Papagayo - Singing Tutorial?

[jaakay]

Hi All,

Just wondered if there was a tutorial on how to lip sync singing using either ASP or better still, Papagayo.

If anyone knows of one, could they point me in the right direction.

Earlier today I watched a Papagayo tutorial by MKelley and he mentioned that he would do one if people requested it. Not sure if he ever did. I also don't know whether he still posts on this forum as he has another forum of his own (can't get registered on it to ask the question - keeps saying there is an error).

Anyway, any help would be appreciated.

Thanks in advance.

[GCharb]

I don't think you will find one of those, but whether it is signing or speaking, it basically works the same way, you import the sound file in AS, then you break down the consonants and vowels on the timeline.

[hayasidist]

wow - that question made me think!!

Papagayo uses (by default / English) the Preston Blair phoneme set - and so there's the corresponding set of 10 mouth shapes (AI, O, E, U, etc, L, WQ, MBP, FV, rest) which themselves are mappings onto the CMU phoneme set (the Carnegie Mellon University Pronouncing Dictionary - a machine-readable pronunciation dictionary for North American English) of 39 basic sounds.

All of which is great for speech ... but what about singing???

Will the AS / Papagayo waveform analyser pick out words from music? And even if it does, my intuition says that that you'll need a new phoneme set that takes into account dynamics (mouth shape for a fortissimo "O" different from a pianissimo "o"??), "decoration" (tremolo "o" different again???) and pitch (high note "o" different from low note "o"???).

or maybe the results of the analyser output and phoneme set is "close enough" and therefore good enough - with any "extra" mouth shaping etc handled as a "task left to the animator"!??? Now there's an experiment for when there's a lull in the schedule....

[Víctor Paredes]

There is a point important here. In normal speak and in singing the mouth is not as important as you think. Most (almost all) of the magic come from the body movement. You don't necessarily need a set full of different mouths for any little intonation change.
Even before of define each mouth, take your character and decide where the accents are and which poses or movements that accents need. Talking or singing involves the whole body acting, not just the mouth.
Forget the mouth for a while and try to make your character sing only with the body. Once you get it, the mouth is the easier part.

[SvenFoster]

Interesting selgin. I've always hidden the mouths on my characters until I'm pretty much done. Then I'll do the lip syncing.
Seems I've been accidentally doing something sensibly

[GCharb]

As selgin said, you animate the body with accents on the signing, then the mouth, I personaly dont use phoenemes for speech, or signing, I just feel it, I am a big Richard William's fan I must admit.

[funksmaname]

I find listening to the melody and raising or lowering the head/body (a lot, or a little) along with intonation helps sell it (in both speech and singing). if the pitch goes up, lift, if it goes down, drop.

regarding papagayo, you don't have to use the mouth shapes that come with it - infact you can just use it to put down any 'instruction' along the time line and then feed that dat file into AS as actions... but that's a different story

[hayasidist]

yes - I agree that the body acting is key to the magic. (And I keep Richard W's ASK on my desk too!)

In long shots the mouth is barely visible and I've found that just flapping the lower lip up and down more or less sync'd with sound works "ok" (there is always room for improvement!!) when combined with the body movements along the lines that Funks, Selgin and Gilles say (and one day I'll do that better than I do right now too ).

Depends on how you choose to count them, but there are about 300-ish pairs of muscles that move the skeleton around, with around 50 in the head and neck that change facial features such as eye position, mouth shape and head position. So I'm guessing that no-one rigs their character with an anatomically correct set of bones (to act as the muscles that shift limbs and skin around) but ...

and agreeing with what Selgin says - it's about poses and movements - so for CU work it's poses and movements in head that are important. Which I guess that means that "template" mouth shapes are good starting points but only a part of the whole answer?