Skip to main content
Resonance’s voice cloning feature lets you create custom voices from audio samples, perfect for branding, personalization, or creating unique voice characters.

​
Overview

Custom voices are created from audio samples and become available to all members of your organization. You can clone voices using two methods:
  1. Upload - Provide an existing audio file
  2. Record - Capture audio directly in your browser

​
Creating a Custom Voice

1

Open Voice Cloning Dialog

Click Voice cloning from the sidebar menu, or navigate to the Voices page and click “Create Voice”.
2

Choose Input Method

Select how you’ll provide the voice sample:
Upload an existing audio file from your computer.Supported formats:
  • All audio formats (MP3, WAV, M4A, FLAC, etc.)
  • Maximum file size: 20 MB
Upload methods:
  • Click “Upload file” button
  • Drag and drop audio file into the dropzone
3

Preview Your Audio

After uploading or recording:
  • Click the Play button to preview
  • View file size and duration
  • Click X or Re-record to start over
4

Enter Voice Details

Give your voice a descriptive name.Examples:
  • “CEO Voice - John Smith”
  • “Brand Narrator”
  • “Customer Support Agent”
  • “Hero Character”
Select the category that best describes this voice’s intended use:
  • General - Multi-purpose voice
  • Audiobook - Long-form narration
  • Conversational - Natural dialogue
  • Customer Service - Support and assistance
  • Narrative - Storytelling and narration
  • Characters - Voice acting and characters
  • Meditation - Calm, soothing guidance
  • Motivational - Inspiring and energetic
  • Podcast - Interview and discussion
  • Advertising - Promotional content
  • Voiceover - Professional VO work
  • Corporate - Business and training
Choose the language and locale for this voice.Search from a comprehensive list of language codes:
  • English (United States) - en-US
  • English (United Kingdom) - en-GB
  • Spanish (Spain) - es-ES
  • And many more regional variants
Default: en-US
Add notes about:
  • When to use this voice
  • Tone and characteristics
  • Best use cases
  • Any special considerations
This helps team members choose the right voice.
5

Create Voice

Click Create Voice to process and save the voice.The system will:
  1. Upload the audio to secure storage (R2)
  2. Process and optimize the voice sample
  3. Make it available to your organization
  4. Add it to your Team Voices list

​
Using Custom Voices

Once created, your custom voice appears in:
  1. Text to Speech page - Under “Team Voices” in the voice selector
  2. Voices page - In your organization’s voice library
All organization members can:
  • Generate audio using the custom voice
  • View voice details
  • Preview the original voice sample
Custom voices are organization-scoped. Only members of your organization can access them.

​
Managing Custom Voices

​
View All Voices

Navigate to /voices to see:
  • All custom voices created by your team
  • Built-in system voices
  • Voice categories and metadata
  • Search and filter options

​
Delete a Voice

Deleting a voice is permanent and cannot be undone. Any existing audio generated with this voice will still be accessible, but you won’t be able to create new generations.
To delete a custom voice:
  1. Go to the Voices page
  2. Find the voice you want to remove
  3. Click the delete/trash icon
  4. Confirm the deletion
The voice and its associated audio file will be removed from storage.

​
Voice Quality Tips

Recording Environment

  • Use a quiet room
  • Minimize background noise
  • Avoid echo (use soft furnishings)
  • Position mic 6-12 inches away

Audio Sample

  • Speak naturally and clearly
  • Maintain consistent volume
  • Include varied intonation
  • Aim for 30-60 seconds minimum

File Quality

  • Use lossless formats when possible (WAV, FLAC)
  • Minimum 16-bit, 44.1kHz recommended
  • Avoid heavy compression
  • Keep under 20 MB

Voice Characteristics

  • Choose expressive samples
  • Include natural pauses
  • Avoid monotone delivery
  • Match intended use case

​
Best Practices

​
For Professional Voiceovers

  • Record with a quality microphone
  • Use a pop filter to reduce plosives
  • Include a variety of emotional tones
  • Record multiple takes and choose the best

​
For Brand Voices

  • Use the same speaker across all samples
  • Match your brand’s tone and personality
  • Create separate voices for different use cases
  • Document usage guidelines in the description

​
For Character Voices

  • Exaggerate character traits in the sample
  • Include emotional range (happy, sad, excited)
  • Record longer samples (60+ seconds)
  • Test with varied text before finalizing

​
Troubleshooting

  • Grant browser permission to access microphone
  • Check system microphone settings
  • Try a different browser (Chrome recommended)
  • Verify microphone is connected and working
  • Ensure file is under 20 MB
  • Verify it’s a valid audio format
  • Check your internet connection
  • Try converting to WAV or MP3
  • Re-record with better audio quality
  • Reduce background noise
  • Use a longer audio sample
  • Ensure clear, natural speech
  • Refresh the page
  • Check you’re in the correct organization
  • Verify the voice was successfully created
  • Contact support if issue persists

​
Organization Sharing

Custom voices are shared across your entire organization:
  • All members can use team voices for generation
  • Voice library is organization-wide
  • Permissions are managed through Clerk Organizations
  • See Managing Organizations for details

​
Next Steps

Build docs developers (and LLMs) love

Get started for freeTalk to us