Trouble with exporting; seperated polygons

[anomaly]

So I've been trying to find a way to convert my model to .xsi format but no matter what route I take, I keep running into the same damn problem.  My ideal method was to make them in Gmax, export to .md3, import into Blender 3D, and use OM's .xsi exporter.  That didn't work as when I exported to .xsi all the polygons where separated from each other and I ended up with a model with a huge number of vertices.  So I tried converting .md3 to .obj using Lithunwrap, and then use blender to export to .xsi: same problem.  Even exporting a model made in Blender does this too.  Thinking it was a problem with OM's exporter, I tried using Lithunwrap to convert .md3 to .x.  Looking at the .x file, it too had all tri's separated.  However, if I use Gmax or Lithunwrap to export to .obj I end up with a proper model with each sub-object in one piece.  So, for some reason, converting to .x and .xsi results in seperated, detached polygons.

And, yes, I did make sure the model was in one piece before exporting from Gmax.  Also, all of these tests were done with untextured and unanimated models, if that matters.

So my questions are does anyone know what I'm doing wrong, and what methods do you successful people use to convert to .xsi?

[Nielk1]

If you mean a square became 2 triangles, that is inevitable. XSIs for BZ2, aside from the flames, are triangulated. If any mesh is not triangulated and is not for the flame is bugs out and AVs.

Unless I misunderstand your issue.

[Clavin12]

if you mean all the polygons became their own thing couldn't you just join the meshes(if they were separate meshes) and remove double vertices?

[anomaly]

If you mean a square became 2 triangles, that is inevitable. XSIs for BZ2, aside from the flames, are triangulated. If any mesh is not triangulated and is not for the flame is bugs out and AVs.

Unless I misunderstand your issue.

No, that's not what I mean.  What I mean is, each triangle becomes detached from its adjacent polygons.  So each triangle has its own 3 vertices that it no longer shares with any other triangle, which leads to the huge vertex count.  

if you mean all the polygons became their own thing couldn't you just join the meshes(if they were separate meshes) and remove double vertices?

All triangles are still in their respective sub-objects, if that's what you were wondering, just not attached to adjacent triangles like they should be.  And it wouldn't be a matter of deleting extra vertices but welding them together.  Which is not something I would attempt in notepad.

[Nielk1]

So a square, instead of having 4 verts, has 6?

[anomaly]

So a square, instead of having 4 verts, has 6?

Exactly.

[Clavin12]

In blender welding vertices is the same as deleting double vertices.

[anomaly]

I know a few people here use blender for modeling and have presumably used OM's exporter.  Do your models contain a large number of vertices?  And if they don't, could someone lay out the basic steps you take between finishing the shaping of the model (before texturing and animating) and exporting.

[Clavin12]

Define large.

[Nielk1]

Define large.

Previous posts in this topic make what he means clear.

[mrtwosheds]

Does this make any difference to the .msh that bz2 creates?

[Nielk1]

I figure he can't get that far because his models blow up BZ2.

[Clavin12]

I don't understand the problem. Can he just weld vertices?

[anomaly]

I don't understand the problem. Can he just weld vertices?

Because they would be dewelded once I exported again.

But, I think this is all irrelevant now because I now think that this is how .x and .xsi is supposed to export.  I really should have done this in the first place but I only just now checked some of the stock .xsi's and pretty much all of them have at least a decent number of duplicated vertices so I think my models should be fine.  Also, some graphical problems that I was having with one of my models that I thought was caused by the detached polygons was actually caused by something else.  So, I guess detached polys  are actually the norm.

I feel kind of stupid for not realizing this before.  Sorry for the trouble.

Also, I know there is a poly limit on models and I think it was raised with TA 1.5 but, is there also a vertex limit?

[VSMIT]

The vertex limit for 5.1 is 60k verticies.  Note that this does not mean verticies shared by multiple polygons.  It means individual verticies attached to a polygon (3 per poly).