Last modified by Maarten Struijk Wilbrink on 2021/11/08 11:40

Facial animation, emotions & speech

Blendshapes is a technique often used for facial animation. It is something created by the maker of the asset / model to mimic movement for things like speaking and facial expression. 

If you need your character to have semi-realistic facial movements, make sure the character model has blendshapes built-in. This is often mentioned on the page where you purchase your model (as an example, in this case it is called an ‘advanced facial rig’, or you can tell by the images with facial expressions. 

To use blendshapes, you can most easily use an asset to control this, such as SALSA. This asset includes a section called ‘LipSync’, which takes in any audioclip, and uses this to mimic speech movement. Another component is the ‘EmoteR’, which controls facial emotional expressions. The final component is ‘Eyes’, for both randomized and controlled eye-movement. 

This last section is especially important if you want your characters to appear lifelike. 

XWiki 14.10.13