@Makumba666
That might surprise you but the goal of any facial mocap system is not to recreate exact expression of the actor. That would be too restrictive. Imagine instead of Gollum we would see Andy Serkis ? Or Andy Serkis in place of Caesar in "Planet of the Apes" ? That's what you would get if the facial mocap system would recreate exact expression.
A facial mocap system is meant to "map" the expressions of the actor to a character expressions which may have completely different facial muscles structure or arrangement for the same conceptual expression.
Faceware is a markerless realtime facial mocap system, it analyses in realtime the configuration of the actor face and associate certain actor expressions to certain expression codes (in iClone). If you use the Faceware Analyzer and Faceware Retargeter (the Faceware standalone applications) then you control everything, in fact you "train" the application to encode the expressions depending on the actor expressions. You are free to do this on a global expression basis or even up to single face muscle basis if you want to follow FACS for instance (the Facial Action Coding System) like Weta with there facial animation system (morph based btw).
In iClone, with the special Faceware version, this has been pre-coded for you, you just have to calibrate the actor and normally a subset of the actor expressions will be properly detected, unfortunately not to the FACS precision. So you can't change the type nor the number of expressions that are detected and encoded but you can change how each of those expressions will look on your characters by creating custom morphs, either by sculpting those directly in ZBrush or whatever, or, if you want, by creating a bones based facial rig and capture or bake the mesh after having set each bone to your liking. So ultimately you can use bones or morphs, but the final data used is a collection of morphs for efficiency reasons (a single sculpted morph may represent dozens of displaced muscles).
Whether you use the Faceware iClone plugin or the full Faceware standalone applications, one thing to understand is that there is no direct relation from a marker to a bone nor morph because Faceware is markerless, Faceware internally creates tracking points that may not be in the same place every time and may depend on each actor face so they cannot act as "identified" trackers such as "head", "right foot" or "left hand".
Mocap does not always mean using markers. It is convenient for body mocap albeit other systems exist as well, but it has been almost "proven" to be much less efficient for facial mocap, hence the use of "markerless" systems nowadays, Faceware being the leader, afaik.
--
guy rabiller | GOETZIWOOD STUDIOS
"N.O.E." (Nations Of Earth) Sci-Fi TV Show, Showrunner.
Edited
6 Years Ago by
grabiller