What Technique do you use for lipsyncing?

[RyanSurrey]

Switch layers, Mouth rig or other?

Can you show a simple sample of it?

[Greenlaw]

I use switches but the way I use the switches varies from project to project.

I used to like using Ctrl+Alt+Right-click over the mouth in the workspace. This method pops up a list near the character's face and, assuming you've named your phonemes clearly, you can simply select from the list to create your keyframe. Tip: I usually have a separate SBD control for 'emotions,' which can variably morph the mouths from neutral to positive or negative expressions of the same phoneme. This method is a very fast and direct way to do lipsync, and I used it for a few episodes of All Hail King Julien on Netflix last year.

However, there are a couple of downsides to this approach. First, the keyframes will be on the switch layer and not in the bone layer. That might seem obvious, but it means you need to keep track of keyframes in multiple layers. To work around this, you can use the Timeline Visibility option in the layers panel to keep both the bones and mouth channels up at the same time. Alternatively, you can Shift-Alt+Right-click the mouth drawing, and it will select the Mouth group in the layers panel and call up the mouth keys in the timeline. (Assuming you're not using Group layers in your Switch, which I often do.)

Another method I use often is to link the Switch to an SBD Action. This way, I can keyframe the mouth shapes in the Bone layer, which can be easier to manage. You'll want to use Copy Previous Key mode and set the dial to Step mode for this to work correctly, otherwise, the switch will interpolate through your drawings linearly.

The downside is that the mouth shapes need to be dialed through linearly, unlike the first method where you can directly select any part of the list. Not a big deal if you only have 11 phoneme mouths and maybe a handful of 'special expression' mouths, but it gets hairy if you have dozens of mouth shapes to scroll through. That's where the first method might be better.

One reason I used to prefer the first method was that it was easier to identify a phoneme by name, while the second method required you to recognize the mouth drawing visually. Victor and I recently came up with a solution that works really well when animating lipsync: create a separate switch with text layers for each phoneme and place that next to the SBD. This way, you get name feedback for the mouth as well as visual.

Tip: to keep the SBD together, I like to parent all my SBD's to a single 'mover' bone that has Independent Rotation enabled. Then, when I need to move the SBDs to another part of the workspace, I just move this single bone. If you parent the mouth text switch to that bone, it too will stay 'attached' to the mover bone and remain next to the mouth SBD.

Final tip: in addition to using the switch, I also like to 'squash and stretch' the mouth switch layer itself while the character is talking. This can help make a limited set of mouth drawings look a lot smoother and more sophisticated with only a little extra effort.

Typically, I'll just use Layer Binding on the Mouth Switch to bind it to the head bone or a Mouth 'mover' bone parented the head bone. You'll also want to reposition the origin for the switch group centered over the mouth drawing. Now, you can use the transform tool to 'squash and stretch' drawings to create fake 'inbetween' drawings, and you can move the group or the mouth mover bone it's attached to, to reposition the mouth over the face as needed. (I prefer using the mouth mover bone to keep the keyframes in that layer.) I suggest doing this as a second level of animation after you've completed the lipsync animation with the switch.

Okay, one more tip: In addition to having the audio track in the scene, I like to make use of Markers to indicate where parts of the dialog begins and ends. I don't break down every syllable, just the key phrases in the speech. Once you have the main phrases broken down like this, you can usually tell where the words, syllables, and pauses are by looking at the waveform itself.

Hope this helps.

[RyanSurrey]

I use switches but the way I use the switches varies from project to project.

I used to like using Ctrl+Alt+Right-click over the mouth in the workspace. This method pops up a list near the character's face and, assuming you've named your phonemes clearly, you can simply select from the list to create your keyframe. Tip: I usually have a separate SBD control for 'emotions,' which can variably morph the mouths from neutral to positive or negative expressions of the same phoneme. This method is a very fast and direct way to do lipsync, and I used it for a few episodes of All Hail King Julien on Netflix last year.

However, there are a couple of downsides to this approach. First, the keyframes will be on the switch layer and not in the bone layer. That might seem obvious, but it means you need to keep track of keyframes in multiple layers. To work around this, you can use the Timeline Visibility option in the layers panel to keep both the bones and mouth channels up at the same time. Alternatively, you can Shift-Alt+Right-click the mouth drawing, and it will select the Mouth group in the layers panel and call up the mouth keys in the timeline. (Assuming you're not using Group layers in your Switch, which I often do.)

Another method I use often is to link the Switch to an SBD Action. This way, I can keyframe the mouth shapes in the Bone layer, which can be easier to manage. You'll want to use Copy Previous Key mode and set the dial to Step mode for this to work correctly, otherwise, the switch will interpolate through your drawings linearly.

The downside is that the mouth shapes need to be dialed through linearly, unlike the first method where you can directly select any part of the list. Not a big deal if you only have 11 phoneme mouths and maybe a handful of 'special expression' mouths, but it gets hairy if you have dozens of mouth shapes to scroll through. That's where the first method might be better.

One reason I used to prefer the first method was that it was easier to identify a phoneme by name, while the second method required you to recognize the mouth drawing visually. Victor and I recently came up with a solution that works really well when animating lipsync: create a separate switch with text layers for each phoneme and place that next to the SBD. This way, you get name feedback for the mouth as well as visual.

Tip: to keep the SBD together, I like to parent all my SBD's to a single 'mover' bone that has Independent Rotation enabled. Then, when I need to move the SBDs to another part of the workspace, I just move this single bone. If you parent the mouth text switch to that bone, it too will stay 'attached' to the mover bone and remain next to the mouth SBD.

Final tip: in addition to using the switch, I also like to 'squash and stretch' the mouth switch layer itself while the character is talking. This can help make a limited set of mouth drawings look a lot smoother and more sophisticated with only a little extra effort.

Typically, I'll just use Layer Binding on the Mouth Switch to bind it to the head bone or a Mouth 'mover' bone parented the head bone. You'll also want to reposition the origin for the switch group centered over the mouth drawing. Now, you can use the transform tool to 'squash and stretch' drawings to create fake 'inbetween' drawings, and you can move the group or the mouth mover bone it's attached to, to reposition the mouth over the face as needed. (I prefer using the mouth mover bone to keep the keyframes in that layer.) I suggest doing this as a second level of animation after you've completed the lipsync animation with the switch.

Okay, one more tip: In addition to having the audio track in the scene, I like to make use of Markers to indicate where parts of the dialog begins and ends. I don't break down every syllable, just the key phrases in the speech. Once you have the main phrases broken down like this, you can usually tell where the words, syllables, and pauses are by looking at the waveform itself.

Hope this helps.

Thanks Mate,

The switch layer thing you said, it was a good and quick one. but i cant simulate the smile/frown with mouth while talking.

i think i would rather go with mouth rig. any tips for that?

[Greenlaw]

There are many ways to deal with that.

These days I prefer adding an Emotion or Expression SBD for the mouth. This way, I can dial a 'neutral' mouth shape toward a smile or frown during the animation using the same drawings. In other words, The SBD's center position would be neutral, turn right morphs the mouth shape to a smile, and turn left morphs it towards a frown or angry mouth. This control is completely independent of the Mouth SBD which controls which 'phoneme' drawing appears. The Emotion SBD reshapes the individual phoneme drawings for expressions. For clarity, I usually add 'ani' after the name of the drawing so I know that it's a shape that can be altered by the Emotion SBD.

Some animators will make the expressive changes directly on the timeline. They do this by selecting the drawing and using Magnet and Transform points to alter the drawings. This can look more natural but it's also a lot of extra work. To reduce that work, you can make Regular Actions for the most commonly used alterations.

Personally, I like to use the Emotion SBD most of the time, and will manually keyframe edits with the Magnet tool when I need something very specific.

Alternatively, you can probably use a Mesh Warp to deform the mouths but I don't really recommend that. It's doable but I'd personally rather have specifically drawn targets for smiles and frowns--it's easier to stay 'on model' this way. If you want to use mesh warp, you'll probably need to add a lot of points to your vectors and avoid using Beizier points to make the shapes you want. This is probably a better approach if you're using bitmap images. You can link the Mesh Warping to an SBD or make regular actions as you see fit.

Some people will go through the trouble of building a bone rig for the mouth but I think that's just going to make it difficult to animate. You can try it but I don't really recommend it myself.

If you really want flexibility, you can make a group for each mouth shape, and break the mouth into at least three layers. I typically use Mouth, Teeth and Tongue, and use the Mouth to mask the Teeth and Tongue layers. By separating these elements, it make is easy to create new mouth shapes very quickly and you can even add morphs to change the left/right angle of the mouth (group) drawing. Naturally, you might control this in a Head Turn SBD.

[RyanSurrey]

There are many ways to deal with that.

These days I prefer adding an Emotion or Expression SBD for the mouth. This way, I can dial a 'neutral' mouth shape toward a smile or frown during the animation using the same drawings. In other words, The SBD's center position would be neutral, turn right morphs the mouth shape to a smile, and turn left morphs it towards a frown or angry mouth. This control is completely independent of the Mouth SBD which controls which 'phoneme' drawing appears. The Emotion SBD reshapes the individual phoneme drawings for expressions. For clarity, I usually add 'ani' after the name of the drawing so I know that it's a shape that can be altered by the Emotion SBD.

Some animators will make the expressive changes directly on the timeline. They do this by selecting the drawing and using Magnet and Transform points to alter the drawings. This can look more natural but it's also a lot of extra work. To reduce that work, you can make Regular Actions for the most commonly used alterations.

Personally, I like to use the Emotion SBD most of the time, and will manually keyframe edits with the Magnet tool when I need something very specific.

Alternatively, you can probably use a Mesh Warp to deform the mouths but I don't really recommend that. It's doable but I'd personally rather have specifically drawn targets for smiles and frowns--it's easier to stay 'on model' this way. If you want to use mesh warp, you'll probably need to add a lot of points to your vectors and avoid using Beizier points to make the shapes you want. This is probably a better approach if you're using bitmap images. You can link the Mesh Warping to an SBD or make regular actions as you see fit.

Some people will go through the trouble of building a bone rig for the mouth but I think that's just going to make it difficult to animate. You can try it but I don't really recommend it myself.

If you really want flexibility, you can make a group for each mouth shape, and break the mouth into at least three layers. I typically use Mouth, Teeth and Tongue, and use the Mouth to mask the Teeth and Tongue layers. By separating these elements, it make is easy to create new mouth shapes very quickly and you can even add morphs to change the left/right angle of the mouth (group) drawing. Naturally, you might control this in a Head Turn SBD.

DUDE! That is awesome. I never thought of that.
What would be a easy way for 360 body turn?

[Greenlaw]

A 'easy' way?

There are actually many, MANY ways to create a 360 rig. The approach you take depends on how smoothly you want the turn, the style of animation, and the design of the character. This is really too broad a topic for a forum post.

I would recommend looking at some of the rigs that come with Moho. Not all are '360' rigs but there's enough turning in there to help you figure it out by breaking apart the rigs and reverse engineering it. (That's pretty much how I learned the basics, plus I got a lot of helpful advice from users in this forum whenever I got stuck.) The recent ASP 11 and Moho 12 sample rigs are very good for studying--they showcase a lot of different techniques you can learn from.

But since you're asking for the 'easiest' way, take a look at some of the generated rigs from Moho's character wizard. I don't find the results very useful for the productions I'm involved with but the tool does generate very basic '360' characters that might be good to learn from. The system basically outputs a unique rig for each view and places them in a single switch. That's about as easy as you can get, and if your animation tends to be 'snappy', you can do convincing turns with it. And if your character is symmetrical, you can use fewer rigs by flipping the L/R versions.

If you want a more smoothly turning setup, you'll want to use SBD's to slide elements to one side or the other, and probably reshape the contours of the character. Typically, I'll set up a separate SBD for the head and body, and sometimes a separate SBD to turn the upper torso separately from the lower torso.

It's possible to make a rig that does a smoothing turning 360 using a single rig (I did that not long ago for a TV show that hasn't come out yet,) but it can be very difficult to do. Most of the time, you really don't need that level of complexity--in a lot of TV shows that use rigged or semi-rigged puppets, you'll notice that characters rarely go off 3Q views, and they keep the turning range very limited for most scenes.

It's more common for me to create rigs that smoothly turn from front view to 3Q and then snap to separate profile drawing. If I'm using the same rig for front and profile, I'll snap the bones into position using Step interpolation in the SBD action. For most cartoon styles, this transition doesn't look too abrupt, especially if the character is animated to move into the turns. If back views are needed, I'll then snap from profile to back 3Q and then smoothly turn to back view. When possible, I'll flip the rig for L/R views, or maybe partially do that if the character is not symmetrical.

For some character rotations, you can do a lot by simply sliding textures inside the shapes--I do this all the time at work and I learned the technique from studying Victor's behind-the-scenes video for Woolbuddies. You can see that here: https://www.youtube.com/watch?v=FxkqxLG3BKQ

For many characters I create at work and in personal projects, the rig usually winds up being a hybrid of different techniques, dictated by character design, clothing and what the character is expected to do based on the storyboard/animatic.

Hope that helps.

[RyanSurrey]

Sorry for being a nerd,
But is there any technique for mouth in head turn?

[Greenlaw]

Again, it depends on the character design and style of animation you choose.

A basic approach: I parent the mouth switch to a bone and move the bone inside the head turn switch. I then use the contour of the head as a mask to clip the mouth.

In one popular style, the mouth never touches the side of the head even in profile (characters talk out the side of their face). That's probably the easiest since it doesn't even require masking.

Another is to allow the mouth to cross the edge of the mask. When the mouth is open, it has a black hole that shares the edge of the mask. In other words, the mouth doesn't 'cut-out' the head. This is also easy--it requires the mask but you don't do anything complex with it; you just let it be.

A more complicated method is to 'pre-cut' a hole in the side view of your character and place a 'cut-out' version of the mouth switch behind the head. This allows the mouth to have any open shape while the character is in profile. This probably looks the most realistic, and is reasonably easy to set up and doesn't require any masking. But I normally make the profile setup for the character a separate switch drawing when I want to do this--it's just a lot easier this way. It would be a lot more complicated if you try to make this a seamless effect--but I find you can hide the transition pretty well if you have the character move into the turn.

I use any of the above techniques depending on the show I'm working on and character design.

To make any of the above look more realistic, you should use groups of mouths inside the switches. Each group should have teeth (top layer,) tongue (middle) and a mouth (bottom) layer which acts as the mask. You can have Exclude Strokes on, which is the easy way to render the strokes over the teeth and tongue, or you can copy a reference layer to the top and knock out the fill color. Once you have the mouths broken down this way, you can counter animate the teeth and tongue layers to fall slightly behind the mouth animation in the head turn--this gives the mouth the appearance of depth or perspective in the turn.

Hope these tips helps. There are actually many other ways to animate mouths. i would experiment and see what works best for you. And don't forget to share any cool discoveries you make.

[Greenlaw]

One big reason why I prefer to make the profile for 'realistic' elements a different switch drawing is because the morphs will probably behave badly during and after 'turning' from front to side views. For example, while it's possible to morph the front view of a realistic human eye into a profile view, you're going to find that the 'blink' morphs created for the front view will not work for the side view. This problem will hold true for mouth shapes too.

A few months ago, I ran into that exact situation. I had set up a realistic character whose eye could smoothly transition from front to side, but the blink created for the front did not work properly for the side. To compensate for this, I had to create a separate 'side view' eye that I switched to during the head turn that had a functional side blink morph built-in. What made this tricky was that I was asked to close the character's eyes during a slow head turn, which meant the blinks for the two angles had to match up pretty closely. I was able to do it but it was harder than I thought it would be. Then again, it didn't help that I hadn't anticipated that request and had to quickly 'hack' my setup. (One of these days, I'm gonna revisit this situation and try to do it right.)

Anyway, if you're going to animate a realistic mouth smoothly in a head turn, you'll probably need to apply the same techniques.

(Fortunately, most of the cartoons I work on don't require that much 'realism'. A blinking 'circle' eye looks pretty much the same in front and side views.)