Narration

Each narrated section keeps its Lyreline source content separate from its Studio narration draft. You can adapt the draft for spoken delivery without changing the canonical procedure in Lyreline.

Synchronize a visual with narration

A Visual Cue associates a narration phrase with a visual in the same section.

  1. Select the narration phrase associated with the visual.
  2. Choose Attach Visual Cue.
  3. Choose the target visual.
  4. Studio estimates when the visual should appear.
  5. Use Adjust Timing to move the visual slightly earlier or later than Studio's estimate when needed.
  6. Preview the section and refine the timing.

Use Reset to Estimate to return a cue to Studio's estimated timing. Cues stay attached to their target visual when visuals are reordered.

Add a natural narration pause

A narration pause adds silence at an authored location without changing the narration audio.

  1. Place the caret where the pause belongs.
  2. Choose Insert Pause.
  3. Select or set the duration.
  4. Studio finds a natural break in the narration where possible.
  5. Preview the result.

The current visual remains on screen during the pause. Authored pauses carry through Section Preview, Final Video Preview, and exported video.

Narration choices

Imported audio

Attach an existing WAV, AIFF, or M4A narration file. Studio copies managed audio into the project, reads its duration, and updates visual timing.

Apple Speech

Apple Speech uses the speech voices available through macOS. Choose a voice and adjust the supported speaking settings before generating narration. Voice availability can vary by Mac and installed system voices.

OpenAI narration

OpenAI narration is optional. You provide your own OpenAI API key in Lyreline Studio Settings. The key is stored in macOS Keychain rather than in project files or preferences.

Lyreline Studio 1.5 uses OpenAI's Realtime voice model, gpt-realtime-2.1-mini, for more natural pacing and expression while preserving the narration workflow you already use. marin remains the default and recommended OpenAI voice.

To generate OpenAI narration:

  1. Enter or edit the narration text.
  2. Choose OpenAI narration.
  3. Choose a voice and, when useful, add speaking instructions.
  4. Generate the narration.
  5. Preview and confirm the spoken wording.
  6. Use Visual Cues, authored pauses, Final Video Preview, and export normally.

You remain in control of the narration text. Studio reads the text you supply; it does not independently decide what the narration should say.

When OpenAI narration is selected, the narration text required to generate speech is sent to OpenAI using the API key you provide. Narration text is sent when you initiate generation. Testing the OpenAI connection also contacts OpenAI using that key. No OpenAI request occurs merely because Lyreline Studio is launched.

If a key is missing, open Lyreline Studio Settings, add and save the key, and optionally test the connection. OpenAI account access, usage, availability, and data handling are governed by OpenAI's terms and policies.

Narration generated by earlier Studio versions remains playable. Existing projects do not need narration regenerated merely because Studio 1.5 uses a newer OpenAI model.

Source updates and narration review

When Lyreline source changes, Studio does not silently replace an established narration draft or generated audio. The Lyreline source and Studio narration remain separate until you explicitly decide how to reconcile them.

For sections that need narration review, choose Review Narration Update… to compare the last reviewed Lyreline source, the updated Lyreline source, and the current Studio narration. Studio also provides an editable Proposed Narration.

You can use Start from Current Narration when the existing spoken wording is mostly correct, or Reset to Updated Source when the new Lyreline wording is the better starting point. You can then edit the proposal before applying it. For example, you can preserve polished narration, incorporate only the meaningful source changes, correct a typo, or blend old and new wording.

Apply Narration Update

Apply Narration Update changes only the Studio narration text. It does not:

  • regenerate narration audio;
  • replace the existing generated audio;
  • change the Lyreline source; or
  • mark the source update reviewed.

If generated narration already exists, changing its draft can cause Studio to show Needs Regeneration. Regeneration remains a separate action so you can finish reviewing and editing the narration before replacing the audio.

Studio supports Undo and Redo for an applied narration reconciliation, allowing you to return to the previous Studio narration while the source review remains unresolved.

Mark Reviewed

Mark Reviewed means that you have reviewed the updated Lyreline source and are satisfied with the section's Studio narration. It does not modify the narration text or regenerate audio.

After required section reviews are complete, the project-level Mark Updates Reviewed action can be used to acknowledge the overall continuation update.

For imported audio, changing narration text cannot regenerate the attached recording. Studio therefore does not offer the editable reconciliation proposal for imported-audio narration. If the spoken content also needs to change, update or replace the imported recording separately.