According do unrealwiki, 2 leaf nodes per poly means the best ratio (perfectly balanced tree).
Taking that statement in account, does it mean that in any given map, 2:1 ratio gives the best performance scenario?
1:1 ratio, too much polies to render. 3:1 ratio, too much VIS calculations wasting too much CPU cycles.
no idea
i'd think that's right, but i don't know how well optimised VIS is nowadays
2 leafnodes for every poly? That sounds ridiculous, indeed impossible, unless I'm misunderstanding the point. I tend to keep to as small a number of leafnodes as possible, having several hundred polys (including prop geometry) per leafnode in general. If a leafnode only has a few polys in it, whats the point in the leaf? You might as well combine it with one nearby if possible because there is never going to be any point in culling those few polys as the calculations involved in deciding to do so would be more expensive than just rendering them.
I'm not familiar with the unreal engine so perhaps their use of the term leafnode is different to that of the source engine. Could you link me to the article please?
I'm with Reno on this one. To me 2 leaf nodes per poly sounds totally wierd. I am not experienced with Unreal technology, but still I am pretty sure that you would not have 2 leaf nodes per poly.
The initial thought about an optimal ratio between polygons and leaf nodes is however a valid question.
http://wiki.beyondunreal.com/wiki/Node_Count (nodes are powers of two numbers?)
Thats like using Octrees and Frustum Detection in scenes around 250 polies :roll:
Aha, Quite differente from HL/Source then, which is explained in a nice way here:
Don’t have an account yet? Register yourself now and be a part of our community!
