Morph Targets

[Wildsided]

This goes out to all the smarty's in the audience.

 

I found a piece of software that'll allow me to make realistic looking faces. The software's exporter will export a neutral face model and then a bunch of morph targets based on that model.

 

So when it's done you end up with Neutral.obj and Angry.obj for example

 

I can obviously import both .obj's into A:M as 2 separate models. What I'm wondering is if it's possible to make a new %pose in model 1 and import or copy over the CP positions from model 2 in order to make a pose slider that shifts from neutral at 0% to Angry at 100%.

 

The two models have the exact same spline allocations as far as I can tell but the CP's are obviously rearranged to match the face expression.

 

I've tried making a new pose for both models, going into angry, setting things control point mode, select all and copy, then try pasting into the model 1 pose but it does nothing.

 

Any ideas? As this would save a ton of work on facial animation.

 

[detbear]

Wild Sided,

 

I've been doing this for some time now.

 

ONE THING YOU MUST REMEMBER: to get the morph to translate correctly, you need multiple "Inbetweens"

 

If you just go from "Neutral" to "Frown", it will go linearly without curves. So you'll need to have more than one "Mold Plate"

 

So....here is what you want to do. Take your neutral Base Face and open the frown morph model in the same model.(so you have both models within one model.)

Then make a new pose. select the cps of the base(within the pose window) and RIGHT CLICK/ Snap to nearest cp. Then it will ask you

for a group to snap it to. (You need to have a group for the frown created)....... Choose "Frown"...

 

The cps of the base face will snap to the frown.

 

WARNING.....If the cps have to move a large distance, especially in the lip area......cps will snap to the wrong cps. THIS is why having several(or a bunch) of

plates is important so that you can snap in incriments over time.

 

Hope that helps. It does work.

[Wildsided]

It does indeed work. Thanks dude, see what you mean about the snapping to wrong CP's but at least I have a point to start from now.

[robcat2075]

That would be a good candidate for a dedicated demonstration/tutorial.

[detbear]

Wilsided.... Any area that is tight with crowded or clustered cps will be easier if you use plates for percentages of the motion.

 

For mouth movements and eyelid movevements I've actually used plates for every 2-5% of the motion. This takes forever, but since there

is very little movement, there are often no "MIS-SNAPS"......Hey, I just came up with a new CG term.

 

Robert.....That would be valuable for people to see. This is why I suggested the "snap to cp" function way back.

 

BTW....I see where 18L has added "Bone to cp snap"......That will add another dimension to this because it will accompany the group constraint. This could

potentially add a way to swap these face morphs between characters so that once you create a "Phoneme set" of morphs on one character, you can

paste them onto a second character by having bones control cps on two separate face shapes.

[Rodney]

That would be a good candidate for a dedicated demonstration/tutorial.

 

I agree!

 

I'm envisioning a pretty sweet workflow but the reality will work much better I'm sure!

[robcat2075]

 

Robert.....That would be valuable for people to see.

 

And they would love to see you demonstrate. You're the one who knows most about it.

[detbear]

I will see what I can do