Lip synch problem

[darkmatter]

Yeah, thanks once again for the help =) To finish the topic, i wanto to say that I eliminate some layers. I had layers for different "mouth emotions", and the only difference betwwen them was the "rest" layer. But i realized that i can change the points of the "rest" layer during animation =) I also realized that i didn't need one of the bone layers (i had bone layers facing both sides). One is enough, because I can flip it during animation. With those deleted, I passed from a 3,4mb file to a 900kb file.

http://www.mediafire.com/?3wrsjnjjy1x

Please, check it out and tell me if there is anything more unecessary to delete

[Genete]

If it is a professional project then you should simplify it as many as possible. If it is a hobbyist then GO!. You'll learn a lot with your own setup and would step by step find the proper setup for each situation.

It depends on how many complexity you want to achieve with only one folder for each character. My experience is that you should have as less layers as possible. Every kind of situation responds to a different setup.

What Vern and I wanted to advice you was that the work flow for that setup would be very hard to manage and very very confusing,
I don't want to critique your new setup. Just if you're happy with it, it would be good.

Only one suggestion that (hope) you'll like it.
To make the upper part of the legs to be connected with the hip you should connect those points to the hip bone. To do that just do following:

1) At frame 0 select the 2 upper points of your legs. There are 4 upper points. Make them peak points.
2) At frame 0 displace those points to the lower part of its T-shirt. Don't worry as well as there is a bone Off set they would go to the proper place.
3) Bind those two points to the hip bone. That's the one that make rotate all the character (the most root bone). Then when the Off set of the bones works the points would go closed to the hip bone and will be always at the same position.

This would allow you move the legs. without loose connection to the hip what looks a little weird.
If you did not understand my explanations tell me and I'll send you the file. I prefer you try it and if not ask again.
BTW the new setup is more or less OK.
-G

[darkmatter]

WHOW that solved my leg movement problem =D thanks for the tip!

And no, it's not professional. This is some kind of "old passion". I always loved animation =) So yes, it's a hobby =D

[Genete]

You're welcome!

[mooncaine]

Known issue: before you import a Papagayo .dat file, you must have your mouth Switch layer keyed with a keyframe at the beginning and end of the project timeline. I don't know if the issue happens when import a .dat file, in general, or if it's a Papa thing.

Of course you must have a "rest" or "closed" mouth layer in the Switch, and that layer should be on the bottom, under all the other layers in the switch.

If you do these things before importing the Papagayo .dat file, it should work as planned, using the "rest" layer sensibly, but it's never perfect, so plan to tweak it, similar to Vern's workflow.

Personally, I'm finding it's about as quick for me to just do the lip sync in ASP, because the Papagayo interface requires so much clicking and dragging, and sometimes gets "stuck" where I can't drag out a word that's beeen shrunken too small. Undo doesn't undo that, and at that point, the only thing to do is quit messing with it. All in all, Papa hurts my wrists more, so far.

Maybe I am trying to be too precise in Papa, and should let it work, then tweak in ASP. I'll try that next time a scene comes up with lots of talking.

[heyvern]

I like doing lip sync in Papagayo myself for a couple of reasons.

One is the "faster" playback and simple checking of the mouth shapes to the sound. You have that big bold image of the phonemes on the right.

The other is being able to double click on one word or phoneme or sentence to preview that section to make sure you got it right. In AS you can't preview into the animation because of how sound starts and stops.

I also like the scrubbing better in Papagayo.

I will do tweaks in AS afterwards.

-vern

[slowtiger]

Right now I'm working on a project with a lot of lipsync, and I do it all by hand - two reasons:

1. Papagayo on Mac doesn't work properly, so I gave it up.

2. Doing it by hand requires about 4 to 10 switch keys per second, which is done faster by hand than going though the hassle with opening another program, import, tweak, export ...

The files I work with are between 2 and 10 seconds long. The importance of correct lip sync isn't that big - it's much more important to adjust the acting to the sense of the dialogue.

[heyvern]

1. Papagayo on Mac doesn't work properly, so I gave it up.

The Mac version for papagayo works fine for me though it can be a bit slow. I use it all the time.

2. Doing it by hand requires about 4 to 10 switch keys per second, which is done faster by hand than going though the hassle with opening another program, import, tweak, export ...

The hassle you speak of is not like that for me. It is two passes on the lip sync. The first in Papagayo the second in AS. Most of the work is done in Papagayo. The final "acting" tweaks are done in AS.

The importance of correct lip sync isn't that big - it's much more important to adjust the acting to the sense of the dialogue.

There's the key difference. I need the lip sync to be as good as possible. Papagayo is still "better" (better is a relative term) at this than doing it from scratch in AS. It isn't a perfect solution but it does the job.

The playback on the Mac in AS is so out of sync with the audio there is no way to preview lip sync in AS without rendering the animation. At least in Papagayo the preview is in sync.

I find it very difficult to match the phonemes with the audio in AS. I use Papagayo to lay down a matched set of keys for the dialog that can be used as a base for tweaking in AS. To get to that point by hand in AS would take twice as long... trust me... I've tried it. Even with all the flaws of Papagayo it still saves me a ton of work.

-vern

[slowtiger]

The playback on the Mac in AS is so out of sync with the audio there is no way to preview lip sync in AS without rendering the animation

That's not my experience. It is in sync - but everything plays slower than it should, although I have the "skip frames" option ticked. Usually it plays at about 21 - 24 fps where it should do 25 fps (in my current project).

[heyvern]

The playback on the Mac in AS is so out of sync with the audio there is no way to preview lip sync in AS without rendering the animation

That's not my experience. It is in sync - but everything plays slower than it should, although I have the "skip frames" option ticked. Usually it plays at about 21 - 24 fps where it should do 25 fps (in my current project).

My point exactly. How can you preview or pin point the dialog lip sync if the playback is off... even a little?

My experience is that the audio is completely out of sync with the animation in AS. I can't rely on it unless I render it.

At least by using papagayo to get it "exact" I know if a phoneme on frame 65 is an "O" of a specific word in a specific phrase and I want to exaggerate a head nod or turn or change the shape of the mouth or widen the eyes at that spot after importing to AS, I'm not guessing. I can even use the imported dat information for timing the acting of the rest of the character without even HEARING the audio or having it in sync.

If there were some way to have the text track visible in AS that would be great. It is very hard to lip sync with out any knowledge of the word spoken and no way to see it. In Papagayo you have that visual clue right there. The word and its length relative to the wave form is clearly visible.

I can double click a specific phoneme in Papagayo and hear if it hits the precise part of the word exactly. AS can NOT do this.

To be honest, if Papagayo doesn't work for you and you don't use it, you can't really judge its benefits. I find it INVALUABLE for doing lip sync.

-vern

[Genete]

I can double click a specific phoneme in Papagayo and hear if it hits the precise part of the word exactly. AS can NOT do this.
-vern

That's true. But you can limit the playable area to a region where the word or phrase sounds with CTRL right and left click over the time line to generate the red and green squares of playable area. Using this you can repeatedly play an specific portion of animation (like an specific word) as many times you want (just press space bar) and focus only on move or insert the keyframes for the phonemes. It could help a little for manual lip sync I guess...
-G

[slowtiger]

How can you preview or pin point the dialog lip sync if the playback is off

Actually, it's not. AS stretches the audio if the playback rate is slower than normal. In the range of 4 frames off, it's audible as a kind of stuttering. A bit annoying, but still in sync. If the playback rate is real bad, the sound is pitched down. I haven't yet found the reason for this extreme slowing down, sometimes it is a specific file, sometimes the number of edits unsaved, whatever.

[heyvern]

I can double click a specific phoneme in Papagayo and hear if it hits the precise part of the word exactly. AS can NOT do this.
-vern

That's true. But you can limit the playable area to a region where the word or phrase sounds with CTRL right and left click over the time line to generate the red and green squares of playable area. Using this you can repeatedly play an specific portion of animation (like an specific word) as many times you want (just press space bar) and focus only on move or insert the keyframes for the phonemes. It could help a little for manual lip sync I guess...
-G

So has that been fixed now? In the past changing the start and stopping points caused the audio to START at the starting point instead of frame 1.

EDIT:
Yeehaaa! I never knew you could do THAT! Pretty cool. I like it.

Thanks for the tip Genete.

-vern