Clone a speaker’s voice to synthesize text into speech