Speaker-conditioned TTS with emotion & energy control
Synthesize speech in a cloned voice with optional blending