BACK TO TEACHING HOMEPAGE

BACK TO COURSE HOMEPAGE

   
     

CA515: Facial Animation Studies

This course focuses on analysis of the animation cycle, traditional animation techniques, and the study of motion and experimental animation. Students will focus on facial animation and gesture techniques. Specifically, students will build self-portraits, record audio, and perform animation tests with rigged busts of their own heads.

 

 
     
JUMP TO A SPECIFIC WEEK:

 

Week 9: Lip Sync using Voice-o-Matic (VOM). Blocking Pass of Expressions

Voice-o-Matic (VOM) is a Maya plugin built by Di-o-Matic that automatically creates lipsync animation baked to KEYFRAMES (not everyframe) which can then be tweaked in any version of Maya, as VOM leaves no trace of itself in the saved file. While the animation may not be perfect, the placement of the keys is a HUGE headstart and timesaver for any artist. This week's demo will look at how to use this plug effectively.

First up, we need to install VOM if it is not already installed. When prompted to find the location of the Maya install folder, locate to the "bin" directory in the version of maya you choose to use, here I am using Maya 2009:

Note that if you are using VOM on Windows Vista, you may have to run maya in adminstrator mode for the plugin to function correctly.

To begin with, lets set our animtion time preference to 30fps, as that was the recording speed of the audio we have previously generated:

Here is a look at the Rig being used in this test. The Soldier Head has a sliderboard with the following Viseme shapes:

ah, ay, ee, oo, ch, fv, m, th, luh, r, p, s

As well as many other controls for expression that will be used in the next pass of the blocking animation after lip sync.

Next up, we need to load the VOM plugin. You can find this under Windows/Settings and Preferences/Plugin Manager.
Set VoiceOMatic.mll to loaded:

You should then see a VOM menu in the animation Menu set. Choose Open Voice-O-Matic to bring up the following dialogue box. The box is numbered in 4 keys parts with icons at the top.
In stage 1---- we import the audio file at match the scene duration to the audio clip
In stage 2---- we associate the Visemes we have created with possible phonemes VOM can detect. The we set up controls from the Viseme link to our actual Viseme animation curves
In stage 3---- we process the lip sync itself, choosing how much anticipation and smoothing should be calculated into the sync
In stage 4---- we set the timeline keyframes

 

 


Step 1---- Importing the Audio
Click the open icon in the step 1 box and select your .wav audio file. Say yes to "Do you want to set this sound as the active one"

Also say yes to this second option:

And then Blamo--- your audio is set in the timeline.

 


Step 2a---- Mapping the Phonemes to Visemes
This is arguably the longest step. Click on the Phonemes Mapping button. You will see a list of Visemes on the left (these are the defaults) and highlighted Phonemes on the right. Clicking on the phoneme buttons plays an audio recording of what that sound should sound like (aka, what VOM is listening for).

I recommend starting from scratch here. Click the "X" icon to delete out all of the Visemes, then recreate your own using the names you have to govern your own sliderbox edit rig controls. In my case these names will be: ah, ay, ee, oo, ch, fv, m, th, luh, r, p, s

You can create a new Viseme by clicking on the furthest left "blank white paper" icon. Name it appropriately and hit OK.

 

Now that we have created each of our Visemes, we need to associate the sounds that they will be controlled by. Select a Viseme on the left and then the phonemes you desire on the right.

Here is the complete mapping of each of my phoenmes:

 

 


Step 2b---- Mapping the Visemes to the animation curves

Now that we have the phonemes hooked up the visemes in VOM, we need to hook the Visemes up to our animation controllers in Maya.

---To do this, Select a control curve (in this case the "ch" curve) and hit "s" or set a key in the channel box to to set an animtion key on the curve.
Note that in my scene all channels for the animation curve are locked besides translate Y, so hitting "s" is essentially the same thing.

---Next, select the curve and open up the hypergraph input and output connections window. Choose the animation curve that gets attached to the
controller curve. We need to set a key to generate this animation curve, and we need the animation curve in place because this is what VOM will
map too and essentially re-key.

-with the animation curve selected (in this case "ch_translateY") in the hypergraph and the viseme selected i the VOM window (in this case "CH")
click the red checkbox to add the animation curve to the viseme mapping.

-Repeat for each Viseme

-Click the Disk icon and save out your mapping file. Luckily you wont have to map this again!

 

 

 


Step 3---- Process the lip sync

Now for the easy part, decide how much smoothing and anticipation you want, and hit "process lipsync". I liked values of 0 and 0, which was a little robotic, but also tweakable later.

 

 


Step 4---- Set the Keyframes for animation

Note that each of my setup controllers goes from a value of 0 to .8 on the translateY


As a result, I simply set my values range from 0.0 to .8 and my keys will be set within this range.
I have found that setting your animation curve settings to "Plateau" gives the best results, but of course you could also use
"Spline", "Stepped", "Clamped" or any other.
When done, just hit Create Keys and BAM your keyframes are set.

 

These keyframes create tweakble animation curves in maya that can be hand edited to your needs.

Here are two videos where you can see VOM in action:

 

 

 

 

 

BACK TO TOP

JUMP TO A SPECIFIC WEEK: