We provide a huge number of options for facial animation. We can work with audio, video, multi-camera data and depth sensors. In many cases, this data is captured at the same time as body motion data, but sometimes it's not possible to capture every performer on a motion-capture stage. However, this shouldn't stop you considering great facial capture during voice-recording. The best audio studios (such as Side, London - where this clip was recorded) can supply facial capture hardware and services. This data can be rapidly processed by Cubic Motion. Here we show the draft result as a simple Maya playblast.
A common concern is head motion. Ultimately, the facial performance will be placed in a scene, which will to some extent determine the required head movement. However, facial movements are an integral part of performance, and those movements picked up during voice recording time can really add to the expressiveness of a result. Cubic Motion provides head motion data from a reference camera (free-of-charge on large projects) alongside the animation from the head-mounted camera. This motion can be used wholly or partly in the final integrated result.
This simple but effective solution ensures that the face itself is always captured with maximum integrity whilst also providing head motion. We believe this is far superior to methods which try to capture facial detail and head pose from a single device, where the integrity of facial capture is likely to be compromised by pose changes.
Shout out to 3lateral - creators of the "Lydia" rig shown in this playblast.