Skip to content

Lip Sync and Mouth Shapes

Lip sync makes a character’s mouth follow speech. You attach a voice clip to the character and Glintia works out the mouth shapes for every sound. If a word comes out with the wrong mouth, you can fix that one shape by hand instead of re-recording.

Select the character on the canvas, then click Lip Sync in its toolbar, or press Ctrl+Shift+L (Cmd+Shift+L on a Mac). The panel has three ways to make a voice, a search box and your library of voices.

Move the playhead to where the character should start speaking, then choose one of the three options.

  1. Click Text to Speech
  2. Type the line, choose a voice and adjust the delivery. If Glintia can tell the character’s gender, it suggests a matching voice
  3. Click Generate, listen, and click Regenerate for another take
  4. Click Add to Timeline

Voices, delivery settings and costs are covered in Text-to-Speech.

Recorded and uploaded audio is transcribed so the mouth follows the words. Transcription costs 0.25 AI credits per minute of audio (a quarter credit at minimum), and an audio file must be 25 MB or smaller to be transcribed. Text-to-speech voices come with their mouth shapes already.

  • The clip is added to the selected character at the playhead and appears on the character’s Voice row in the timeline. Until a character has a voice, that row shows Add lip sync, which opens this panel
  • The clip can’t run past the end of the character’s time on screen. If it would, it’s cut short, so lengthen the scene and the character’s bar first for a long line
  • A character can have several voice clips, but they can’t overlap. If the new clip would overlap one that’s already there, you’ll see “Move the cursor to an empty spot”
  • Drag a voice clip on the timeline to retime it. The mouth moves with it

Every voice you generate, record or upload is saved to the Lip Sync panel, so you can use it again in any design.

  • Click a voice to add it to the selected character at the playhead
  • Hover to preview it, or click its play button
  • Filter with All, Pinned, Recent, and by source (AI voice, Recording, Upload) when you have more than one kind. Sort by Newest, Longest or A–Z, or use Search voices…
  • Pin to the top keeps a voice handy
  • On text-to-speech voices, New voice with these settings opens the editor with the same voice and delivery and an empty script, and Edit Voice opens that voice’s script for changes
  • To delete a voice, click the delete button, then click again to confirm

A No lip sync badge means no mouth shapes were generated for that clip, so the character’s lips won’t move with it.

Open Settings in the character’s toolbar and find the Lip Sync section. These settings apply to this character on the canvas, on top of the character’s own defaults.

  • Lip Sync switch: turn it off to keep the mouth still while the audio plays
  • Mouth Energy: from Mumble through Normal to Big
  • Timing: nudges the mouth up to 200 ms Early or Late against the audio, if it looks a beat ahead or behind
  • Jaw Movement: how much the lower face lengthens as the mouth opens, from Off to Strong
  • Head Movement: how much the head moves while speaking, from Off to Strong

Click Follow the character’s default settings to undo your changes.

Each voice clip has a mouth track: a row of blocks, each one mouth shape held over a stretch of speech. You can change which mouth a block uses, or move it in time. Edits are saved with the clip and used in playback and exports.

  1. Select the character and click Mouth shapes in its toolbar (it appears once the character has speech)
  2. If the character has several voice clips, choose one from the list
  3. Click a block in the strip. The panel shows its start and end time
  4. Pick a different Mouth shape, or use Move in time with Earlier and Later to shift it by about one frame
  5. Click Reset to generated to put the whole clip back the way it was

The shape picker lists every mouth in the character’s library, including extra shapes that no speech sound uses, so you can pick one deliberately. A block drawn with hatching is too short to show; it’s absorbed into the mouth before it.

  • Remove: click the X on the voice clip in the timeline
  • Keep the sound but stop the mouth: turn off the Lip Sync switch in Character Settings
  • Replace: remove the clip, then add another one at the same spot. For a text-to-speech line, you can instead click Edit Voice on it in the Lip Sync panel, change the script, generate it, and click Save Changes to update the clip on the selected character

Each character’s mouth follows only its own voice clips, so a conversation is a series of clips on different characters.

  1. Select the first character, move the playhead to its line, and add its voice
  2. Select the second character, move the playhead to just after the first line, and add its reply
  3. Repeat for each line, then drag clips on the timeline to tighten the gaps
  4. Play the scene to check that nobody talks over each other
  • The mouth doesn’t move: check that the Lip Sync switch is on in Character Settings, and that the voice doesn’t show a No lip sync badge. For a character you rigged yourself, the mouth must be mapped in the rig wizard
  • The mouth looks a beat late or early: use the Timing slider in Character Settings
  • One word looks wrong: fix that block with Mouth shapes
  • The voice stops early: it ran past the character’s end time. Lengthen the character’s bar, remove the clip and add it again