Add support for packed audio

[dmorn]

This is a variant of AAC used in audio-only variant streams of HLS (https://datatracker.ietf.org/doc/html/draft-pantos-http-live-streaming-19#page-6). In short, the file is prepended with a series of ID3v2 tags; one of those declares the timestamp (in MPEG-TS terms) of reference for the packets in that file.

When this issue is closed, the parser is supposed to be able to parse packed audio files by parsing each ID3v2 tag it encounters, shifting the Buffer's PTS when needed.