Video frames from a sequence of video frames and corresponding audio are received and processed by a media player. A first portion of the sequence depends on fewer reference frames then a second portion of the sequence. The number of reference frames per frame is limited. The first and second portions of the sequence are treated differently. The first portion is received at a lower data rate, with greater security and processed with a more robust protocol stack than the second portion. The first and second portions may be stored in a single queue and/or separate queues. The first portion of frame data is utilized to compensate for lost frames by repeating and/or interpolating video frames. The media player constructs a local video frame and synchronizes corresponding audio content with it. The received, repeated and/or interpolated video frames are assembled into display order and decoded.