 actor, I expected the cube of B~D side to be loaded at the same time.
Is this right?
I’ll say it on the premise that it’s right.
Is there anything that adversely affects the performance due to “something” other than drawcall that I don’t consider, since the mesh of the parts I don’t need is also loaded?
I was concerned about that.
Thank you. So let’s get to the bottom line.
In any case, I don’t have to divide the zones if I’m gonna instance duplicated Props(use same mesh, same material).
Is it right?
