an0malaus posted at 12:41PM Tue, 20 August 2019 - #4359894

Given that I've explored the DSON .duf & .dsf file formats and have a reasonable python parser script, it occurs to me to question whether FBX format contains sufficient information to transfer inherent characteristics of the source application. E.g. DS uses a different subdivision algorithm to Poser, is the algorithm encoded in FBX? Unless it is, Poser won't be able to regenerate the same subdivision mesh, unless the subdivision vertices are exported in FBX too, which removes the point of subdivision. The HD morphs of DS are encrypted (to prevent piracy, one presumes) with no documentation of the format. Does FBX expose those HD morph deltas? if it does, then the encryption is pointless. If not, then there's yet another source feature which cannot be replicated via FBX transfer to another software platform. Rigging conversions from dual quaternion to tri-ax for Poser compatibility must be performed before FBX has any possibility of providing a meaningful transfer mechanism into Poser, but be aware, that such conversion will not result in the same appearance of bent joints, because the internal algorithms are inherently different and cannot be communicated by simple data transfer.

In addition to the above-mentioned, the Genesis figures also have "auto fit" and "auto follow" features that probably wouldn't transfer over via FBX. It's more than just rigging and morphs that would have to carry over for 100% Genesis compatibility in Poser.