LipService 3d facial sculpting and animation plugin and standalone for lightwave. 

LipServiceTM    
(C)opyright 1999 Joseph Alter, Inc. All Rights Reserved

  So, these are 'phonemes'. Phonemes are just the basic elements of speech. You'll notice that there's 18 of them, yet there's 26 letters in the english alphabet. What gives?

Well, go ahead and spell out 'shoe'. You'll notice that the spelling isn't how the word sounds. It actually sounds like 'sh-oo'. If you were to construct that word with phonemes you would only need two. But wait, there is no 'sh' phoneme?! No, the 'ss' phoneme is visually the same as an 'sh' would be, so there's no 'sh'. Phonemes, to a speech pathologist, are the basic sounds that make up words, not spell them. Phonemes to an animator, are the basic visual shapes that make up the sounds that make up the words.

The phonemes you see here are actually 'Joey' (the fat bald guy) from 'Jersey'. You can see that the targets are very assymetrical, and not idealized perfect mouth positions. A guy from Jersey is going to form his sounds quite a bit differently than one from Texas. We'll get to sculpting later, but it's important to mention that the reason why I built modeling tools into LipService is just this, you have to have the audio track there and some animation to see how your targets are working for your character if you don't want him to look like a robot. He has to actually look like he's forming his words. In LipService, you can lay down animation before your targets are even finished being sculpted, then scrub through the animation see whats wrong, and modify the targets. This easy interplay between the modeling and keframing processes will save you alot of time/guesswork and give you much more human characters.

In LipService, creating a key is just a matter of clicking on the phoneme with the left mouse button. You will see its lable appear on the timeline (see below). Deleting it is just a matter of clicking on 'backspace'.

Creating animation is very fast. Just step through the sound with the arrow keys a frame at a time, listening for the 'eh'-'eh'-'ee'-'mm' (etc) sounds and clicking on the matching phonemes as they change.

This stepping may sound funny, but it is extremely prescise and *frame accurate*. Additionally it is much easier to pick out the phonemes than 'slo-motion' scrubbing

At any point you can roll playback in sync with the 'a','s',and 'd' keys to see what you've done.

Creating keys is not limited to phonemes, any cell (including cells in the performance groups covered later) can be keyed. Again, just click on the icon to make a key.

Notice below on this section of the timeline, there's a '1' and '2' button. The '2' means that you are actually keyframing on the secondary animation track. Making keys here keyframes things that you want added to the first timeline (secondary motion). This can be things like smiles, frowns, whatever.
The important thing here is that you are not limited to a 'one track' keyframe for targets. This is important, usually morph programs let you keyframe blends, then construct a shape out of that. LipService is not a morph program, it's actually going to make a splined motion path for each and every vertex (we'll discuss that under 'interpolation'. This allows you some of the same functionality as morphs without the crappy interpolation.

Also worth mentioning here, since I use this alot:
LipService's 'scene' files make excellent dope sheets. I frequently print them out and paste them to my walls by my workstation and draw little stick figure poses in the margins to plan out my body animations in sync with the dialogue. Dope sheets date way back to early Disney days as an effective way to plan animation and make sure everything is in sync.