I have been working on a project parsing and transmitting mp3 files over a network to a hardware decoder. The MP3 format specifies that frames may not be independent due to 'byte-reservoir or back data'; where data for a one frames lies in the proceeding frame.
I can find no reference to this in the ASF specification; but was wondering if a decoder can decode a data object (or frame or packet etc..) on it's own, or whether it needs several. And if so where is the information relating to this stored?
Thanks
Tim
Looking at the decoder source, it seems like each packet contains 1 superframe. Looking at the superframe init code, a superframe is equal to a frame if there is no bit reservoir. So it looks like individual superframes/packets should be independent of each other, but I'm not certain.