The player's model can support "voice-to-mouth