Alrighty, then. I created a phoneme rig for ManuelBastioni Lab which looks like this:
As you can see each one of these dials represent the mouth shape which is formed to make the associated sound. These were taken from these references here and here.
The next step is to write a blender add-on which takes an audio clip, recognizes the phonemes, and then animates the phoneme rig shown above. Sounds simple eh? As I mentioned in earlier posts, there is a tool called rhubarb which does a similar job, but you know what, after flip-flopping on whether I should create my own or use rhubarb, I decided to create my own. And so, here is an outline of what the program will do:
