Common workflow for facial rigging a character?

johnnycore

Hey guys,

Im currently working on a character which I plan to sculpt, texture, (muscle) rigg, animate, and render.

It will be a detailed model of Sylvestre Stallone as Rambo, when its finished I will share it to the houdini community fully textured and muscle rigged.

Ive modeled a base model in Houdini and I've send it to 3D Coat for sculpting and texturing.. and while I was working on the head I was already thinking ahead a few steps on what would be optimal for facial rigging..

After looking around on the internet I'm not sure what I should do with facial rigging, I see some people are using blendshapes, some people use bones, but what I usually see is the head being detached from the body.. is this a common workflow?

My idea would be rigging the whole body including the face.. when I take a look on the internet I dont see many people are doing it this way.

Can anyone fill me up on this subject?

Thanks in advance!

rafaels · Posted: Thu Dec 22, 2011 4:59 pm GMT Post subject:

Hi,

I believe you're doing the right thing. Can't imagine any impediment to rigging head together with the whole body. I think that bones or curves to rig a head are the most flexible workflow, unless you know for sure the kinds of expressions you want to make and can sculpt them all to blendshape... The only thing is that you should probably include toggles to hide the head rig since it could be a lot o visual clutter while animating the body.

Cheers

edward · Posted: Sat Dec 24, 2011 3:33 pm GMT Post subject:

I think one should able to separate the head into a separate object (after weighting) so that you can do your facial rig without causing extra work to be done to the rest of the body.

rafaels · Posted: Sat Dec 24, 2011 3:37 pm GMT Post subject:

I see... I don't usually separate, unless the model is already like that.
What would be the ideal way to "plug" it back onto the body then, in houdini?

edward · Posted: Sat Dec 24, 2011 3:42 pm GMT Post subject:

You don't. Keep it separated as two objects, each one with their own Deform SOPs.

EDIT: The drawback of this method is that it complicates the lighting pipeline.

rafaels · Posted: Sat Dec 24, 2011 3:51 pm GMT Post subject:

It must not complicate enough then otherwise it wouldn't be an option, I guess... Smile

grayOlorin · Posted: Mon Dec 26, 2011 10:40 am GMT Post subject:

Are.the lighting pipeline complications dueto a norml seam around the neck? That could be addressed with an attribute transfer on boundary points right?

johnnycore · Posted: Mon Dec 26, 2011 3:23 pm GMT Post subject:

Hey guys,

Thanks for your responses, I think I might try the two deform SOP's, my only question about this is what kind of complications would I occur inside the lighting pipeline?

phrenzy84 · Houdini Adept Joined: 02 Aug 2005 Posts: 203

This is the first question i had over at odforce.

I think keeping the model as one for the final result will be easier in the long run. But there are workflows to keep things separate in the setup process.

I have yet to experiment just yet but if you have a complete model and delete everything below the head, rig/capture the face and then the body with separate capture, transfer the combined weights onto the full model.

Again i have yet to experiment, but i want to keep the transfer of attributes down to a minimum.

But something as simple as a visibility sop to hide portions while working or even in the final rig for display purposes can work. That's is if your animatable mesh is already done.

Here is the thread, im gonna be asking all sorts of questions to do with facial rigging.

http://forums.odforce.net/index.php?/topic/14518-facial-rigging-qa/

grayOlorin · Posted: Thu Dec 29, 2011 3:30 pm GMT Post subject:

Note too that you can indirectly drive a full model with separate rigged ones via a lattice SOP with small radius. Houdini's procedural nature allows for a LOT of flexibility in this area