To achieve the best results when cloning a voice and to avoid failure, please follow these guidelines based on where you are building the voice:
For Desktop App Voice Building:
- Audio Quality: The sample must be high quality with no background noise. Sounds like air conditioners, fans, or other people talking can interfere with the build.
- Recording: Record in a quiet space or upload audio that was recorded in a silent environment.
- Technical Specs: Upload at least 15 minutes of clean audio. For ideal results, use mono-channel, 44.1Hz, 16-bit audio in .wav format featuring a single voice.
For Web Platform Voice Building:
- File Limits: Upload audio files up to 7.5 MB.
- Supported Formats: We support MP3, WAV, and M4A.
- Content: For the best results, the sample should be clear, free of background noise or music, and feature only one person speaking.
Comments
1 comment
yyyyy
Please sign in to leave a comment.