Nonmanifold

# 1 06-01-2006 , 01:12 PM

Registered User
Join Date: Dec 2005

Posts: 187

Problem---->

Nonmanifold geometry cannot be converted to a subdivision surface.
// To clean up nonmanifold vertices, use Polygons->Cleanup with the nonmanifold option. //
// Error: Command polyToSubdiv failed. Open Script Editor for details. //

this error comes up when i try convert polys to sub divs, i know its to do with one of the corners of a cycliner which i closed up (wth resize tool) so it came to a point. When i use the Nonmanifold cleanup option it allows me to convert into sub divs but the area were i close these vertices up looks like a bunch of petals and they are all seperate and crap.

I tried going to the area and selecting vertices and merging them all together but this then said i have to do a Nonmanifold cleaup, and when i do the cleanup it converts back into seperate faces.

Can somone pls explain whats goin on?

Also i have been watching the tutorials and he says to try keep like vertices with only 4 sizes or something, why is this? The vertice that keeps on giving this Nonmanifold error has about 20 lines connecting to it before i do the cleaup but after they all disjoint from it and i get about 30 more vertices.

Anyways thx
josh

Quote

# 2 06-01-2006 , 01:35 PM

Moderator
Join Date: May 2005

Location: Manchester Uk

Posts: 6,300

You need to try to keep it into quads so that when you convert it to sub d's you dont get the petally things.

Try to keep it into quads then convert, to see whats going on try messing about with a many faced poly cube (made up entirley of quads) by converting that into sub d's, then remove some edges and ver tices to create 5 and 3 sided faces and do the same to see the difference.

Quote

# 3 06-01-2006 , 04:13 PM

Registered User
Join Date: Dec 2005

Posts: 187

hmm

k i tried what u suggested n i see that when i convert it like gives a lot more side etc to play with but other than that it dont do much,
do i have to convert into sub divs cos i preferre workin in polys so i guess i could just use the smoothing thing for polys, is it smart to work in sub divs or are polys better, + i duno how i could get round the petally problem because i need them to be joined up in order for it to look like a sharp point.

Quote

# 4 06-01-2006 , 04:21 PM

Registered User
Join Date: Dec 2005

Posts: 187

These are my frog pics
1)is what it is in polys
2)With the wire frame
3)Close up of the area with the pain
4)What happens when i go to sub divs

Attached Thumbnails

Click image for larger version

Name: frogproblem.jpg
Views: 350
Size: 86.7 KB
ID: 26086

Quote

# 5 06-01-2006 , 04:23 PM

Registered User
Join Date: Dec 2005

Posts: 187

Attached Thumbnails

Click image for larger version

Name: closeupfrog.jpg
Views: 353
Size: 124.4 KB
ID: 26088

Quote

# 6 06-01-2006 , 04:25 PM

Registered User
Join Date: Dec 2005

Posts: 187

Attached Thumbnails

Click image for larger version

Name: frogproblemwires.jpg
Views: 346
Size: 114.7 KB
ID: 26089

Quote

# 7 06-01-2006 , 04:27 PM

Registered User
Join Date: Dec 2005

Posts: 187

Sry 1st wos the sub divs but far out, 2nd two are polys and this is the sub division problem close up.

Attached Thumbnails

Click image for larger version

Name: probelmcloseup.jpg
Views: 350
Size: 158.6 KB
ID: 26091

Quote

# 8 06-01-2006 , 05:04 PM

Moderator
Join Date: May 2005

Location: Manchester Uk

Posts: 6,300

if you wanted you could go exclusivley in polys, sub d's are helpful to tweek the smoothed object into exactly what you want as the smooth polygon can produce some strange results if you dont get it correct

Quote

# 9 07-01-2006 , 12:34 PM

Registered User
Join Date: Jan 2005

Location: vienna

Posts: 1,095

some important notes on cleaning up non-manifold geo. first of: know what is causing nonmanifold geometry!

look at the maya docs, there it says:

a) three or more faces share an edge.
b) two or more faces share a vertex but no edge.
c) adjacent faces have opposite normals

what does that mean?

1) it does not mean you must have all quads or tris or 5sided. you can have whatever n-gons you like (til 255 or something like this). n-gons wont cause nonmanifold geo.

2) your mesh must not be watertight but can haver holes if the topology follows the above rules.

3) in most cases it´s the normal thing. I´ve overseen it a hundred times, searching for hours n hours. so check your normals and make sure the all stick at the same side of your topology.

Quote