Help, I have this strange mesh problem appearing in my model. Whatever is wrong also effects new spheres created in the file, and partially in files that this object is copied to.


Steps to reproduce.

1. make a new file in Moi
2. create a sphere (200 mm radius in my case) - sphere looks fine


3. open the cracked tower.3dm file (see cracked tower.zip attached to this post) in a new instance of MoI. Top of the tower has a mesh problem.


4. copy and paste the tower into the new file containing the sphere
5. create a new sphere next to the original - everything still fine
6. now create a new sphere by snapping to the centre of the tower - this one has the meshing problem.


Creating new spheres in the cracked tower file also have the meshing problems no matter where they are created.





This translates over to the meshed export as well.


This wasn't there in my original model created with the previous beta.



I have found these 'glitches' cropping up on other part in files I've worked on since the latest beta but I'm not sure what causes it. It does seem to be things that are created using spheres.

Can anyone else reproduce this?

Hi Tony, that definitely looks like a bug in the new mesher - the "pole" areas of surfaces tends to be a problematic area for calculating the surface normals, it usually needs some special attention and it looks like something is wrong with that.

Next week when I return from vacation I'll take a look at your example file and it should probably not be too difficult to fix up, thanks for posting the bug!

- Michael

Cheers Michael,

Enjoy the rest of your vacation.

Hi Tony, thanks for posting this very clear bug report and the example file for this sphere pole problem!

This was a bug with dealing with the surface normal calculation in that pole area, in certain circumstances.

The bug involved some numeric precision stuff, and actually in the current Jan-19 beta if you move this object to the world origin (reset from the custom cplane that is currently set back to the world one to get to world 0,0,0), it will actually prevent it from happening.

That's why it did not replicate when you tried drawing a sphere in a completely new file, because a completely new file does not have that custom cplane in it. So those spheres were closer to the world origin and when they were it happened to avoid the problem.


But I've fixed this up for the next v2 beta so that it will work properly in its current location and not get those weird wrinkles in that pole area.


- Michael

Thanks Michael,

I've gone back to the Nov-21 beta for the time being as that doesn't seem to have the problem.

Will you be able to have another beta out before the Nov-21 one expires on Feb 21?

Hi Tony,

> I've gone back to the Nov-21 beta for the time being
> as that doesn't seem to have the problem.

It has the same problem only in the display mesh, but not the export mesh.


> Will you be able to have another beta out before the
> Nov-21 one expires on Feb 21?

I think so.


- Michael

Thanks Michael,

>>It has the same problem only in the display mesh, but not the export mesh

Yes, I noticed that, but the target is modo so it's the export I was needing.


>> I think so.

I hope so :-)

Though, is the export mesher in V1.0 the same as that in the Nov-21 beta. Could I model in the Jan beta then open and export from V1?

Hi Tony,

> Though, is the export mesher in V1.0 the same as that
> in the Nov-21 beta.

It's almost the same, just a couple of bug fixes are the only difference such as the "Avoid smaller than" setting did not work completely properly in v1.


> Could I model in the Jan beta then open and export from V1?

Yeah I think that would most likely work.

- Michael

>> Yeah I think that would most likely work.

Cheers, still hopeful for a new one first :-)