Bad Mesh size (> 64K faces)

Posted By: Slin

Bad Mesh size (> 64K faces) - 07/16/11 18:07

Where does this restriction come from?
I would expect the vertex count to be limited by the datatype/size of the indices, which in case of 2byte would be 65536 vertices used by as many triangles as I want. Is there any chance that this could be a bit relaxed in the near future?
And what about 32bit indices?
Are submeshs of models culled sperately if they are big? Do they share their vertexbuffers (it seems so, and probably also the indexbuffer is shared, as the 64k faces restriction is for a complete model, counting all submeshs)?

I am just wondering, as the 64k restriction seems a bit outdated. I currently for example have some water mesh with 64k vertices but about 100k faces, which I would like to use as one mesh, not having to split it into several meshes...
Posted By: Lukas

Re: Bad Mesh size (> 64K faces) - 07/16/11 20:19

I think I remember jcl saying that it is, as you suggest, because of the 2byte indices and the actual limit is actually 65536(-1?).
Posted By: jcl

Re: Bad Mesh size (> 64K faces) - 07/18/11 08:42

Both vertices and triangles use 2-byte indices and thus are limited to 64K. Most 3D cards today support 32 bit indices, but some old cards don't, that's why we still keep this limitation. Submeshes are just render passes of a mesh, and can not be culled separately.

However, a water mesh with 100K polys should not be a problem - just load it as terrain. Terrain can have an unlimited number of triangles and vertices.
Posted By: Slin

Re: Bad Mesh size (> 64K faces) - 07/18/11 09:01

What are triangles indexed for?
The problem with terrain is, that my terrain would need to have a resolution of something like 100kx100k vertices, which I havenīt tried for a long time and can may be created per script, but at least MED when I last tried it was not able to handle terrains with a higher resolution than 1024x1024 or something like that.
Also, I donīt need a plane, but just some rivers to be covered. I could of course masked terrain, but that would still cause way too many vertices to be sent and processed by my graphicscard. And I am also not sure if the terrain LOD would be able to reduce the polycount as much as I would like it to.
Posted By: Rackscha

Re: Bad Mesh size (> 64K faces) - 07/28/11 12:31

Originally Posted By: jcl
Both vertices and triangles use 2-byte indices and thus are limited to 64K. Most 3D cards today support 32 bit indices, but some old cards don't, that's why we still keep this limitation. Submeshes are just render passes of a mesh, and can not be culled separately.


Can you tell me the most recent 3DCard that does not support 32bit indices? Just for the interest...and for the question(if to old) of keeping the compatibility o.O.

Greets
Rackscha
Posted By: jcl

Re: Bad Mesh size (> 64K faces) - 07/28/11 13:07

I don't know the most recent card with 16 bit indices, but our legacy support goal is 2006, and several cards and systems back then didn't support 32 bit.
Posted By: HeelX

Re: Bad Mesh size (> 64K faces) - 07/28/11 13:50

Would it be possible for you, that if a user sets a special flag, that you deactivate that legacy support for 2 byte indices in the engine?
Posted By: jcl

Re: Bad Mesh size (> 64K faces) - 08/01/11 08:24

Sure, in a future version we will support 32 bit indices, either always or with such a flag.
© 2024 lite-C Forums