...
The default Lip Sync engine is based on a modified version of uLipSync. uLipSync uses “Mel-Frequency Cepstrum Coefficients (MFCC), which represent the characteristics of the human vocal tract.” uLipSync is supported on all platforms (including WebGL).
Volume-Based Lip Sync
This engine is now deprecated, although still available for use. Volume-based lip sync (MotiveUniversalLipSync) uses the volume of the audio clip to determine lip and mouth shapes. It is less sophisticated than uLipSync, but it may be a better option for non-human characters (cartoon characters, etc.) It is available on all platforms.
...