Definition
Voice cloning synthesises a target language in the caller's own vocal timbre, so the translated voice sounds like them rather than a generic model.
Voice cloning uses a short reference sample of the caller's own voice to synthesise speech in a different language. Modern models (Cartesia, ElevenLabs) deliver this at low enough latency to use on live calls; the recipient hears the caller's own vocal character but speaking a language the caller does not. ChromePing supports an opt-in Cartesia voice-cloned TTS path for callers who have consented to a short sample; the default TTS path uses a generic voice from Telnyx. Voice cloning adds about 150 ms to the pipeline but can carry emotional nuance that a generic voice loses.
Related terms
Text-to-speech (TTS)
Text-to-speech is the last stage of a translated call, it reads the translated text aloud in the target language's voice.
Speech-to-text (STT)
Speech-to-text turns spoken audio into text in near real time, the first stage of translated calls.
Call translation latency
Call-translation latency is the time between the moment a speaker finishes a sentence and the moment the other party hears the translated version, ideally under 1.5 seconds.