A channel stream is received and demultiplexed into a video packetized elementary stream (PES), audio packetized elementary stream (PES), and program clock reference (PCR). Indexing circuitry stores the video PES and the audio PES in a buffer, locates a presentation time stamp (PTS) in the video PES and stores that PTS in the buffer, locates a start of each group of pictures (GOP) in the video PES and stores those locations in the buffer, and locates a PTS in the audio PES and stores that PTS in the buffer. Control circuitry empties the buffer of an oldest GOP in the video PES if the PCR is greater than the PTS of a second oldest GOP stored in the buffer, and empties the buffer of each PES packet of the audio PES that has a PTS that is less than the PTS of the oldest GOP stored.