CompuServe Messages

Hi Mike, << My principle interest is in character creation, so MOUSE and KEYCAP weren't very interesting to me. >> Well then, you miss the whole point. The *principles* used in the simple keycap and mouse directly apply to some of the problems encountered in constructing characters, and other more organic shapes. Both of those meshes have surfaces constantly varying in all 3 dimensions. For example, how would you easily create the keycap in 3DS ? Note the radiused corners and concave top. << I have SKIN, and could see how you used it for the face, but bear with me for a minute — whether you used it for the arms and legs or not, >> The entire model is skinned with Skin.axp. Mapping is designed to cover some seams. For example, I *didn't have* to have the head as a seperate object, but it is. The "choker" map hides the seam though. That was the point of that example. Did you d/l the MESH ? << I could not attach them to the torso no matter how I tried. Either the boolean failed, or it made the objects "go away." This worried me — are the objects you create not completely legitimate 3DS meshes >> First of all, the Boolean approach is not ideal if you want to attatch the limbs. You need to construct the mesh with that in mind, and precisely weld vert's together. Second, there is a simple "trick" explained in my book, which is used to allow the exported lattices to be legitimate meshes FOR SKIN.AXP. If the meshes weren't legitimate, they wouldn't import at all. WRT to your "legitimate" train of thought, I guess you could say THIS TYPE OF EXPORT is "more legitimate" than most, as the vertex ordering makes sense to Skin.axp. There are several types of exports to 3DS as mentioned in the doc's. The female mesh was intended to be re-skinned, the "trick" was used, and THAT causes the problem you describe. For your application (NON-SKIN), you would export differently. For example, in the MOUSE mesh upload, note the cuts for the mouse buttons. These were trim curves in R3D, which were exported with the mouse, extruded, capped, and used as Boolean operators in 3DS with no tessellation required. << What I'D like would be a modeler that I could manipulate as I'm "watching" — incidentally, this is also the same problem with MetaBalls. >> Well, this is it. And this equally supports both types of modeling approaches: KNOWING WHAT YOU'RE DOING up-front, designing with control curves (contour lines), or starting with a rough "blank" mesh, and using the TON of deformation tools to mold your mesh. I've never seen a program with so many types of deformation tools. << How close does this modeler come to your "holy grail?" Or, to put it another way, just what advantage does this modeler have over Deform/Fit? >> I apologize in advance if this makes me sound like a smart-a**, but if you have downloaded and READ all the doc's I posted, and you can't answer that question yourself…..then you will likely have a problem using this or any other powerful spline modeler. Of course, that's where the book comes in <g>. There is no modeler on earth that explicity fits my holy grail definition. I am kinda' hoping Gary might create one though, in R5. R3D 4.xx might be close. Functionally, this is a very good modeler for most anything. Regards, BILL