HooTmAn said:
I did optimize the ori. mod, this is also a propper PL.. did you even try this submod?!
I think i made a good varrity for the hitboxes too, so it does not feel claustrophobic at all..
I made new lods too, pls stop takling about my submod like i did not even spend time at it.. thx
It's great that you did a submod and adapted that mod to work correctly with POP. It's great for people who can afford the performance hit, and I'm not trying to diminish your work.
I was talking about the original mod. But let's be objective so you know what I'm talking about. Just made a fresh folder with your last version.
-There are 6 trees with lods (in tree_f) and 8 bushes that seem to have them, but there's only one that makes them disappear at lod2.
-Trees are generally much more complex. For example, tree_10_a is 4603 vertexes. For reference, the native model on this one is 86 vertexes. We limited 3.8 armors to a grant maximum of 2000, with lods going as simple as 50. It is not an exception: the whole of native "xtree_meshes_b" is 4926 vertexes, while PL's one is 30969.
-Some trees use very HD textures. For example, trees that use material sp_pine_needles_2 use it hundreds of time, and it's 512x512 px. For reference, every helmet in POP use textures of 512x512 max, usually twice (mirrored texture).
-Some hitboxes are absolutely too complex. Hitboxes require a whole lot of computation, so they're usually simplified with a few shapes. Usually, trees can use as few as 1 cylinder. "bo_aspen_b" uses the original model, which is 2081 vertexes, including irrelevant branches and stuff.
-Running the error finder in openbrf, it reveals the submod has
268 106 errors. Missing collisions, missing meshes, duplicate collisions, textures and materials.
So what I was saying is: that would be great if someone decided to do the work gutekfiutek refused to do, as a modder's resource. Until then, it's still objectively unoptimized, and reserved to people with higher end computers.