Comment by JP_Watts

Can you describe how it pick different voices? Does it need separate audio channels, or does it recognize different voices on the same audio input?