I know we have discussed this before (we, as in me and some of the forum members), but I can't remember where and I still wonder if there is a chance I can improve it. When looking at my Great Goblin model, it has proven to be the one that is obviously the hardest to rig without any major distortions in anims. I suppose this is due to the fact that it was the first full body model I made, and I made it from scratch, as in not using an exported W3D model upon which I sculpt and thus keep the proportions. Great Goblin has been rigged several times, by a number of experienced modelers such as Ridder Geel, Ealendril and now Valheru - but it's still not as good as the other rigs. If we compare to my Yazneg model (rigged by Ealendril) and Fimbul (rigged by Valheru), they outshine Great Goblin by far in terms of fluid movements without major polygon errors, and the only big difference is that I sculpted those models using the generic orc mesh from vanilla game...
So, I guess it's safe to say that the proportions of the model compared to the SKL it's suppsed to be rigged to, is definitely crucial to the outcome of the anims. However, when it comes to the resolution of the model, as in polycount and the number of vertexes that should be binded to the appropriate bone of the skeleton - how much does that affect the same topic? Some say that higher resolution increases level of difficulty to rig, but also vastly improves animations (like in CGI movies), and the other way around. But others seem to believe that higher poly count only increases risk of binding errors and makes it a whole lot harder to rig all parts of the model to the best corresponding bone. So what goes? And do you think the riggers would succeed better if the Great Goblin was remodeled into something closer to for instance the generic troll, than this?
In short, is the proportions all of the problem, or just some of it and how important (good or bad) is the resolution (polycount)?
Edited by Nazgûl, 26 April 2015 - 10:18 PM.