Why Custom Voices Matter

Generic AI voices sound like AI. Custom voices sound like your brand. When every video uses the same voice, your audience recognizes your content instantly. No need to hire voice talent for every project or worry about scheduling conflicts.

Accessing Voice Library

Voice Library lives in Brand Settings:

  1. Go to Settings > Brand Voice & Tone
  2. Click the Voices tab
  3. See all preset voices and your custom voice clones

Preset Voices

ILLIXIS includes 10+ professional preset voices:

Female Voices:

  • Rachel: Young, warm, natural tone
  • Charlotte: Elegant, sophisticated
  • Matilda: Warm, friendly, welcoming
  • Lily: Bright, expressive, British

Male Voices:

  • Adam: Deep, authoritative, professional
  • Daniel: British, refined, eloquent
  • Charlie: Casual, approachable

Multilingual Voices:

  • Sarah: Natural, supports 29+ languages
  • Freya: Expressive, international
  • River: Neutral, versatile (non-binary)

These voices are free to use and work immediately. No setup required.

Creating a Custom Voice Clone

Requirements

You need audio samples of the voice you want to clone:

  • Format: MP3, WAV, or M4A
  • Quality: Clear audio with minimal background noise
  • Duration: 1-2 minutes total recommended (can split across multiple files)
  • Content: Speech only, no music or sound effects
  • File Size: Maximum 10MB per file
  • File Count: 1-25 files (best results: 2-3 files, each 30-60 seconds)

Quality Tips:

  • Record in a quiet room
  • Use a decent microphone (phone mic works if room is silent)
  • Speak naturally, not reading robotically
  • Include varied sentences (questions, statements, emotions)
  • Avoid laughing, coughing, or long pauses

Cloning Process

  1. Go to Settings > Brand Voice & Tone > Voices tab
  2. Click "Create Custom Voice"
  3. Fill in the form:
  • Name: Internal name (e.g., "CEO Voice", "Sarah - Brand Spokesperson")
  • Description: Optional notes about when to use this voice
  • Gender: Male, Female, or Neutral
  • Language: Primary language (default: English)
  • Remove Background Noise: Keep checked for best results
  1. Click "Choose Files" and upload 1-25 audio files
  2. Click "Create Voice"

ILLIXIS processes your audio and creates the voice clone. This takes 10-30 seconds. When complete, you'll see your new voice in the list with a preview audio sample.

What Happens During Cloning

  1. ILLIXIS uploads your audio files to secure storage
  2. Audio is analyzed and a voice model is created
  3. ILLIXIS generates a preview sample: "Hello! This is a preview of my custom voice clone. I hope you like how I sound!"
  4. Voice is ready to use immediately

Cost: Voice cloning is free. You only pay when generating audio with the voice.

Managing Custom Voices

Setting a Default Voice

The default voice is automatically selected when creating new videos:

  1. Find your custom voice in the Voice Library
  2. Click "Set as Default"

The star icon indicates which voice is your default.

Regenerating Preview

If you want to hear a different preview sample:

  1. Click the "Regenerate Preview" button next to the voice
  2. A new preview audio generates with the same text

Deleting a Voice

  1. Click the "Delete" button next to the voice
  2. Confirm deletion

Warning: This permanently removes the voice. Videos that used this voice will still work (the audio is already generated), but you can't create new content with this voice after deletion.

Limits

  • Maximum Custom Voices: 10 per account
  • To create more: Delete an existing voice first

Automation Schedule

ILLIXIS keeps your Voice Library synchronized and up-to-date automatically:

  • Voice Library Sync: Runs daily at 4:15 AM UTC. New voices appear in ILLIXIS within 24 hours.
  • Voice Usage Quota: Checked before each generation request. If you exceed your plan's voice generation limit, you'll see a clear error message before any charges.
  • Custom Voice Processing: After uploading audio samples, voice cloning completes within 2-4 hours. Most clones finish in under 30 seconds, but complex uploads may take longer during peak times.

You don't need to manually sync voices. The daily automation ensures your library stays current.

Using Custom Voices

Custom voices appear automatically in voice dropdowns throughout ILLIXIS:

In Video Production

  1. Create a video from Video Production or Social Studio
  2. Open the "Voice & Audio" section
  3. Your custom voices appear at the top of the dropdown, marked with a star if default
  4. Select your custom voice
  5. Generate video

In Video Scripts

When generating video scripts from articles:

  1. Click "Create Video" from Content Hub
  2. Select your custom voice from the dropdown
  3. The generated voiceover uses your cloned voice

Voice Quality

Word-Level Timestamps

All voices (preset and custom) support precise word-level timing. ILLIXIS uses this for:

  • Animated Captions: Words appear exactly when spoken
  • Subtitle Sync: Perfect synchronization with voiceover
  • Caption Styles: Word reveal, highlight, fade effects

Speaking Rate Control

Adjust voice speed in video generation settings:

  • 0.5x: Very slow (good for educational content)
  • 1.0x: Normal speed (default)
  • 2.0x: Very fast (good for high-energy content)

Language Support

Custom voices work best in the language they were trained on. If your CEO speaks English, the cloned voice will sound best in English. For multilingual content, use preset multilingual voices (Sarah, Freya, River).

Troubleshooting

"Voice cloning failed"

Cause: Poor audio quality or file format issues.

Fix:

  • Re-record in a quieter environment
  • Use MP3 or WAV format (not M4A if possible)
  • Try shorter files (30-60 seconds each)
  • Remove background noise manually before uploading

"Voice doesn't sound like the original"

Cause: Insufficient or poor-quality training audio.

Fix:

  • Upload more audio samples (aim for 2 minutes total)
  • Include varied speech (not just monotone reading)
  • Ensure original speaker speaks naturally
  • Delete and recreate with better audio samples

"Voice sounds robotic"

Cause: This is rare with current voice technology, but can happen with very short training samples.

Fix:

  • Provide more training audio (2+ minutes)
  • Include emotional variety (questions, excitement, calm statements)
  • Check that original audio isn't overly processed or filtered

Preview audio not playing

Cause: Browser audio policy or network issue.

Fix:

  • Click the preview button (browsers require user interaction for audio)
  • Check browser console for errors
  • Try regenerating the preview

Best Practices

Training Audio

DO:

  • Use clean, clear recordings
  • Include varied sentence structures
  • Speak naturally and conversationally
  • Record multiple takes in different tones

DON'T:

  • Use noisy environments
  • Include music or background sounds
  • Use heavily compressed or low-quality audio
  • Clone voices without permission (legal and ethical requirement)

Voice Usage

DO:

  • Set your primary brand voice as default
  • Create separate voices for different show formats (e.g., "Podcast Voice" vs "Tutorial Voice")
  • Test with short previews before committing to long videos

DON'T:

  • Create voices for every person in the company (max 10 limit)
  • Use low-quality clones for important brand content
  • Assume all voices work well in all languages

Legal Considerations

Requirements:

  • You must have permission to clone any voice
  • Only clone voices of people who consent to being cloned
  • Disclose use of AI-generated voices where legally required
  • Respect personality rights and likeness laws

Voice cloning requires explicit permission from the voice owner. ILLIXIS enforces this policy.

Related Features

  • Video Production: Use custom voices in video generation
  • Social Studio: Generate social videos with your brand voice
  • Brand Settings: Manage voice strategy alongside visual identity

FAQ

Q: Can I clone celebrity voices? A: No. You need explicit permission from the voice owner. Cloning without permission violates platform terms and may violate laws.

Q: How many voices should I create? A: Start with one (your primary brand voice). Add more if you have multiple show formats or brand personalities. Most brands need 1-3 voices.

Q: Can I edit a voice after creation? A: No. If you're unhappy with the result, delete and recreate with better audio samples.

Q: Do voices expire? A: No. Once created, voices remain available indefinitely (until you delete them or reach the 10-voice limit).

Q: Can I share voices between team members? A: Yes. All voices are available to everyone in your ILLIXIS account. Set clear naming conventions so everyone knows which voice to use.

Q: What happens if I delete a voice that was used in old videos? A: Old videos keep their audio (already generated). You just can't create NEW content with that voice.

Q: Can I export/download my custom voice? A: No. Voices are tied to your ILLIXIS account and can't be exported as standalone files.

Summary

Voice Library turns one-time voice recordings into permanent brand assets. Clone your founder's voice once, use it in hundreds of videos. No scheduling, no retakes, no voice talent invoices.

This is the professional sound your brand deserves, without the professional voice talent budget.

Ready to lose the stack?

One platform. You approve. ILLIXIS executes. Marketing that just happens.

Join the waitlistNo spam, everUnsubscribe anytime
First 20 founding members: 50% off any plan for your first year.

Marketing, Unstacked.