View Single Post
Basically everything is correct - Some more hints.
Old
DrJim
Guest
 
Status:
Posts: n/a
Thumbs up Basically everything is correct - Some more hints. - 05-16-2007, 11:08 AM

Quote:
Originally Posted by Shaba1 View Post
Ok. I got lost in most of that reply.
Can't say I blame you - I got lost rereading it - and I wrote it!

Not sure about the Blender details - I don't use it much myself so I'm not sure about the commands - but everything else sounds fine. And yes, you only have a hierarchy of meshes, not a skin/armature (or similar definition) model. As I mentioned, you can really see this around the joints of a character

Some more details that can lead to problems if ignored:

1. To get the overall objects motions easy to program, you do need to be careful where you set the local origin/pivot point of each sub-object. Usually a good location is the point at which a sub-object joins its parent object.

2. Something I always have problems with is the direction of surface normals. This is particularly a problem with some Blender models that just use surfaces with textures on both sides (for hair, for example) - but almost any object that is not completely closed (i.e., there are openings in the mesh for doors, etc.) can have this problem.

My suggestion is to stay with low polygon models (the lower, the better) so you have a chance at fixing the problem manually. Also, if you cut a mesh into two or more parts, add surfaces to close the cut areas.

Comment: This means effectively you have to ignore the subsurface modeling (or other smoothing) options of your object creation program - a very unfortunate tradeoff, since one already has the unrealistic joint problem.

3. Despite the fact you are importing an .ase model as a texture, all you are actually going to import is the mesh (there is a link to a tutorial on importing objects on the Alice main page which covers this type of thing in some detail). You will then have to import the texture for the sub-object in a separate step. Not difficult - the only problem you might have is a mismatch in scale, which can be fixed with a couple of test runs to find the correct calibration factors - but a bit confusing if you don't expect this.

Again, good luck. I'll be glad to answer questions if I can - but I suspect you will rapidly move beyond the things I've tried.
   
Reply With Quote