Mesh size/complexity vs Lightmaps?
polycounter lvl 10
Heya folks, I just got a question here about something I've never really seen an explanation for. As a self taught user of UDK I've not really found a tutorial or the like that lays out how big a mesh should or at what point it's just better to break it into different parts for assembly in UDK.
I do not have access to my example picture at the moment sadly, but it's like this: I have made a medieval water wheel for a mill, I soon realized that the support of it needed to be its own mesh but after that the paddles still ate up the lightmap.
What are the priorities one need to make here to get the best result in UDK? :poly105:
- Should one strive to need as little lightmap as possible? Even at the cost of extra standalone mesh parts with their own lightmap sets?
- Should one try to make it one whole mesh anyways but just try to be more efficient as well as amp up the lightmap size from just 32/64?
- Is there a fundamental trick I keep missing?
- How do you go about this problem?
Any advice on this would be greatly appreciated
I do not have access to my example picture at the moment sadly, but it's like this: I have made a medieval water wheel for a mill, I soon realized that the support of it needed to be its own mesh but after that the paddles still ate up the lightmap.
What are the priorities one need to make here to get the best result in UDK? :poly105:
- Should one strive to need as little lightmap as possible? Even at the cost of extra standalone mesh parts with their own lightmap sets?
- Should one try to make it one whole mesh anyways but just try to be more efficient as well as amp up the lightmap size from just 32/64?
- Is there a fundamental trick I keep missing?
- How do you go about this problem?
Any advice on this would be greatly appreciated
Replies
General rule on lightmapping is the bigger the object the bigger the lightmaps need to be. Smaller objects usually use small lightmaps or even vertex lit in UDK.
Just as a follow up then, is there any point at all where it's smarter to break an object into several components? Just for future reference.
Edit: Ugh.. and how do you get it to use dynamic lighting? The static mesh doesnt seem to have that option? But I guess neither me or google are good at finding that <.<
Hm, that just made the shadow turn solidly and utterly black. I did some searching and found that the interpactor used dynamic shadows and switching to one of those gave me the right result.. I think. It gives the usual dynamic shadowing without taking on any baked shadows at least. Or is there another way using just static meshes?