Abstract

A method and system for synchronizing audio and video data signals that are in a signal format wherein the audio and video data signals are not interleaved but originate from separate sources and arrive at a constant frame rate. In one embodiment, the audio and video data signals are in an Moving Pictures Expert Group (MPEG) format. Portions of the novel system operate within the hardware of a video processor unit and portions of the novel system operate within a software driver of a computer system that receives the audio and video data signals. Audio and video data signals are referenced by a presentation time stamp (e.g., VPTS for video and APTS for audio) encoded in the data. Using these stamps, the novel system (via the video processor unit) determines if the video signal is ahead or behind the audio signal by a small number of frames, x, and if so, the video processor unit either pauses or skips frames in its video display unit hardware to regain synchronization. If the video processor unit determines that the video signal is ahead or behind the audio signal by more than x frames, a software driver operating on the computer system is informed of the difference and the software driver instructs the video processor unit to skip or pause frames within its system parsing unit hardware using a host interrupt driven mechanism to regain synchronization. Using this method, the digital video signal is effectively synchronized to the reference digital audio signal.