Audio and speech issues

Fix missing or wrong speech in generated video. Job to be done: write clear dialogue, pick an audio-capable path, allocate enough duration, and use Lip Sync when you need face sync on an existing clip.

• Put spoken lines in clear quotes in your prompt.

• Label narration vs on-camera dialogue when both appear.

• Kling 3.0 Omni often carries native audio; some paths use separate TTS.

• Longer scripts need more seconds (or an extend chain) so words are not cut off.

• For precise face-synced audio on an existing clip, open the asset’s Lip Sync studio — not a generic Director plan item.

• [@me] can carry configured voice when you cast yourself; do not expect the Director to cast [@me] uninvited.

• Dialogue without quotes or speaker labels

• Asking for a long monologue inside a 4–8s Veo single

• Expecting free EdgeCut AI to match Kling native audio quality

• Trying to “lip sync” via chat alone without the Lip Sync studio

1. Choose Kling Omni (or another audio-capable path you are offered) when speech is central.

2. Quote the line and state who speaks on camera.

3. Set duration long enough for the words — or ask for an extend chain.

4. Confirm the approval card.

5. If lips are off on a finished clip, open Lip Sync on that asset.

Prompt and model choice: Creative Director. Persistent audio pipeline failures after retries: /help or support@edgecut.ai with timestamps and the asset id if visible.