Voice conversations by way of communications devices are conducted by transmitting symbols representative of a user's voice from a transmitting communications device and recreating the user's voice at a receiving communications device. The communications devices each include a processing engine responsive to a user's voice input for generating speech sample data indicative of predetermined portions of the user's voice. A storage device is coupled to the processing engine and stores the speech sample data. The processing engine also includes a communication module that generates transmission data, indicative of the user's voice spoken during a communication session as a function of the speech sample data and causes transmission of the transmission data to a remotely located recipient of the communication session.