AI Audiobook Voice Conversion

I want to take a single recorded narration and transform it—through AI vocal conversion—into distinct, believable voices that fit a range of book characters. The goal is to breathe life into motivational and emotional works as well as conventional fiction and non-fiction titles, so each character sounds unique while remaining consistent across chapters.

Here’s what I need from you:
• An end-to-end workflow (model selection, training or fine-tuning, and post-processing) that converts one clean reference voice into multiple character voices suitable for long-form listening.
• Output in standard audiobook formats (44.1 kHz WAV or 192 kbps MP3), ready for direct mastering or distribution.
• Settings or presets I can reuse for new projects so I’m not locked into a single production cycle.

Acceptance criteria:
• At least four clearly differentiated voices delivered from the same base narration: two suited to fiction, one for non-fiction, and one adaptable for motivational or emotionally driven passages.
• Consistent pronunciation, pacing, and volume within each generated voice.
• Minimal artifacts—no robotic or metallic resonance when played through consumer headphones.

If you already work with tools such as ElevenLabs, Voice Conversion GANs, or similar neural TTS engines, mention it briefly so I know we’re speaking the same language. Once the process is proven on one chapter, we can expand to entire books.