Problem with UV map from Wings3D to Blender

[Ivanova Shostakovich]

   I use Wings 3D for modeling my creations and making their UV maps. I'm learning Blender but I still find modeling in Wings to be more intuitive. I move my creations to Blender for the texture baking.

   A periodic problem I have encountered is when I want to apply textures to the models in Blender for baking. When I select all faces for the material in the UV editor, one vertex will sometimes be displaced at coordinates 0,0. This is not really a problem per se because I can just grab the vertex and put it where it is supposed to go. But sometimes the vertex will be common to two different UV parts in the map. I don't know how to fix this (yet) in Blender.

   I use Wings 1.4.1 and Blender 2.59.

   Does anyone know why this happens or what to do about it?

[Nacy Nightfire]

I'm not sure I completely understand your question, but I'll point out a feature in the UV window that allows you to select  (or NOT select) shared vertices or shared location:

 

[Ivanova Shostakovich]

   In the image below, vertices from two separate islands are merged when I select the material in UV editing. In different tests, the vertex in question is vertex 0 in the mesh in Wings. I get the UV map all nicely laid out in Wings but Blender merges that common mesh vertex, thereby connecting the two islands that are supposed to be separate.

   I checked out the shared vertex feature. It seems to govern how you can automatically select shared vertices. I'd like to know if it is possible in Blender to get that merged/shared vertex properly separated so that the two islands are mapped the way they were in Wings.

 

 

 

[Drongle McMahon]

Not very efficient. but you can do it by (1) I face select mode, select one of the faces that uses this vertex. (2) Use the menu Select->Unlink Selection. (3) Press G and move it a bit - the vertex will be separated. (4) In vertex select mode, move the other vertices back to where they were before, and for each (5) use B to select them and the separated one underneath and (6) W to weld them back together. Now you hve everything restored except the problem vertex which remains separated. Put the two versions of it in the right places.

There musy be a better way!

[Nacy Nightfire]

What I recommend is:

Set your uv to "sync" mode (press the box with the arrow at the bottom UV menu).  Select the merged vertice.

Bring your cursor back into the UV Edit window where the selected vert is and press CTL V and click (don't move anything just click to seperate it).

Go back into the UV space and move the split vert to each island where you want it.

DESELECT "sync" mode in UV so nothing is selected in UV space.  Important or it will remerge.

Return to the split vert in the UV edit window.  Box select the vert in wireframe mode it to select each instance of the vert - it will be split but in the same location.  Finally press CTL M to remerge the vert.

 

[Ivanova Shostakovich]

   Drongle, that worked for me. It didn't take terribly long. Now I know about the unlink tool.

   I got a script that displays vertex indices in Blender. The indices assigned in Wings are preserved when I open a dae in Blender. Thinking back, I suspect that every time I have seen this problem, vertex 0 was the culprit. This project I am working on has two meshes combined into one dae. The UV of the second mesh was unaffected by this problem. It seems as though Blender doesn't like vertex zero or something. But Blender vertex indices start at zero natively. I don't know why this happens. Rather than try to figure it out, perhaps I'll just do all my UV unwrapping/mapping in Blender.

   [Edit]  I'll try that too Nacy. [Edit]

   Thank you Nacy and Drongle for your help.

[Gaia Clary]

---

Ivanova Shostakovich wrote:

   The indices assigned in Wings are preserved when I open a dae in Blender. Thinking back, I suspect that every time I have seen this problem, vertex 0 was the culprit. This project I am working on has two meshes combined into one dae. The UV of the second mesh was unaffected by this problem. It seems as though Blender doesn't like vertex zero or something.

---

Hi, Ivanova. Please note that the collada support has been improved pretty much since blender 2.59. The upcoming blender 2.64 will contain a lot of fixes, mostly for the Collada exporter, but also a few for the importer.

Would you mind to report that issue together with an example .dae file to the blender bug tracker ? Alternatively if you do not want to register to blender.org you can send me the .dae file and i will take care of that issue.

[Ivanova Shostakovich]

   I have since upgraded to Blender 2.63. You are right Gaia, the collada importer does a much better job. As far as reporting the bug goes, I had been using version 2.59. The bug tracker you linked me to is for 2.6x. Does that make a difference? And are they going to be interested in a bug with an older version?

   I don't like running into problems like this. However, learning can be a painful process.

[Gaia Clary]

So i presume the bug does no longer occur with blender 2.63 ? In that case i'd say its solved and i would just forget about it.