A guy did that with Avatar. Is near insuperable.
Sure you can, with a decent video capture. I hope I can probe it with my animation. Thanks for the copyright warning, I’ll have to use something maybe much older.
This is it: eyelets and black tape.
Cool! great animation! Instead of a monologue, I thought in something with more gestures and less lipsync, with a joke or a blooper.
I think weta were using spray over masks to apply on actors in the same place each day, a stencil in the shape of actor’s face.
Final test is comming! 
An 11 seconds animation with Sintel Facerig Prototype.
Download audio here. (It belongs to 11second.com)
Comments please!
[ATTACH=CONFIG]223906[/ATTACH]
AWESOMEEEE!!! that is face animation in extreme! the results are incredible, but it seems that there is only facial motion capture, so, it does not seem to be useful for animators. Nevertheless the final face look absolutely alive ! Thanks for sharing!! 
Pretty sure they could also build a movement lib too. Just need to put subject through an obstacle course.
Milito, do you need help rendering? (my computer can be put to render a range of frames if you want)
Actually I do, I’ll send you a private message
Final test
It is an eleven second animation with Sintel characters that apply face capture tecniques and animation with face rig prototype v2.1.
I was wondering whether or not producing a “Making of” video, please, comment if you would like it.
Cool, did you have to loop the performance from your own features (assuming this is an external audio file)?
Hello 3point, what do you mean by looping the performance? If you mean hearing the same audio repeatedly until getting the final capture, yes I did jaja
Looping in the movies traditionally meant getting the actor back into a recording booth to re-record their lines. This would happen if the audio captured on the shoot went poorly. They would have to repeat their lines to precisely match the filmed performance. They would loop the playback over and over again, until they got it right.
Good job, looks amazing. How much fidelity did the curves have? Down to the frame for keys? That is one kf for each frame.
Actually what I did was to loop acting for capture(I did more than 10 captures), the audio is always the same.
About fidelity, It was fully baked, filtered and animated. So It has little corrections through the take before animating.
This is the “making of” video, it shows how the magic works. Enjoy!
Very brilliant work, I shared this thread with some friends who are animators to ask if this might help workflow later when we have a larger project 
Oh that sound great!!, I would love to hear some feedback from colleagues animators please post it after sharing with your friends. Thank you very much Craig.
The mouth is perfect. But to win the competition this would need to then record a second video where the whole head, arms and upper body move with the energy the woman is talking. At the moment seems very static. One would hope to see her arms moving much. I don’t know what they do in the 11seconds but for all to learn it would be correct they say the exact fragment of the movie so one can then go and watch what the actress did in this scene and compare with our interpretation.
And… WHY WAS NOT THIS TECHNIQUE USED IN SINTEL EEEEH??? With match tracking all is much easier! Where was Sergey doing instead contacting Blender Foundation? Playing with the bears??? (joke on the chat)
Hmm, wonder if there is a decent young Harrison Ford model laying around, would be a cool reboot for Indiana Jones.
First of all, thank you very much, I have to say about the body animation, that I Just focused on the face and decent body movements, I did not intent to exagerate the body (because that could became noise while reading the face gestures). Nevertheless the animation stuff took me just 3 afternoons.
About the take, I got that information for you, my friend:
Script: “There’s a moment, there’s always a moment, “I can do this, I can give in to this, or I can resist it.” And I don’t know when your moment was, but I bet you there was one.”
YouTube
jaja I think tracking was not ready at versión 2.5 alpha when Sintel was produced. Sergey was the lead animator, wasn’t he?