WIP Lip Synch Osipa Simple Setup


#1

Here is a little progress on some facial animation. i have only animated the lips and jaw, because i haven’t finished rigging the face…i got tired of techie stuff so i decided to give some animation a shot. this is the best i have personally ever done on facial animating and i must admit it has been very easy using this type of setup. i am very excited about the results so far. this is only the second pass…basically a wide/narrow pass and an open/close pass. any comments are welcome!

http://www.blacksheepnc.com/video/Pass2ShopSmartWEB.mov

thanks,
kungfudork


#2

Hey, this is cool! :thumbsup:

Sincerely,

Carl Raillard


#3

really good no crits here!


#4

Can you shine any light on how to rig the null to the mouthshape? I have been trying to get it to work, but expressions in A:M clashes with my brain :shrug:


#5

thanks for the comments!

emmerzeelpictures: i can’t recall offhand exactly how i did it right now, but i am working on another character. so, as soon as i refresh my memory i will post step by step here. hopefully it will help you out.

-kungfudork


#6

thank you :thumbsup:


#7

looking good


#8

GAHHH! No! I hate this type of setup. I just got through dealing with it on a contract and I fought with it at every turn. Sliders are much more intuative, easier to track and can more easily be built/rebuilt in a pinch.

Remember. . . SIMPLE IS GOOD!


#9

Yeah, simple is good… so use 2d sliders where it makes sense. Some parts of the face move in 2d, like the corners of the mouth for example. Surely it makes sense to have 2d sliders for that.

The slider could control a pose which could be edited easily for re-building surely.


#10

wegg: i can go back and build and rebuild the sliders, they are all still there under poseslider window. the interface only connects some of the sliders to work together, but i can always fine tune animation with the individual sliders. to me this is very simple and intuitive(sp?). sorry you had a tough time.

here’s an update of another pass:
lipsynchpass2(quicktime 645Kb)

-kungfudork


#11

Wegg: What I like about the osipa setup is that if you buy into his simplification of lip synch, then the way that the null drives the open/close and wide/narrow sliders IS simple. You’ve only got this one little doodad to control so many mouth shapes. Once you start using it it gets to be as simple as the old phoneme sliders because you start to remember the position of the nulls to get a certain shape. Give it a second look Wegg! I think you’ll like it!


#12

Well I have been doing this a LONG time and have personally lip synched hours of facial animation to a very high quality. This type of setup is fine for “muppet” like facial animation but for what I was trying to do. . . I fought with it tooth and nail. Also . . . I love the freedom of throwing in my own shapes whenever it is called for on key points. This took that control away from me and left me limited to the ranges perscribed by the guy who set up the rig as anything I tried to add just didn’t blend in. I fail to see the benefit. But I also hate the auto head leveling. . . and auto hip balance and all the other “auto” “make your life easier” gizmoes that a lot of technically minded people seem to gravitate towards. For me. . . if I can’t recreate the entire rig in an hour or so and know where everything goes. . . then it is either too complex or your software sucks. A:M rules. . .

If you were the TD on my project. . . the FIRST thing I would say would be “Thats great. . . how do I turn it off?” <shrug>


#13

Just some thoughts.

The head movement does not really seem to relate to the words. Its sort of moves around but is not emphasising the dialogue at any point. I would try and lose some of the movement, and only hit a couple (or even just one) of key points. The most obvious one to my mind is the ‘smart’ word. Hit that one, and then perhaps the s in s-mart. You have a little sideways jiggle on the second ‘shop’ which is a also a little late and should be more of an up down action (if you leave it in…) Even that is a little predictable and might not work so well, but it gives you a starting point. The head movement is also very linear and computery. Try offsetting some of the channels so that they peak at different points. Eg. Shift the z-channel back a couple of frames. That should get rid of that roboticky sort of feel a little.

The sync is ok so far, but really only works for me when he says ‘shop s-mart’. The earlier stuff doesn’t quite hit the mark. The lips also seem to bounce a bit on the first ‘shop’? I’d go back and rethink the initial sync, as it doesn’t hit the ‘notes’ so to speak particularly well to my mind.

Also, the blink is sort of just stuck in there. Blinks are always done with a purpose (in animation anyway). I’d move it so that it coincided with a head movement, or maybe even just speed it up. Its a fairly slow blink. I’d crack it out a bit quicker.

This is all just my opinion btw way. Use it or not as you want.

In any case I hope it’s useful.

Cheers

Shaun


#14

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.