Okay, I guess I can live with the red and green dots. Now, a serious question:
Is there any way to convert a switch layer, after it's been lipsync-ed, to a normal vector layer so I can control the interpolation? I was hoping the "blend morphs" feature would handle this, but I just checked the manual (imagine that), and I don't think it works on switch layers. Can anything be done to ease in and out of mouth shapes? Thanks!
Switch layer manipulation
Moderators: Víctor Paredes, Belgarath, slowtiger
-
Fonce Falooda
- Posts: 41
- Joined: Fri Jun 05, 2009 9:25 am
- Contact:
My personal opinion: if you need to have ease-in on lipsync phonemes you're seriously on the wrong track.
Open you copy of "Illusion of Life" (or rush to the library). Pages 455-468, "Dialogue". You will notice that the actual phoneme mouth shapes are the least important part of all dialogue.
The meaning of a line of dialogue should be delivered, in order of importance, by:
- body pose,
- gesture,
- face expression,
- eye direction,
- mouth shapes.
This is the classical animation approach to dialogue. I know that I sound ridiculously old-fashioned here, in a time where wordy talking heads express their page-long lines as stone-faced as possible, maybe because it's considered to be "cool" to not move any muscle except your tongue. But you can't have all characters talk like worst-case botox accidents.
It has been mentioned multiple times here: you are not restricted to your basic phoneme mouth shape set. You can use as many switches as you like, for different expressions or different loudness. And you can utilize those to have additional breakdown poses, which will serve your purpose as good as some ease-in curve.
In the very few cases when you really have to stretch a syllable for effect you might as well go into the switch layer and move the points.
Open you copy of "Illusion of Life" (or rush to the library). Pages 455-468, "Dialogue". You will notice that the actual phoneme mouth shapes are the least important part of all dialogue.
The meaning of a line of dialogue should be delivered, in order of importance, by:
- body pose,
- gesture,
- face expression,
- eye direction,
- mouth shapes.
This is the classical animation approach to dialogue. I know that I sound ridiculously old-fashioned here, in a time where wordy talking heads express their page-long lines as stone-faced as possible, maybe because it's considered to be "cool" to not move any muscle except your tongue. But you can't have all characters talk like worst-case botox accidents.
It has been mentioned multiple times here: you are not restricted to your basic phoneme mouth shape set. You can use as many switches as you like, for different expressions or different loudness. And you can utilize those to have additional breakdown poses, which will serve your purpose as good as some ease-in curve.
In the very few cases when you really have to stretch a syllable for effect you might as well go into the switch layer and move the points.
-
Fonce Falooda
- Posts: 41
- Joined: Fri Jun 05, 2009 9:25 am
- Contact:
Genete- That's really close! Except I think I'd have to do a mess of manual clicking to do a whole sentence, let alone a whole short! But it's obviously possible. Thanks!
Slowtiger- I'm well aware of body language and facial expressions. I just think linear interpolation of mouth shapes is the weak link in the chain. I don't want to put all that work into everything else and have the character's mouth move like a robot. It'll be distracting, especially if I deal with all the other stuff you mentioned.
Slowtiger- I'm well aware of body language and facial expressions. I just think linear interpolation of mouth shapes is the weak link in the chain. I don't want to put all that work into everything else and have the character's mouth move like a robot. It'll be distracting, especially if I deal with all the other stuff you mentioned.
On the contrary - I think the greatest obstacle to overcome is the thinking of having all lip sync completely automated. But this leads to uniform, flat speech (visually).have the character's mouth move like a robot.
In the end it doesn't really matter how you achieve your result - as long as you don't just treat all lip sync the same. Some sentences, even single words, look best when done automatically without interpolation, others look better with interpolation, and the rest will look best when completely done by hand, with different phonemes and even point tweaking. But I really dread the look of characters all moving their mouths the same.
BTW has this changed in v6? Can I have interpolated and step switch keys within the same timeline? This alone would improve that part a lot.
Fonce,
I wouldn't dismiss what ST is saying.
For our first six episodes we used lip sync interpolation. Now we've switched to step animation and the sync looks MUCH better. It depends on a whole lot of things (how much dialog you have, how fast your characters speak, the style you are animating, etc.) but in general, in today's post-MTV world, people are talking VERY fast and complete interpolation just doesn't work -- it gives a "mush mouth" sort of view to things.
Our episodes now are so much crisper and look so much better. But here's a good test for you -- find something you like, either movie or television, and do some analysis yourself. Whatever is it that you enjoy, bring it into your editor and advance it frame by frame (make sure you've turned off or otherwise disabled field rendering, so you can actually see the individual frames properly). When I do this I can clearly see that the sync is not interpolated -- but as I said, YMMV.
I wouldn't dismiss what ST is saying.
For our first six episodes we used lip sync interpolation. Now we've switched to step animation and the sync looks MUCH better. It depends on a whole lot of things (how much dialog you have, how fast your characters speak, the style you are animating, etc.) but in general, in today's post-MTV world, people are talking VERY fast and complete interpolation just doesn't work -- it gives a "mush mouth" sort of view to things.
Our episodes now are so much crisper and look so much better. But here's a good test for you -- find something you like, either movie or television, and do some analysis yourself. Whatever is it that you enjoy, bring it into your editor and advance it frame by frame (make sure you've turned off or otherwise disabled field rendering, so you can actually see the individual frames properly). When I do this I can clearly see that the sync is not interpolated -- but as I said, YMMV.
-
Fonce Falooda
- Posts: 41
- Joined: Fri Jun 05, 2009 9:25 am
- Contact:
ST and MK, I appreciate what you're saying, but I've done SOOOOO much lip syncing by hand that was by default "step interpolation", that I'm sick of it. I can get that in Flash or Toon Boom. Done a load of it in Flash with the "swap symbol" button, and then using two different plugins to do the same thing, only faster. Before that I was using Magpie Pro on bitmaps!!! 
The thrilling part of doing dialogue in AS here is doing the switch layer and having it tween the shapes for you, but it's exactly as you're saying. I DON'T want flat, uniform speech, so I'm still a little amazed that you only get linear interpolation on the switch morphing!
As it stands, I guess I'll use step interpolation until I neeeed something morphed, then I'll use vern's trick to pull the mouths out of the switch layer. Thank you both for thinking about this, and for all the insight I've gleaned from other threads you've weighed in on. I'll post stuff when it's ready.
The thrilling part of doing dialogue in AS here is doing the switch layer and having it tween the shapes for you, but it's exactly as you're saying. I DON'T want flat, uniform speech, so I'm still a little amazed that you only get linear interpolation on the switch morphing!
As it stands, I guess I'll use step interpolation until I neeeed something morphed, then I'll use vern's trick to pull the mouths out of the switch layer. Thank you both for thinking about this, and for all the insight I've gleaned from other threads you've weighed in on. I'll post stuff when it's ready.