Help - Search - Members - Calendar
Full Version: Weighting a Face Tutorial
Hash, Inc. Forums > Technical Direction and Development (Learning Animation:Master) > A:M Tutorials and Demonstrations > Featured Tutorials > Holmes Bryant's Tutorials/Manuals
HomeSlice
This 24 page illustrated tutorial attempts to guide you through an efficient and methodical approach to applying control point weights to a face. It focuses on manually assigning weights using the Control Point Weights Dialog in AM.

Although the the example in the tutorial uses a model which was boned with the LiteFace rig, the tutorial should give you a basic workflow you can apply to any mesh or rig.


(updated 2/02/09) fixed a few errors, made the cheek and nostril weighting a little easier, and added sections for eyelids and brows.
: Click to view attachment
steve392
Thank's a lot Holmes thats a good tut and very helpfull
NancyGormezano
Oh yes - that looks excellent - THANK YOU!
johnl3d
Thanks
robcat2075
Looks good!



"you're on your own" biggrin.gif
itsjustme
Nice tutorial, Holmes!
Fuchur
QUOTE(itsjustme @ Feb 1 2009, 12:53 AM) *
Nice tutorial, Holmes!


Just a question: Why are you going to all this trouble to rig a face with bones?
What is the advantage over poses which handle cp-animations for example for the lips?

There are parts where I can see the advantages (for example the jaw and the eyes of course),
but especially the mouth will be more flexible with cp-poses?

*Fuchur*
itsjustme
QUOTE(Fuchur @ Feb 1 2009, 04:50 AM) *
QUOTE(itsjustme @ Feb 1 2009, 12:53 AM) *
Nice tutorial, Holmes!


Just a question: Why are you going to all this trouble to rig a face with bones?
What is the advantage over poses which handle cp-animations for example for the lips?

There are parts where I can see the advantages (for example the jaw and the eyes of course),
but especially the mouth will be more flexible with cp-poses?

*Fuchur*


I can't say why Holmes uses bones...I just know why I generally do. Using a bones rig makes it easy to get a consistent result and easier variability. Separating jaw movement from lip movement is easier with bones, the percentage of bone movement is easier to tweak than pushing CP's around, you don't have to worry about setting multiple keyframes so that everything arcs correctly, it's easier to break movements into smaller parts while seeing how they relate to the overall face movement, etc. If you have a muscle based face, it relies on the CP's being identical if you want to transfer poses to a different character...unless you are using a distortion box. I would also argue that a bone rigged face can be pretty flexible if done right.

That being said, no matter which way you go, it's going to be a fair amount of work to get things looking good.
HomeSlice
QUOTE(Fuchur @ Feb 1 2009, 02:50 AM) *
Just a question: Why are you going to all this trouble to rig a face with bones?
What is the advantage over poses which handle cp-animations for example for the lips?

There are parts where I can see the advantages (for example the jaw and the eyes of course),
but especially the mouth will be more flexible with cp-poses?

*Fuchur*

I've animated both ways. Personally, I prefer to tweak controls right in the same window I'm animating in, instead of having the controls in a seperate Pose Slider window, but I'm not sure that one way is always better than the other.

A bone rig is much more flexible. You can create a wider variety of expressions, even with a rig as simple as LiteFace. Also, if the controls are placed in a way that makes sense to the the animator, it is faster and easier to animate. You grab a Null right on the screen and drag it around.

A face with poses can be be faster and easier to set up, especially for a simple character, but the animator is limited only to what the poses can do. If you set up enough poses to mimic the flexibility of a bone rig, you end up having a few dozen pose sliders you have to deal with in the Pose Slider window. That gets tedious and tiring.

For example, let's say you want to set up a face with muscle poses and you want it to be as flexible as a simple bone based rig such as LiteFace (Squetch and 2008rig have more bones and are probably even more flexible). You would have to make poses for:

--------------------------------------------------------
Outer Eyebrow Left up/down
Outer Eyebrow Right up/down
Outer Eyebrows up down

Inner Eyebrow Right up/down
Inner Eyebrow Left up/down
Inner Eyebrows up/down

Inner Eyebrow Right in/out
Inner Eyebrow Left in/out
Inner Eyebrows narrow/wide

Eyelid Left Top up/down
Eyelid Left Bottom up/down

Eyelid Right Top up/down
Eyelid Right Bottom up/down

Eye Left open/close
Eye Right open/close
Both Eyes open/close

Eye Left Mad/Sad
Eye Right Mad/Sad
Eyes Mad/Sad

Cheek Left up/down
Cheek Right up/down
Cheeks up/down

Cheek Right in/out
Cheek Left in/out
Cheeks narrow/wide

Nostril Right up/down
Nostril Left up/down
Nostrils up/down

Nostril Left in/out
Nostril Right in/out
Nostrils narrow/wide

Lips Left Smile/Frown
Lips Right Smile/Frown
Lips Smile/Frown

Lips Left narrow/wide
Lips Right narrow/wide
Lips narrow/wide

Lips Left Sneer
Lips Right Sneer

Lips Top up/down
Lips bottom up/down
(these two poses are for showing the teeth and for shifting the mouth up toward the nose or down toward the chin)

Tongue in/out
Tongue up/down
Tongue left/right
Tongue Curl up/down
----------------------------------------------------------
I think that's 45 pose sliders you have to deal with. (not counting the jaw)

LiteFace has two options (Squetch and 2008rig are similar):
If both sides of the face do the same thing - 12 on-screen controls grouped into 7 groups around the face - or wherever you want to place them. (not counting the jaw)
If each side of the face does something different - 14 on-screen controls placed right on the part they control, plus 3 controls for the Tongue which are off to the side.

The controls are much faster and easier to find this way.
----------------------------
Another option is to set up your 45 muscle based poses and hook up the on-screen controls to the pose sliders. But I don't think that would save you much time. In the LiteFace rig, positioning the bones and installing the rig takes about 15 minutes (less if you work fast). The rest of the time is spent with weighting the control points. So it all comes down to the time it takes to define 45 muscle based poses versus the time it takes to weight the face. I suppose it would just be a personal preference then.
robcat2075
Almost everything on the face moves in some sort of arc and bones are great at making things move in arcs.

It's hard to make things move in arcs in SmartSkin.

I remember long ago when Victor Navone showed how he had rigged the face of his "Big Bang" character with bones. That was one of those big "aha!" moments.
John Bigboote
Great tutorial Holmes. Man! I wish I could present info as clearly as you do. One question... is your method compatible with the ZignTrack A:M/Track utility?
HomeSlice
Wow, thanks for taking the time to look at it and post responses. I have uploaded an updated version in the first post. I fixed a couple of errors, made the cheek and nostril weighting a tad bit easier and finished working through the face bones.

QUOTE
One question... is your method compatible with the ZignTrack A:M/Track utility?

I don't know. I don't have A:M Track - I don't even have a web cam. Someone else will have to test it. You perhaps?
NancyGormezano
QUOTE(HomeSlice @ Feb 1 2009, 06:53 PM) *
In the LiteFace rig, positioning the bones and installing the rig takes about 15 minutes (less if you work fast).


We folks out here in the Land of Perpetual Lazitude have only this to say: Bless you child, bless you.
Luuk Steitner
QUOTE(John Bigboote @ Feb 2 2009, 06:24 AM) *
One question... is your method compatible with the ZignTrack A:M/Track utility?

Yes, it is.
nimblepix
Thank you so much for this.
I'll be needing it soon. : )
Paul Forwood
Very well done, Holmes!
mtpeak2
QUOTE(John Bigboote @ Feb 2 2009, 12:24 AM) *
One question... is your method compatible with the ZignTrack A:M/Track utility?

The rigged face model I posted awhile back was designed for A:M Track. It also has controls that you can add additional facial motion to the action or animate the face directly.
HomeSlice
QUOTE(mtpeak2 @ Feb 2 2009, 06:03 PM) *
The rigged face model I posted awhile back was designed for A:M Track. It also has controls that you can add additional facial motion to the action or animate the face directly.


Yes, I want to plug Mark's face rig here. LiteFace is designed to have the fewest number of bones possible while still being useful. Mark's face rig is designed to give the greatest number of options and facial articulation possible. It also has some nifty automation features, like the eyelids can automatically follow the eyes, and the cheeks can automatically rise and fall according to what is happening with the rest of the face. They are complimentary products.
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.
Invision Power Board © 2001-2013 Invision Power Services, Inc.