Lip-Sync and Switch Layers
A switch layer is a stack of drawings where only one shows at a time, and lip-sync is the classic use: a set of mouth shapes that flip in time with speech. FoundryInk handles both from a single Switch / Lip-sync panel. This page covers the whole thing: drawing your mouth shapes, showing and keying them by hand, mapping shapes to sounds, and getting the timing to feel like real talking.
1What a switch layer is
A switch layer holds several drawings stacked in one place, but shows exactly one of them at any given frame. Flip which one is showing over time and you get flip-book animation: a blinking eye, a spinning coin, or the star of this page, a talking mouth. Because only one shape shows at once, you never have to hide the others by hand. The switch does the choosing for you.
2Draw your mouth shapes as sub-layers
Lip-sync starts with the shapes. Draw each mouth as its own sub-layer inside the switch: one for a closed mouth, one for a wide open mouth, one for a rounded mouth, and so on. Each drawing sits on its own sub-layer so the switch can pick between them. Keep every mouth registered to the same spot on the face so they do not jump around when they flip.
3Open the Switch / Lip-sync panel
With your switch layer selected, open the Switch / Lip-sync panel. At the top it shows the current frame as Frame {n}, so you always know where the playhead sits. Below that are tiles, one per drawn sub-layer, each labelled with its shape name (a fresh shape gets a default name like Shape 1). This panel controls everything that follows.
4Showing now: one shape at a time
The panel always shows which shape is live right now. Only one sub-layer displays at a time, and the panel makes it clear which one is currently showing. This is the heart of a switch: the others are still there, tucked away, just not visible. Understanding that only one shows at once is what makes the keying in the next step click into place.
5Click a shape to show it and set a switch key
Move the playhead to the frame where a mouth should appear, then click that shape's tile. This does two things at once: it shows that shape on the stage and it sets a switch key at the current frame (the action reads as show that label and key it at this frame). From that frame onward, the switch displays your chosen mouth until you tell it otherwise. Keying is just clicking the right tile at the right frame.
6How a mouth holds until the next key
A switch does not fade between shapes. Once you key a mouth at a frame, it stays on screen, unchanged, right up until the next switch key tells it to become a different mouth. So if you key a wide mouth at frame ten and a closed mouth at frame twenty, the wide mouth simply holds for those ten frames and then snaps to closed. This hold-then-snap behaviour is exactly how hand-drawn lip-sync works, so lean into it.
7Read and clean up the Keyed frames strip
Every switch key you set shows up as a chip on the Keyed frames strip, giving you a tidy map of where the mouth changes across the shot. If you set a key at the wrong frame or picked the wrong mouth, click its chip to clear that key. This makes fixing a fumbled sync quick: you do not have to re-record anything, just clear the bad chip and click the right tile at the right frame.
8The small set of mouths you actually need
You do not need a mouth for every letter. A convincing set is small: a closed mouth for M, B, and P, a wide mouth for E and I sounds, a rounded mouth for O and U, an open mouth for A, and one or two in-betweens for F and L. Five to eight shapes carry almost any dialogue. Drawing a huge library of mouths is a common beginner trap that costs hours and buys very little.
9Map mouths to sounds in the Phoneme shapes table
For a faster start, the panel has a Phoneme shapes table. Each row is a mouth sound, and next to it is a picker that starts on Not assigned. Open the picker and choose which of your drawn shapes represents that sound: point the closed row at your closed mouth, the open row at your open mouth, and so on. This tells FoundryInk which drawing means which sound before you place a single key.
10Key a phoneme row, or apply them all
Each row of the Phoneme shapes table has its own Key button, so you can drop a single mapped mouth at the playhead without hunting through tiles. When you have a run of shapes mapped and placed the way you want, click Apply phoneme shapes to commit the mapping into your switch. This connects the table (which sound means which mouth) to the timeline (when each mouth appears).
11Sync to the audio
Lip-sync is only as good as its timing, so bring your dialogue in on an Audio row and work against it. Step through the shot frame by frame, listening for where each sound lands, and key the matching mouth there. The transport shows your position as Frame {n} / {total}, and the Previous keyframe and Next keyframe jumps let you hop between the mouths you have already set to check the spacing.
12Preview and refine the rhythm
Hit Play and watch the mouth run against the sound. First passes almost always feel a beat late, because a mouth should start opening just before the sound, not on it. Nudge keys a frame or two earlier where the sync lags. Toggling stepped playback like On 2s can also give speech a snappier, more hand-drawn cadence than a mouth that changes every single frame.
13What good lip-sync looks like
Good lip-sync is not busy, it is legible. The mouth holds calmly through steady sounds and snaps open on the strong vowels, so the eye reads speech without noticing the trick. Over-animated mouths that flutter on every frame look chattery and cheap. When you can play the shot with the sound off and still roughly read the rhythm of talking, you have nailed it.
14Reuse the switch for blinks and more
Everything you learned here works for any switch, not just mouths. Build a second switch layer for the eyes with open, half, and closed drawings, and key an occasional blink the same way you keyed a mouth. A well-timed blink every few seconds does almost as much for a face as the lip-sync does. The panel, the tiles, the keyed chips, and the hold-until-next-key behaviour are identical.
Lip-sync questions
That is what a switch layer does. It holds all your mouth sub-layers but displays exactly one, chosen by your switch keys. The others are still there, just hidden until a key brings them forward.
Move the playhead to that frame, then click the mouth's tile in the Switch / Lip-sync panel. That shows the shape and sets a switch key at the frame in one click. The mouth then holds until your next key.
Find its chip on the Keyed frames strip and click the chip to clear that key. Then park the playhead at the right frame and click the correct tile. No re-recording needed, just clear and re-key.
No, and you should not. A mouth holds until the next switch key, so you only key where the mouth actually changes. Fewer, well-placed keys read as cleaner speech than a key on every frame.
It maps each mouth sound to one of your drawn shapes. Each row starts on Not assigned; pick your shape for that sound. Once mapped, use the per-row Key to drop a mouth or Apply phoneme shapes to commit the whole set.
Five to eight is plenty for almost any dialogue. Start with closed, open, wide, and round, and add a couple of in-betweens only if a close-up needs them. A giant mouth library is more work than it is worth.
A mouth should start opening just before its sound, not on it. Nudge the lagging keys one or two frames earlier. Perfect frame-on-frame alignment reads as late to the eye, so anticipate the sound slightly.
Absolutely. Build a switch with open, half, and closed eye drawings and key an occasional blink exactly like a mouth. The panel, tiles, keyed chips, and hold behaviour are all identical.
Your character can talk
You drew a set of mouths, mapped them to sounds, and keyed them into speech that holds and snaps like the real thing. A talking face is one of the biggest leaps a character can make, and you just made it. Next, tackle limbs that pose themselves: point a hand at a target and let inverse kinematics bend the whole arm to reach it.