Hey guys,
Im currently studying at Staffs uni , and am doing a project on quadruped motion capture data.
This involved me looking into the pre production stage, writing up the paper work, actually capturing a dogs motion then applying them to a 3d mesh.
Heres a video of what i ahve achieved, any critys/comments would be greatly appreciated thanks.
[ame]
http://www.youtube.com/watch?v=bMw4ovECpbw[/ame]
Replies
For the little I can tell from that capture, it looks like almost no information has been captured for the spine or the tail. The data for the neck and head looks sporadic, even random, and while the legs clearly have the greatest level of data retained from capture, they're not moving like a dogs legs move. Look at this video from around 1 minute in, watch how the dogs back legs 'kick' the front legs forwards, I can see that your data has this to a degree, but it is 'kicking' the front leg forwards in an very stiff way, and the front legs dont seem to be bending at all as they move forwards.
[ame]http://www.youtube.com/watch?v=vIbSO5zAV_4[/ame]
It would be good if you would take a real render at 30fps or something and people could crit you further, but clearly the biggest factor here is a lack of detail in the capture, many of the bones simply haven't retained any useful information.