CompuServe Messages

#Skin.axp

    21-Feb-94 00:30:29
Sb: #85216-#Skin.axp
Fm: Yost Group 76702,413
To: M. G. BATCHELOR 71532,1214
Answers: 1) It appears from your Skinning.gif that the .axp takes cross-sections or 3D definitions in their CONTROL HULL form, which I assume are created conventionally in the 2D Shaper & Lofter, etc.. GY: Actually, we're working on a PXP that derives mesh-based cross-sections from lofted models, because that's the data structure you need to use in the keyframer. Q. Is there any provision for creating or displaying them in their true curve form before creating the patch ? GY: Nope. Just cross-sections before render-time. 2) By any chance were all the various elements all skinned in one pass ? GY: They were all skinned at once by five instances of skin.axp. 1 on the body, 2 on the horns, and 2 on the fangs. 3) Were the Ortho Sweep type objects (e.g.-horns) concatenated with the CS – Swept object (body) ? I don't think so here, but can they be ? GY: They were just part of the hiearchy. 4) If a non-patch object was heirarchially attatched to a DEFORMING (as opposed to TRANSFORMING) patch object, would it still maintain proper position and orientation as well ? GY: Yup. 5) Is there a provision to reparameterize / remap the patch up or down once it's created ? (If not, please strongly consider) GY: You can change patch parameters at any time. 6) Once the patch object(s) is constructed, can it be "seen" & manipulated with conventional Bend, Taper, Squash, etc.. or are new functions coming ? GY: Sure they can. They're just mesh objects. 7) The Skinning.fli APPEARS to be parametrically rendered, although I doubt it for various reasons – therefore I ASSUME conversion to a zillion polys at render time. Q. Is this true, and if so, can I assume traditionally long render times for a zillion polys ? GY: Yup, like Alias, it makes triangles at render-time. Note that those frames took 4 minutes per frame on my 486/50 at 320×200. – G