Voiceover Generation

You can automatically generate voiceover audio files for video scripts using the persona’s voice.

If the narrator of a voiceover should not be the persona, you can generate the audio manually in a third-party platform like ElevenLabs and upload the audio file in the batch format (See more here).

Creating a Persona Voice

Before generating voiceovers, you need to create a voice for your persona. This is done from the Personas page in the Positioning section.

Opening the Voice Design Modal

1. Navigate to Positioning > Personas
2. Find the persona you want to create a voice for
3. Click the Voice button on the persona card

Voice Design Process

The Voice Design Modal guides you through creating a unique voice that matches your persona’s characteristics.

Step 1: Enter a Test Script

The test script is used to preview how different voice options will sound. It should:

• Be at least 100 characters long
• Include a variety of emotions and tones to test the voice’s range
• Represent typical content the voice will narrate

A default test script is provided with various emotional cues. You can modify this script or write your own.

Step 2: Generate and Edit the Voice Design Prompt

The voice design prompt describes the characteristics of the voice you want to create.

Options:

• Click Generate Prompt to have AI create a voice description based on your persona’s characteristics (age, gender, personality, etc.)
• Manually write or edit the prompt to fine-tune the voice characteristics

The prompt should describe:

• Voice tone and quality (warm, authoritative, friendly, etc.)
• Speaking style (conversational, professional, energetic, etc.)
• Age and gender characteristics
• Any specific vocal qualities

Step 3: Generate Voice Options

Once you have a test script and design prompt:

1. Click Create New Voice (or Change Voice if updating an existing voice)
2. The system generates multiple voice options based on your prompt
3. Each option is rendered with your test script

Step 4: Preview and Select

For each generated voice option:

• Click the Play button to hear the voice reading your test script
• Compare the different options
• Select your preferred voice by clicking the checkbox

Step 5: Save the Voice

1. Ensure a voice option is selected
2. Click Select Voice to save it to your persona
3. The voice is now available for voiceover generation

Testing an Existing Voice

If your persona already has a voice configured, you can test it:

1. Open the Voice Design Modal
2. Enter or modify the test script
3. Click Generate Speech to hear how the current voice sounds
4. Use the Play button to listen to the generated audio

Changing a Persona’s Voice

To change an existing voice:

1. Open the Voice Design Modal
2. Modify the voice design prompt if needed
3. Click Change Voice to generate new options
4. Select and save the new voice

Generating a Voiceover using a Persona’s Voice

Once your persona has a voice configured, you can automatically generate voiceovers for video batches.

Opening the Voiceover Modal

From the batch’s Format tab (during Briefing or Editing stages):

1. Click the Actions dropdown
2. Select Attach Voiceovers

Voiceover Options

The modal presents two options for adding voiceovers:

• Generate: Automatically generate voiceovers using the persona’s voice
• Upload: Manually upload pre-recorded voiceover files. (Read more here)

Generating Voiceovers

Requirements:

• The batch’s concept must have a persona with a configured voice
• If no voice is configured, this option is disabled with a message directing you to create a voice first

How it works:

1. Click the Generate card
2. The system generates voiceovers for each creative in the batch:
   • Main creative: Generates audio for all scenes combined into one voiceover
   • Variation creatives: Generates audio for the first scene only (the hook/variation scene)
3. Once complete, the voiceovers are automatically attached to each creative