Problem Mesh problem: need help

[Andy44]

I've been revamping my simple passenger module add-on, but I've run into a problem with the mesh. In Anim8or this all looks fine, but in Orbiter weird things are showing up. In the below pic you can see the aft starboard solar panel is missing some of its face, the forward starboard panel is skewed, and if you look close, the porthole on the stern hatch looks like it's "broken". In fact, part of the porthole pane is like shard that pokes into the mesh interior. Depending on the scenario and what other vessels are loaded, I get different errors.

Anybody know what causes this and what the fix is?

 

[Donamy]

Make sure you only have single faced polys, Orbiter can't read double faced polys.

 

[Andy44]

Well, I eliminated the rear door group and replaced it with a simple face, and deleted the solar panels and rebuilt them from simple rectangular solids. Still have problems with the new solar panels, and no idea why. Something so simple I can't understand what I'm doing wrong.

 

[NukeET]

Are all the normals in the same direction?

 

[Andy44]

As far as I can tell. But I run into weird things like two segments between the same two points, and one of them is "front" and the other one is "back". I used the shell tool to give the mesh an interior wall, and I think that's where the problems started. I don't know where these comes from. I'm thinking of scrapping the mesh and starting over fresh, though I dread it. I hate throwing away hours of work.

 

[NukeET]

As far as I can tell. But I run into weird things like two segments between the same two points, and one of them is "front" and the other one is "back". I used the shell tool to give the mesh an interior wall, and I think that's where the problems started. I don't know where these comes from. I'm thinking of scrapping the mesh and starting over fresh, though I dread it. I hate throwing away hours of work.

Don't you just love the weird s Anim8or inserts into meshes? I've seen similar things happen with my meshes, and 90% of the time I've had to start over.

 

[Zatnikitelman]

You and me both
On the surfaces that are exhibiting these problems, are there other surfaces just "under" them?

 

[Andy44]

No, even after I delete the faces, the problems are still there. I thnk I'm going to have to scrap and start over. >:-|

 

[Donamy]

Get AC3D, you won't be disappointed.

 

[T.Neo]

Get AC3D, you won't be disappointed.

Is AC3D payware?

 

[Woo482]

XSI mod tool is quite good and its free If you are looking for a new program you should try it

 

[T.Neo]

Anim8or hasn't failed me, yet.

 

[Y.A.X.]

I use GMax with MAX2MSH, for me it's the best choice.

But, I had a problem like that, and it was the texture (a bad alpha channel).

 

[Hielor]

Wings3D.

 

[Bj]

Don't you just love the weird s Anim8or inserts into meshes? I've seen similar things happen with my meshes, and 90% of the time I've had to start over.

Anim8or is not exactly a 'meshing' program. Its more for animating 3D objects. Yes it is easy to use, but G-max is my favorite.

 

[Donamy]

Is AC3D payware?

Well worth the $70 us

 

[tgep]

I've run into this problem myself and here are some of the reasons it occured.

1 ) Group sides were reversed
2 ) Anim8tor inserted a flag on the group ( remove them manually with txt editor )
3 ) the group faces are too close together which causes a " bleed over " effect

Sometimes these things don't show themselves until the flight testing stage and the mesh is running in Orbiter.

 

[liber]

1 ) Group sides were reversed

Right about that...with Anim8or go on single object...mean ungroup if thay are in group...than select object go on Edit in Anim8or and Flip Normal...this should revers back and front faces of object.

 

[Zatnikitelman]

Try selecting the object, then use the fix-normals command. I have no idea what it does, but it "looks" like a normals problem (even if it isn't) so it might work.

 

[martins]

Before spashing out on meshing software, make sure this isn't just a z-tearing problem.

Depth resolution is performed by z-buffering, where the depth of each pixel is compared to the current buffer entry before drawing. The problem is that z-buffers have a limited resolution (16, 24 or 32 bits integer), leading to a discrete distribution of depth levels. As a result, two very close parallel surfaces may end up in the same "z-bin", making it impossible for the renderer to establish the correct coverage. This typically results in artefacts similar to the ones in your picture.

The severity of the problem is affected by different parameters: the z-buffer bit depth, the total dynamic range of the viewed volume, the distance of the objects, and how close the surfaces are to each other.

Try both 16 and 32-bit video modes. If the problem is more severe in 16-bit mode, then it is probably a z-tearing problem. Another indicator is if the problem disappears if you move the camera closer (although orbiter's dynamic range adjustment may to some extent compensate this effect).

If you are confident that the mesh is ok otherwise (e.g. doesn't contain superfluous surfaces, such as surfaces pointing into the interior of closed, invisible volumes) then the only way to fix this problem is by avoiding close parallel surfaces. This could mean cutting window holes into the hull instead of just placing a window pane on top of the hull surface.

Some renderers also have a provision for "z-biasing" which allow to prioritise surfaces relative to each other, but this is a bit of a hack and can lead to other artefacts. Z-biasing isn't currently exposed to the mesh format.