Okay here is a question, why is it, when you bring a model into UV mapper, weld the verticies then split the model, you can save everything, but when you open it up again, you can actually weld it and the same exact number that was removed the first time is now removed again. Then I split it and it goes on and on and on.... When you split the vertices do you create more? Sharen A never ending saga...Sharen

Also, when you weld things in Poser, then export you can get a completely different bunch of verticies, saved, looks like the size grows twice as much, then bring it into UV and split vertices your file is so much smaller then when you did it the first time. Sharen

See I am confused and thus, not making sense.

Hmmm, This is a guess. First of all the phrase "split verticies" implies that the number of vericies will be doubled, I guess I always imagined it being a function much like beveling, i.e. if you have a 4 sided poly, and split the verticies, each corner will become 2 verticies instead of 1, with additional edges to conect the new verticies to one another, so in this case what was one poly has now become five, albeit the new ones are to narrow for you to see. If the new verticies occupy the same coordinates as the old (and why not) they are eligable for welding, hence your never ending cycle. -Gort

when you split vertices, you are physically separating the mesh vertices at the points which exceed the angle you specify..so yes, this does add vertices. However, it will only add vertices to the areas which exceed the angle you specify to split...so this mean you can weld and split all day long and every time you weld it will remove the vertices created by splitting. Usually the default angle is fine, but sometimes you may want to change it (depending on the model). Splitting is done to force hard corners in your mesh so that poser wont smooth the crap out of the areas you dont want smoothed. welding when you first create your model will fix the areas where your modeling program may not export efficiently in the area of "in line" polys (ie vertices that are below the angle you wish to split). Many programs suck quite badly at exporting obj models efficiently. Ive never been clear on the poser weld mesh export option. I never check it so cant help you there. :) BT

Sharen, Splitting a model will create more vertices. If you have a simple welded cube, with a vertex at each corner, it will have eight vertices. If you now split this cube, it will have twenty four vertices, one at each corner of each face. This is sometimes not obvious, because the model will look just the same in wireframe mode in Poser. A welded model will look different to a split model when rendered, because Poser tries to smooth across welded edges. Not all programs work to the same precision, in terms of the number of decimal places with which vertex coordinates are stored in exported files. I do not have UVmapper or Poser here to check (at work), but you may find that some of the changes in file size that you are seeing is due to the change in precision of the vertex coordinates stored in the files. Hope this makes sense, and is useful.

hmmm, well I assumed all this, but I messed around with this and between splitting verticies and welding them between Poser and UV I was able to reduce one file quite nicely, but back and forth.....I went....Sharen

Attached Link: http://www.uvmapper.com/uvmpedge.html

Hi Sharen, Your way for create a model for Poser is right: - Weld vertices - Split vertices - Save a model without normal faces informations Splitting model create more vertices... After splitting don't weld the model again! See the link for understand how split vertices work and how works for Poser model. Hope help you, Mav :)

Thank you Maveris...it does help....Sharen

Although I have read this tut over an over again, my question now, regarding this tut....at the end it says, And there you have it - a model with both smooth and hard edges in Poser without requiring any additional facets =) If we split the verticies to the entire model at first, why is it considered that we have smooth and hard edges? Sharen

From reading this tutorial, the model is first welded throughout, and then split, with a splitting angle of 43 degrees. This will split the mode around the edges of the indentations, which are coloured black in the final image. The final model has a mixture of welded and unwelded vertices, and therefore a mixture of smooth and sharp (hard) edges, stemming from the 43 degree splitting angle. Whilst the final model has no additional facets, it does have additional vertices, compare to the fully welded version.

Oops, mode = model

Well imagine a stored cube. That has 6 faces, 12 triangles and 8 vertices. Now to store each edge seperately takes 4 verts to describe each face so that comes out to 24 verts total. But since the 24 verts are the same 8 verts used over and over the verts are shared. Internally the mesh is stored with a vertex table (the 8 vertices) and a triangle table containing indices into the vertex table to describe each triangle. This way only 8 verts are stored but 24 can be referenced since the indices in each triangle can point to the same verts over and over. Vert splitting creates verts by itterating through each triangle and creating a unique vertex and index for each triangle vert. So your box stored as 8 verts and 12 triangles becomes 24 verts and 12 triangles. Now smoothing occurs across triangles that share the same vertices. Its assumed these triangles are connected. A ray called a normal is created for each triangle. A normal is a ray of length 1 that is perpendicular to the surface of the triangle. What smoothing does is create an averaged normal across adjacent traingles and uses that for lighting calculations. Now a system can theoretically find all vertices occupying the same spot and smooth those as well but its usually assumed that any no adjacent vertex is not to be smoothed. This is the difference between a soft and hard edge. A hard edge is one in which the vertices are not shared between triangles. A soft edge is one in which its vertices are shared.

Sharen - Welding then splitting then welding again inside UVM (as long as settings remain the same) just performs the same split/undo-split operation over and over. On the other hand, Poser can weld but does not have a corresponding split - so if you weld in Poser and want to split again, you must export to UVM (I know you know this, but for the sake of others it is here). When you export from Poser you will notice that the file size is LARGER with welded vertices than when imported from UVM with split vertices. This seems contrary to the notion that welding should reduce the vertex count and splitting should increase it (which are both true). But looking at the file size alone does not paint the full picture, because there is a lot more in an OBJ file than just vertex information. In this case, I believe the difference in file size is due to the fact that when you export from UVM, you do so without including normals (which we all know are ignored by Poser anyway). The funny thing is that when Poser exports an OBJ that had NO normals when imported, the normals are re-created and inserted into the file (as Vn lines) by Poser - therefore the larger file size due to the presence of normals. As far as I know there is no way to export from Poser without normals (as you can in UVM), so they are always added to the file even though Poser doesn't need them. For this reason, it is best to use the UVM exported OBJ rather than the Poser exported OBJ as the file that is placed in the Geometries library because the UVM version without normals will always be smaller than the Poser version. At least that's what seems to work best for me ;-)