Hello,

Has that person you commissioned checked this thread https://drive.google.com/file/d/0B0ciRKrWV2vsN25yUlRJdHExd0E/view on how the new importer works?
Here is also the full documentation https://drive.google.com/file/d/0B0ciRKrWV2vsa0hXeHBDMzl3STA/view to create an avatar with the necessary specs for FaceRig.

Hi,

Could you please send another mail to the art support to identify your model and see what was the last status. I remember discussing with your commissioned artist but I don't know what was his data (model name, email, that kind of things).
I don't think making the skinning and animations at Holotech would really be a solution, since you've already paid for this service somewhere else. Rather we could try once again to guide your artist in Facerig specific stuff. But please note that there are several artists across the globe that did import very complex models in Facerig using only provided documentation and the example model. So we know it's possible to accomplish this goal :-)

Hey, I'm the other artist mentioned with the same problem. We're both using maya 2014 to make the model and export the collada files, and I've spoken to other artists creating avatars in Blender that have similar issues. The consensus seems to be that the software used to do the final collada export matters a great deal - artists who create avatars using 3ds max have little difficulty, since that's what's used in-house at holotech and, presumably, what the facerig importer has actually been tested on. Collada as a format is kind of an unsupported mess, so different applications will spit out collada files with different idiosyncracies and not all of them play nice with the facerig importer.

So executive summary - if you want to build avatars for facerig, use 3ds max.

Hi root,

It's true, Collada is a messy format, however because it's free and was supported by Alias, Autodesk and Sony it gained some traction and was available in many widely used 3D applications. That's why it was chosen by our programmers as the bridge format for Facerig.
And is also true that most of the import problems came from Blender users, it seems that Blender collada export writes data in a way that is not correctly interpreted by Facerig collada importer.
In Holotech we are using both 3dsmax and Maya and we are bringing data successfully from both of them. But these two software have different transform axis and units set-ups so these differences must be taken in account when working with them. But really, making some simple test on very simple data would quickly give an outline of how the models should be prepared for Facerig. Keeping things simple at first would ease identifying the issues and find the appropriate set-up, like axis orientation and absolute model dimensions.
There are users that not only succeed to import their avatars in Facerig, but managed to make feature-full avatars, maybe with the exception of animated normals feature. And those users succeeded using guidance from the documentation alone.