Having got 'experimental' on the thing, I had a bit of an idea :D
I left the h_saddle bone exactly where it was. In the skeleton, it had co-ordinates 0,0,0 when I looked at the extracted skeleton file in notepad. It would figure, then, that the animations are worked from this base-point.
I also noted that animations are immediately translated away from the model file into what translates as 'ground' in game. So...in theory...any messing about with where ground was or where the root was would screw up the stock animations and require all of them to be moved up/down.
So...I moved EVERYTHING ELSE and left the saddle bone where it was. Your animation utilities will translateth estart position of all the bnes, and apply the rotations from there, so this just means that all the other bones pivot correctly in the animations....they are just moved forwards!
As I suspected, all the 'turn' animations pivoted arounbd the saddle ( root ) bone, and this now looks much better. I still have some 'issues' to resolve overall...but this seems to have done what I wanted it to do. Everything does seem to be calculated from the relative position of the root, so leaving the root where it is and moving everything else seems to do what I need. It also means the rider/chariot join will be at the saddle point ...which was something else I was concerned over. I noticed in RTW when making silly robots that you could not safely mount the rider too far away from the 'saddle' bone, or you got 'drifting' when the thing moved. I think I have avoided that by doing it this way too.
Now....back to the grindstone![]()
Bookmarks