[Linphone-developers] Bug? RTP/H263+ P-bit

[nicolas.canceill]

Hello every one,

Using Linphone, I have noticed a strange behavior from the H.263+ (1998) 
packetization mode over RTP.

I am not an expert: both the ITU-T H.263 recommendation and the RFC4629 (RTP 
format) are a little complex 
for me. However, I am pretty sure of one thing:

In a H.263/RTP packet, if the P-bit of the H.263 header is 1, then the H.263 
payload must start with either a 
Picture Start Code, a Group-Of-Blocks Start Code, a Slice Start Code, an End Of 
Sequence, or an End Of 
SubBitStream; and from this start code must be removed the first two 0 (zero) 
bytes.

My problem is: while using H.263+, Linphone seems to break this rule, and sends 
packets with a P-bit value of 
1 but with a payload starting with a 0 (zero) byte, which is none of the 
expected start codes.

For example, in a packet starting at a picture start code: 04 00 as a header, 
and a payload starting with 80, 81, 
82, or 83 (in hexadecimal).
But I keep finding 00 04 headers followed by 00: is this a bug? otherwise, how 
is it compliant with the 
specification?

Thanks for any help.

Best regards,

Nicolas


Une messagerie gratuite, garantie à vie et des services en plus, ça vous tente ?
Je crée ma boîte mail www.laposte.net