They are from the <Animation> and <source> and <sampler>. I’m just trying to figure out why I have 3 sets of X, Y for a single time point.
Maybe I’m just confuzed by bezier curves…
Between any two time points I have a Bezier Curve Segment using the 3 given tangents as the 3 parts of the 4 control vertices that represent that curve? But then why would I need 9 total tangents to represent a 3 point curve? Shouldn’t I just need a single In tangent XY pair and a single out tangent XY pair for each of the points?
And since its in 2D how does that flow over to the 3D shape the movement represents?
Unfortunately Collada animations are in a somewhat confused state right now since they don’t match exactly what the DCC apps do. There’s a long, boring thread about this issue here. It seems that some DCC apps (Max, Maya) have chosen to start exporting animation curves in the DCC-style instead of the standard piecewise cubic curve format that’s described in the Collada book.
That’s one issue. Another issue is that Max stores independent curves for X, Y, and Z translations, but (in some cases) merges them when exporting to Collada. The six values you’re looking at for a control point are timeX, valueX, timeY, valueY, timeZ, and valueZ.