The program ss_tess is designed to significantly speed up the loading into SoundSync of complex datasets with many deforming surfaces. It works by pre-tessellating each surface node for every keyframe in keyframe-then-node order rather than the method SoundSync uses, which is node-keyframe. Using this method, the ss_tess program improves loading performance because it does only one viewframe function per keyframe, rather than one viewframe per keyframe per node, as SoundSync does.
There is a fixed limit of 80 keyframes per surface, which translates to no more than 80 different "heads" in each SoundSync file.