Abstract: A system is provided. The system includes a first gateway operable to receive a real-time protocol (RTP) packet including video data from a first endpoint; and a second gateway operable to communicate with a second endpoint. The first gateway is operable to communicate the RTP packet to the second gateway in a sub-channel portion of an frame-based protocol.

Abstract: A method of generating speech from text messages includes determining a speech feature vector for a voice associated with a source of a text message, and comparing the speech feature vector to speaker models. The method also includes selecting one of the speaker models as a preferred match for the voice based on the comparison, and generating speech from the text message based on the selected speaker model.

Abstract: A method for compensating for asynchronous clock sources is presented. A buffer is provided for storing frames received from a packet network. The buffer having a preselected playout delay. Upon detecting a buffer overflow, a drain operation is performed on the buffer and the playout delay is modified to provide a sufficient number of buffer locations for frames to be stored. Upon detecting a buffer underflow, a fill operation is performed on the buffer and the playout delay is modified to provide a sufficient number of frames to be played out.

Abstract: A system is provided. The system includes a first gateway operable to receive a real-time protocol (RTP) packet including video data from a first endpoint; and a second gateway operable to communicate with a second endpoint. The first gateway is operable to communicate the RTP packet to the second gateway in a sub-channel portion of an frame-based protocol.

Abstract: A method for adjusting audio playback is disclosed. The method includes storing audio packets in a jitter buffer, and playing out the audio packets at a first rate. The method further includes determining that a capacity parameter for the jitter buffer is out of compliance with a predetermined standard, and based on the capacity parameter, determining a second rate for playing out the audio packets. The method also includes applying a pitch-invariant time scale modification algorithm to change the rate of playing out the audio packets from the first rate to the second rate.

Abstract: A system enables conferencing of multiple participants within a single voice communication session. A replay module records conference audio during the conference and, at appropriate times, provides this buffered audio to selected participants of the conference.

Abstract: A method for compensating for asynchronous clock sources is presented. A buffer is provided for storing frames received from a packet network. The buffer having a preselected playout delay. Upon detecting a buffer overflow, a drain operation is performed on the buffer and the playout delay is modified to provide a sufficient number of buffer locations for frames to be stored. Upon detecting a buffer underflow, a fill operation is performed on the buffer and the playout delay is modified to provide a sufficient number of frames to be played out.

Abstract: A method of generating speech from text messages includes determining a speech feature vector for a voice associated with a source of a text message, and comparing the speech feature vector to speaker models. The method also includes selecting one of the speaker models as a preferred match for the voice based on the comparison, and generating speech from the text message based on the selected speaker model.

Abstract: A system enables conferencing of multiple participants within a single voice communication session. A replay module records conference audio during the conference and, at appropriate times, provides this buffered audio to selected participants of the conference.