AI BriefingSpeech To Speech
Key point
ElevenLabs has unveiled Voice Changer, which converts voices while preserving the original's emotion and intonation, along with Turbo v2, optimized for real-time interaction.
Details
ElevenLabs has released Voice Changer, which changes only the voice while keeping the original recording's emotion, timing, pace, and pronunciation intact. This enables control over subtle emotional expression and nuanced delivery that text-based text-to-speech struggles to achieve.
Users can demonstrate emphasis, pauses, and rhythm for specific phrases through their own recorded or uploaded voice, and then apply these directly to a target voice. Technically, this works by rendering the content of the original voice using the phonemes of the target voice. This is similar to face-swapping technology, with the key being to strike a balance between the target voice's characteristics and the original's emotional features.
Key updates include the following:
- Eleven Turbo v2: A model optimized for real-time interaction, supporting (m)uLaw 8kHz format for IVR systems.
- Enhanced Studio features: Supports gain adjustment, dynamic compression, and metadata (ISBN, author, title) insertion tailored to audiobook production guidelines.
- Voice Library overhaul: Existing voices will be replaced, with 20+ new voices to be added in the coming weeks.
- Pronunciation dictionary: A highly requested pronunciation control feature is being introduced.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
Qwen Releases Qwen3.8-LiveTranslate, a Real-Time Simultaneous Interpretation Model with 2.3-Second Latency