Use joints instead of bony landmarks for the visualization of MoCap-Data

This means PeTrack now requires C3D-files with joint positions added in. This results in a more accurate representation of the skeleton.