Facial motion capture in games

I decided to make a separate post blog post for facial animation in games. This is due to how differently different studios approach this process. As previously discovered with naughty dog, on their making of the Last of Us, the motion capture process was crucial to the development of the game for the most effective and accurate performance for the game. But with the facial animation, they didn't use facial motion capture, but instead they just used very precise facial references, so each actor had a camera fixed on their face to be sent to the animator who then has all the controls for the characters face in Maya. The animator scrubs through the reference and carefully keys in the animation. This is the same process I used for my extended practice animation which was fairly meticulous but I enjoyed having the control. But there are other methods that have been used.

A company called Faceware , for example, are a company that have worked of numerous feature films and top quality games titles. they provide a suite that is used to get facial animations of an actor. It is completely marker-less and uses an analyse method of the video clip of the actor. it then renders the facial animations for your character.


I think the direction of facial animation is going in quite a linier way, from the start of the innovation from the game LA noire who made a big impact due to its increased accuracy, but it also had various bugs that took from the believability, but this was a different approach where the actor was placed in a studio with cameras all around them and then that capture was used in conjunction with the characters.

I looked into a short on-line course which teaches you a cheap version of facial motion capture which I still have to look more into, but its not something I wish to pursue in my career, but I do need an understanding of the processes involved in order to seamlessly place myself in the animation pipeline without having to adjust to new methods and such.