Wireshark-bugs: [Wireshark-bugs] [Bug 11338] Associate the file-mp4 dissector with the audio/mp4
Comment # 8
on bug 11338
from Tyson Key
(In reply to Martin Kaiser from comment #7)
> (In reply to Tyson Key from comment #5)
> > Created attachment 13711 [details]
> > Another trace file containing a payload from a different encoder, Windows
> > Media Player, and UPMPDCLI
> >
> > Unsure if it's the iTunes MPEG-4/AAC encoder, or a bug in the control point,
> > renderer, or server (or a bug in the dissector, or reassembly routines), but
> > it seems that the MPEG-4 payload in the last trace file doesn't get
> > dissected, despite being identified as audio/mp4.
> >
> > This new trace should properly exercise this addition, hopefully.
>
> The MP4 dissector rejects the packet if it doesn't start with a known box
> type. The first box is a "gnre' box. I didn't know about this type as I only
> checked the ISOBMFF documents.
>
> It appears in the list at
>
> http://www.mp4ra.org/atoms.html
>
> If I add this box type, the MP4 packet is recognized. However, the remaining
> boxes don't line up nicely.
>
> According to http://xhelmboyx.tripod.com/formats/mp4-layout.txt
>
> Apple defined some boxes whose name starts with 0xa9, followed by three
> letters...
On that note, I've noticed that for some files (especially the iTunes ones),
dissection stops after the "mdat" box - even though reassembly seems to be
successful, in the case of both code "206 Partial Content", and "200 OK"
responses. Could this be related to the same issue?
You are receiving this mail because:
- You are watching all bug changes.
