Two different and separate audio data streams are processed through a personal computer (PC) system or other computing device so that the separate and distinct audio data streams are heard through separate sound transducers. In a preferred embodiment, chat messages received over a network during execution of a multiplayer game are processed separately from sounds produced by the multiplayer game, enabling a user to hear the game sounds from speakers, separate and distinct from verbal chat messages, which are heard through earphone(s). The earphone(s) are included in a headset, as well as a microphone that enables the user to produce verbal data that are conveyed to a hardware control unit that is connected to the PC system through a universal serial bus (USB) port. The chat audio data are converted into an analog audio signal that is heard by the user in the headphone(s) and is spatially separate and distinct from the game audio data heard through the speakers. Included on the control unit are switches that enable the user to selectively direct verbal data either to the chat session or to a voice recognition engine that converts the verbal data into commands used for controlling the game.