CompuServe Messages

>> I'm often working from photographs, mainly because the people I'm modeling are dead or uncooperative or both << I'd like to see some examples of your work, but I guess I'll have to be patient like the rest. Brandon and I are involved in a users group for 3D artists and animators in LA (which covers all 3D programs, not just 3DS), called the 3D ART FORUM, and I think it might be interesting to show your work on a future program. Something to bare in mind. Our next meeting is March 4th, and will include Babylon 5 FX. Wavefront, Viewpoint, and other stuff. We're expecting at least 150 people. I'll probably be posting more details soon. Are you scanning the images and then loading them as a backdrop behind the SKIN modeler? This is a method I think all modelers should be aware of, but for myself I prefer to use drawings and grids. I don't think I'd be able to isolate enough detail in the backdrop images unless I was working with an interface at about 2048 X1536 resolution. Does 3DS allow you to work with backdrops this size and then zoom in for more detail? I still haven't delved deeply enough into the 3DS to know this stuff. In addition, I tend to rework the drawings as I build the geometry, so I'd have to keep rescanning and reloading the backdrops. Sometimes I draw out the actual shapes of the spline patches on paper, not just the anatomy. Then the job becomes almost like manual digitizing. I got very frustrated using lofted or skinned objects, and then trying to boolean them together. I found myself spending a heck of a lot more time re-editing points with this method, than if I had simply drawn out all the crucial geometry before hand, and build it up point by point, patch by patch. It might seem a little counter-intuitive, and for many jobs this approach would be much too time-consuming, but… well… there you have it. After my experiences with Humanoid, I began to feel that it was unnatural to force biomorphic shapes into a cubic matrix, which is essentially what you are doing when you model a figure with cross-sections. Instead, I wanted an approach that would allow me to create a surface by following the natural contours of a figure. So far, for myself, I find AM's splines easiest for this. But the same method can be used effectively with polygonal modelers, now that there are good smooth tesselation/optimization routines in many programs. For example, even though LW has splines, most LW users I know don't use it when they're trying to do highly controlled modeling because its too sluggish. Instead they often build their models with quadrilateral polys, and then Metaform them (which is similiar to using smooth tessellation). (Oops, I said the L-word) Let me know if I'm rambling. I don't know how much of this stuff you care to hear. (BTW, Brandon did you every come up with a way to do the filleting that you wanted with BlobScuplture or some other program? Have you tried MetaReyes?) – Tim