[AVT] FEC draft

Hi folks,

We would like to discuss the following changes to the FEC draft with
you, and would like to hear your comments.

As decided in previous meetings and discussions, we are in the process
of updating and incorporating the ULP FEC draft with the generic FEC. At
the same time, taking this chance, we would also like to modify a few
issues with RFC 2733. In RFC 2733, the M, P, X, and CC bits in the RTP
header of the FEC packets are calculated from the parity operation, thus
those fields do not confirm to their normal definition as in RFC 1889.

The proposed change to fix this issue is to move the parity data of the
M, P, X, and CC bits into the FEC header part of the FEC packet from the
RTP header, so the corresponding fields in the RTP header can retain
their original meaning.

One of the potential solutions is as below. (E = extension, R =
reserved)

  +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
  |      SN base                  |        length recovery        |
  +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
  |M| PT recovery |P|X|  CC   |E|R|             mask              |
  +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
  |                          TS recovery                          |
  +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

With the addition of M, P, X, and CC bits in the FEC header, one extra
octet is taken. This reduces the mask from 24 to 16 bits. This might
still be enough, as the typical interleaving is something like 10
packets.  

One other potential solution is pretty much as above, but keeping the
mask at 24 bits as it is in RFC 2733. However, by doing that, the FEC
header above will not align on a 32 bit boundary, which might be a good
property to keep.

So the question we would like to hear your comment is: Shall we (a)
reduce the mask to 16-bit, or (b) keep the mask at 24-bit, and break the
32-bit alignment?

Your help is very much appreciated.

Adam

_______________________________________________
Audio/Video Transport Working Group
avt@ietf.org
https://www1.ietf.org/mailman/listinfo/avt