Inbetweens for Lip syncs

[jhbmw007]

Is there an easy way to manually draw inbetweens for the 10 different mouth positions, as opposed to using AS's interpolation/tweening, and still use Papagayo to do the auto-sync?

For instance, if you have a character with exagerated mouth movements, and having the program interpolate the movement from "O" to "AI" doesn't give the look you want. Or would you need to manually put these inbetweens in yourself?

[jahnocli]

Intrepolation is an automatic process where the program works out the point coordinates for a number of poses between two positions. If you want to change these intermediate poses, it would have to be a manual process. Unless I'm talking through my arse again...

[slowtiger]

It was mentioned elsewhere already that one could produce more than the required layers for lipsync. This way, after letting AS do the automatic lipsync, one would edit and add more keys into that switch layer's timeline.

[Touched]

I always use Papagayo as a base, and then tweak the lip sync in AS from there, either changing which phoneme it used or moving or deleting keyframes to get better tweening. As Slowtiger said, you could just make another switch called "O to AI" that has the interpolation you want -- Papagayo would leave that layer alone since it ignores any layers other than the ones named for the phonemes it uses -- and insert keyframes for that layer at any point you want. I'm not sure if you consider that manual or not, but it's less manual than adjusting the points of the AI or O layers themselves for any time you want it adjusted, which you can do as well. Once the keyframes are imported from Papagayo, it doesn't check the DAT file again, so you're free to make whatever adjustments you want.

[slowtiger]

Some workflow tip: after you applied the DAT file to the lipsync layer, you can duplicate it, that way you'll have a backup in case you fuck up the original layer.

[heyvern]

You could create "extra" phonemes, mouth shapes, for those extremes and add them into the papagayo lip synch.

For instance use the regular mouth shapes. Then go back to the word or words you wont to exagerate. Double click the word and insert and extra mouth shape reference or replace the standard mouth shape (which actually just corresponds to a switch layer).

So let's say you have an exaggerated "O" layer in your switch layer. It is called "BIG-O".

In papagayo just double click the word where you need an extra big "O" and type in the name of the layer; "BIG-O". When loaded into AS that layer will be used.

This isn't automatic in Papagayo but might be better to have it stored in the papagayo file so all of that type of thing is in one place.

You could always just select that alternate mouth shape in AS for the switch by hand after importing the "regular" lip synch DAT file.

-vern

[jhbmw007]

Thanks heyvern. I didn't know you could do that in papagayo. Hmm- let's say you had a voice clip that had 2 sentences (to keep it simple). In the first sentnence, you want them spoken normally. The second sentence is yelled- so you'd need different mouth positions for all the phenomes.

Would it be possible to create 2 voices for each sentence, save them as seperate dat files and then apply them both to the same switch layer (to preserve interpolation in the layer). Or would not after applying the second dat file to the switch, it would wipe out the first one?

[heyvern]

Hmm....

I don't exactly understand why you would need to have two different voices for one character...

But no it wouldn't work. Only one dat file per switch.

But wait! There's more!

You could duplicate the mouth switch and load the second sentence/voice in that one. Put in a "blank" layer for each and use that to turn them "on and off" in papagayo, insert a key for the "blank" to turn it off, or you could just do that in AS. So basically both duplicate switches would have a layer that is empty. If you set that layer in the switch the mouth disappears for that switch allowing the other one to show.

You won't be able to have interpolation between the two switches. You could have one layer that is identical for both (maybe rest?) and use that to transition.

I've done this before for head turns that have lip synch switches. So the head can turn left to right with a "fixed" mouth and the lip synch switches with a "blank" layer. Then I turned on the alternate "flipped" mouth on the other end of the turn.

-vern

[Touched]

For the second situation, where you want a normal mouth and a yelling mouth, if you don't want to mess around with the points of the mouths, you could do what I did in a previous project:

Have 2 sets of phonemes/2 switch sets. I had "happy" and "angry", you could have "normal" and "yelling". Apply the same Papagayo DAT file to both switch sets. Put those 2 sets into another switch layer which does not have any DAT file applied to it. Now you can switch between the normal and yelling mouths at any point on the timeline, and there won't be any interruption in their speech.

[heyvern]

The only draw back to that would be interpolation. With nested switches like that you couldn't interpolate between them. You couldn't go from a "normal" to an "exaggerated" with interpolation... you could of course interpolate within the sub switches.

I still like the idea of having ALL of the mouth types in one switch layer. Then you can interpolate smoothly between any of them. You could load the standard dat file and then go through in AS and spice it up by changing to exaggerated versions at key moments.

I have noticed that most commercial animation seen on television these days doesn't even use interpolation for lip sync... so... I am probably alone in that preference.

Actually most animation on television has very simplistic limited animation overall (Simpsons, Family Guy, King of the Hill, etc).


-vern

[Touched]

I prefer the all-in-one, too, for the same reason -- interpolation (and of course tidiness.) When I want to change the mouth shapes, I use point animation in the phonemes themselves.