Question on Collada format details

Looking at man_walk.dea, the TIME arrays are listed as count="30", but have 31 entries in their bodies.  The TRANSFORM blocks are listed as count="480" (30 frames * 16 floats per transform), but are actually 496 floats long (31 frames * 16).  The imported model holds 30 frames.  Does the original model that this was exported from contain 30 or 31 keyframes?  Is this a bug in the collada exporter of wherever the model came from that the counts are off, or is there some boundary issue involving interpolation that puts an extra unit of information after the last keyframe?


That appears to be a bug of the exporter. Good catch, I'll bring it up with the tech artist and see what he can do.

frame 0 for the tool, and frame startin in 1 in code? surely not. It surely sounds funny. But we had a prob like that with max eons ago in a company. that and the morph of first with last…(or a keyframe setup in negative)