Because our project adopts CRIWare as our audio solution, we have 3 choices to integrate facial animation and audio system:
1. Use FaceFX to drive the audio
2. Extends the FaceFX Audio Interface to have the audio functionality for CRIWare
3. Ignore the FaceFX's audio functionality, use CRIWare's original function instead
Our team would like to know if we choose option 3, how do we let the FaceFX animation have the same start, end, and duration as the original audio file?
For instance, the original audio's duration is 2.793s, after creating a new FaceFX animation with the audio, the detail of the FaceFX Anim files shows its properties:
Our team would like to know if there are documents or approaches to achieve this goal?
The FaceFX animation extents are defined by the the first and last keys. Time 0 is the time that the audio file starts, so any negative keys are supposed to happen before the audio starts. You can read more about negative keys here:
A couple of things to try:
1) Pad your audio with silence at the start and the end so that there is always time for the animation to ramp in & ramp out during the silence
2) Try using the NoNegativeKeys.facefx analysis actor when you analyze. This prevents negative keys in your gesture curves (blinks, head rotations, etc.) . It won't fix the curves in your mapping, but the gesture curves can take the longest to ramp in.
3) Integrate your audio playback such that the facial animation starts the audio playback at time=0.