So I was wondering if any of the game artist vets could answer my question. When modeling why not combine everything? It saves polys and looks the same.
depends on if you have per pixel lighting or per vertex. At this point combining/clipping whatnot objects is pretty standard practice, as texturing does a lot of the work of making the form of a model.
However if you were going into an engine or platform that had decent processing power but limited memory, and you had to use very small textures, or vertex colors (like the current comp) then the extra 6 polys from something like that would be beneficial because you can use them for lighting data to make the model look better while saving on texture space/memory.
Agreed, it's engine specific. We had a light mapping engine on my last next gen project so clipping it in would cause a lighting seam as well as waste a lot of map space, I try to make all my models seamless now within reason, I also only model double sided now. Lighting looks junk real time or baked on single sided models.
So if you guys had to give advice to someone making a environment/prop portfolio what would you suggest? I know its kind of an 'it depends' question, but would you model with extrusions or clipping/combining?
Replies
However if you were going into an engine or platform that had decent processing power but limited memory, and you had to use very small textures, or vertex colors (like the current comp) then the extra 6 polys from something like that would be beneficial because you can use them for lighting data to make the model look better while saving on texture space/memory.
That said I'd keep props as much a single mesh as is reasonable for their shape.
Ie a palm tree could have its leaves separate from its stalk. but a rock formation will likely want to be a single solid mesh.