I am practicing lip syncing and I used a quote from the movie Caddyshack to test it out. This is about 1 evenings work. I would appreciate any advice from people on how to smooth out the animation. Please check out my animation and let me know what you think
Hmmm, that’s quite jumpy isn’t it? I can’t guess why except that you may have used too many keys, too close together (I’m also viewing it on a dicky player). I always set one or two keys per word initially, then add in-betweens where they’re needed. Some words might only get one shape key.
The most important thing, just as with poses, is to hit key visemes (mouth shapes) that make the lips “read” properly. One example is near the end where he says “problem”. The “P” is imperative to that word reading properly yet I can’t see his lips close, even in slow motion. They almost close, but not quite. What this means is that at full speed, his teeth are always visible and we don’t experience the “P” sound visually. If you get what I mean. Obviously some characters don’t allow for full closure (like Wu-Man) but your character looks capable of sealing his lips so he needs to.
If you’re willing to upload him at this stage, by pm if you prefer, I’d be happy to take a look at him. Nice character by the way.
Thanks for the reply AndyD. I was thinking that there was too much jaw movement before but to many keys makes a lot of sense. I was trying to add keys for each phoneme. I will try your method out. I think the only part of the lip syncing that looked ok was “Good” and that would make sense because there were only 3 phonemes(keys) that I added. Whereas problem has 7.
Definitely better but I still feel it’s a bit tight. The lips still jump from shape to shape. Most shapes go close to 100%.
When I’m lip-syncing, I sit there and try to imitate the audio as close as possible, accent, tone, everything. I try to be the character. Doing this I can feel what my mouth is actually doing as I speak. In this case, my lips barely move at all on “Do you take Drugs Danny”, they do subtle “oh”, “ah” movements with a slight emphasis on “you” and “drugs” and my mouth almost settles to neutral on “Danny”. My jaw barely moves at all. This happens because the voice actor is speaking so fast. In fact, I feel more action in my eyebrows than in my mouth in that part of the speech (I often find this).
I think you’re probably right that you may have too much jaw/cheek movement for the range of motion in the mouth area. But I also feel you need to conentrate a little more on the supporting actions. A slight tilt of the head on key phonemes/words, and a bit of eyebrow action goes a long way to selling the lip-sync. You’ve got some in there already but it’s a bit stagnant. The head moves and stops, moves and stops, the brows do the same.
Even something as simple as not having him stare straight at the camera can make a difference because we can believe he’s speaking to someone.
I believe (and I think it’s a widely held belief) that the animation should work pretty well with no lip sync at all (think of the “flour sack” principle of showing emotion with nothing more than a simple shape). If you can get this to work, then lip-syncing is a breeze.
Wow! Thanks AndyD, I think everyone could learn a lot from that post. Hmm…Perhaps next I will animate the movements first and then the lip syncing. I will definitely pay more attention to the emphasized words and non-emphasized as well. I really appreciate you sharing your knowledge on this subject.
Thanks If you’re interested in further reading, take a look at the lip-sync tutorial at keithlango.com. That’s where I started to think seriously about better lip-sync (not that I’d really done any before then anyway) and he explains it with a couple of interesting examples.
The idea of syncing actual audio/sounds rather than words is the reason I feel scripted/automated sync solutions would rarely be as good as manual syncing. I have read of one piece of script software (for Maya I think) that attempts to address this problem.
Yeah, you dont really need to animate visemes for every simgle syllable. Also, the head movement is a little jerky for my taste. But it still looks pretty good.
Well, you’ve gotta be happy with your improvement That’s much better.
I’ve been at this very point myself and I’ll tell you what I was told… push it. Unlike your first Caddy Shack clip, this new audio has energy and you have to show it in your facial expressions and in the poses.
You’ve added a subtle head twitch on “shut up” which I quite like. This is what you need to apply right through, and not necessarily so subtle. Hit the key poses hard with a piece like this. Move fast and hold then move fast.
He’s not happy - show it. It’s not easy (frustrates the hell out of me) to do it without it being jerky but that’s where “moving holds” and “eases” come in.
The lip-sync here is definitely much better. The main problem I notice is the “W’s” in both “always”. I’m not really seeing them. “W” is tight, like “ooo” and you may have to hit it early and hold a couple of frames longer for it to register. Actually, it isn’t a bad idea to try and hit some visemes a couple of frames before the audio because the mouth has to form before it can make a sound. Practice will tell you when it’s necessary.
In general, I think you’re onto something and if I was you, I’d keep at this one until you really feel you can’t push it any further. Here’s the 4th revision of my own “angry” piece. It’s not perfect but if you saw the first version you’d see how it progressed. In version one, the thumping fist barely lifted off the table, the body hardly moved at all and the head had a limited range of “nodding” movements! It took courage to keep pushing it further.
Thank you for the words of encouragement AndyD. I have started “pushing” it a little more and I think it is really making a difference. I still think it can be pushed a lot more though. I have seen the earlier angry pieces of yours. I really like seeing the progression that it is making. The latest version is freakin sweet!
