PapaGayo with a song

[laniangeline]

If I'm doing an animation with a character lipsyncing a song for a music video to it would I use Papagayo or just do the dialog manually

[Mikdog]

I find it I get more accurate lip-synching by scrubbing along the timeline and doing it manually without Papayago.

[heyvern]

It's difficult to use papagayo with singing. The main problem is of course the music. With papagayo for me at least it is the visual wave form that helps determine where the phonemes go. With a song the voice is hard to "see" because of all the other instruments and sound around it.

The other problem with singing is that words are never "spoken" in a "normal" way. The phonemes in papagayo for words just won't line up with the phrasing. The few times I tried it I ended up adding in extra phonemes to stretch out words to fit the song. You might as well do it by hand in AS.

-vern

[Samb]

I find it I get more accurate lip-synching by scrubbing along the timeline and doing it manually without Papayago.

same here.
ok, I HAVE to make the lip synch myself, cause papagayo doesn't support german words. but now it's not a problem and looks awesome
but it really takes time.
but singing = no papagayo.
singing is more than "A" "O" "th" etc. it's "A" "low A" "loud A" and so on.

[laniangeline]

Thanks guys. That's what I figured and believe it or not the song I'm using is a Falco song and they sing in German. So I'm going to have to put subtitles on it in English.

laniangeline

[FCSnow]

I have recently finished an animation where I had to lip sync to a song (Santa Baby). It's not easy, but do-able.

1. Start by breaking the song up into small digestable bits (short phrases or sentences).

2. Once Papagayo has broken down the lyrics to phonems. Start at the top and position the phrases over the proper waveforms

3. Next, position the individual word over the right spot on the waveform.

4. This most important and seldom used, the third layer is individual phonems and must be positioned correctly to make the lips move correctly to the song. If the singer holds a long "A" sound, The AI phonem is placed at the beginning. The next phonem is placed at the next sound change what ever it may be. Papagayo will then hold the AI sound until it reads the next lip change.

Good luck with your project.

F.C.Snow

[heyvern]

4. This most important and seldom used, the third layer is individual phonems and must be positioned correctly to make the lips move correctly to the song. If the singer holds a long "A" sound, The AI phonem is placed at the beginning. The next phonem is placed at the next sound change what ever it may be. Papagayo will then hold the AI sound until it reads the next lip change.

This is one of the reasons I like using bones for posing mouth positions with actions when possible.

I can still use Papagayo for lip sync (using a menu script from the forum) and then I can go in and move bones to exagerate the mouth positions for singing.

If you use the "regular" mouth positions it isn't as... dramatic.

It all depends on the sound file and how "clear" the voice is. Also on a Mac Papagayo is very sluggish so doing step 4 is very time consuming.

One time I actually tried to record the voice track separately by speaking the lyrics in time with the music. I used MY voice as the lip sync audio file so the words were clear. It worked sort of okay... but was hard to do. I had to load my voice file along with the original file into an audio editior and sync them up so I could swap them easily.

-vern

[FCSnow]

heyvern wrote:

One time I actually tried to record the voice track separately by speaking the lyrics in time with the music. I used MY voice as the lip sync audio file so the words were clear. It worked sort of okay... but was hard to do. I had to load my voice file along with the original file into an audio editior and sync them up so I could swap them easily.

If you're going to lip sync a song, you have to sing it. Just speaking it won't work. The inflections and sustains would all wrong on the waveform. Since I'm a rotten singer I wouldn't try it anyway.

In my test animation for "Santa Baby" I placed the lip sync setting by hand. Someone commented that at points it didn't look right. So, on the full length version I use Papagayo. At first it was tedious, but as gained experience with Papagayo, it went quickly.

Also, I used the same vector points in all the different lip shapes so I would have a smooth transition from one lip shape to another. I found that the "L" shape is good place to begin.

Today, if I were to try another lip sync song, I would find it so hard. I've climbed my learning curve. It's a gentle slope from here on.

F.C.Snow

[heyvern]

If you're going to lip sync a song, you have to sing it. Just speaking it won't work. The inflections and sustains would all wrong on the waveform. Since I'm a rotten singer I wouldn't try it anyway.

Uh... yes... I'm not a singer either... that is why I said "speaking" the lyrics... to have called it singing would have just been a horrific lie and an insult to actual singers of the world... well... maybe not Britney Spears.

-vern