I've hit a rough patch in my energy. The lip-syncing project requires significant development effort. I'm writing a program in C to do the initial lip-sync animation. However, my day job is also a Software Development job. I'm having a difficult time spending a few hours programming after having spent 10 hours doing programming work... But I just have to suck it up and get it done. I know what I'm suppose to do.
One good piece of advice I got from a couple of professionals in the rigging/animation field, is not to apply the animation directly on the model. Rather give the animator the option to select the key frames to insert. I've been thinking of a good work flow to do this.
I'm now considering, first marking the audio file with the phonemes. Then provide a way from the animator to jump to each phoneme or just use the time line to scrub to the marker. He'd then hit a button to insert the appropriate key frames. I think this would be easier to undo step by step if need be. Then the animator can fine tune the shape of the mouth, before moving to the next phoneme.
I'll have to do some trial and error to find the best workflow.