Zagara has a organic animation on her back, that looks a bit blocky. The light will manipulate the ground appearance, to track my position and allow me to stay there, without having to layer and animate an object to remember it. Then make a transparent light source, (Say hero light radius) that gpu will triangulate my position based off. So the idea is, I can make multiple materials on the ground. So long as there’s a pattern or edge or what ever type of something there, the GPU is already working on this spot that allows me to stay there as well. Is it using a texture? Maybe it’s a long drawn out process… Tessellating this may allow the gpu to render my current position, without using the CPU to confirm it and save it. Is there a vertex there? How does remember where to keep me. Im looking at the ground in game wondering, how it is the game knows where I am standing. Same idea in a different sense, with new tools to accomplish it.Įxample maybe difficult for me to tell, without getting away from the point. Oh yeah 4 LODS will do this I suppose in due process, like building a base mesh then adding more to it each time to produce a result. The limitations will how be many materials you support per object and how many variants of patterns, you allow per material. These spots are then recognized as a material, then GPU then will render these materials as a single object though with different consistencies in shading. You’ll flag a set vertex and edges with a pattern, this flag will be taken by the GPU to produce a pattern in the isolated spots. Assuming there’s not enough vertex’s to work with, you’ll have the GPU add some pattern of geometry based on guides. I am talking about both texture mapping and tessellating, for advancement in material gpu level automated recognition. With Tessellation you’re adding a very expensive operation for no benefit because your camera distance doesn’t move far enough. HOTS IIRC does this which is how the character model on the screen appears so detailed but it doesn’t crash your game with that much geometry in a match. Tessellation is rather expensive anyway, it’s better to use multiple levels of LOD like how Overwatch has 4 levels of LOD for character models. Aliasing is computing the average color between pixels to remove jaggies. Usually a normal or bump map is used for this.Īre you confusing aliasing with tessellation? Tessellation is polygonal geometry. Tessellation is the mathematical addition of triangles to add more geometry to a model that wasn’t originally modeled with that level of geometry. Still a still image is boring and need some attention to details, that I am sure you don’t want to manually setup. A guide for the GPU to shade and light even if unchanging. You’re looking down on the ground unchanged I suppose it’ll only be a guide to texturing the ground, or one sided trees and buildings unchanging also need a guide to texture on. It looks cool and is the length the whole city block, but in order to replicate that in game, there needs to be geometry to support the structure of texturing.Īnd the last example I posted about armor plates bending and sckewing was just an additional though I noticed and remembered to post. Example being we have a wall in our city, people have tagged and spray painted all over. I suppose I also know where you guys are coming from, considering tessellation as a large surface mapping. I mean both examples are most likely done in this game, or well managed geometry hides the things I speak of. Armor and other objects have hard corner edges, when I think they are supposed to be rounded.Įxample detailing embossing textures on a sword or armor, making it look like it’s popping or conclave. I see the same type of thing, on a lot of the characters in game. It’s more like isolated tessellation, for surface texturing and smooth edging.Įxample you’ll see a lot less these days is, wheels that aren’t round but instead blocky.
0 Comments
Leave a Reply. |