Log in

View Full Version : Catch 22 in AVC syntax


Guest
5th February 2009, 03:30
I stumbled upon this while working on my restreaming application and thought it might be interesting to those interested in AVC syntax.

An SPS is activated by the first picture that references it. Prior to activation, an SPS is not available for interpretation of SEI messages. That means that (for example) a pic timing SEI that precedes the first picture cannot be accepted because its SPS has not yet been activated when the SEI is processed. I wonder if the JVT is aware of this problem.

Manao
5th February 2009, 07:08
Read 7.4.1.2.1 : When a sequence parameter set RBSP (with a particular value of seq_parameter_set_id) is not already active and it is
referred to by activation of a picture parameter set RBSP (using that value of seq_parameter_set_id) or is referred to by
an SEI NAL unit containing a buffering period SEI message (using that value of seq_parameter_set_id), it is activated.

Guest
5th February 2009, 07:29
A buffering period SEI message is not a picture timing SEI message! My point is that the latter gets rejected. You can easily see that this happens using the JM reference code.

Trahald
5th February 2009, 16:53
When an SEI NAL unit containing a buffering period SEI message is present, the buffering period SEI message
shall be the first SEI message payload of the first SEI NAL unit in the access unit

ive never seen an HRD bitstream that didnt have buffering period messages paired with the IDRs and preceeding the picture timing sei message. (perhaps im not understanding what you mean)

Guest
5th February 2009, 17:53
Specification of pulldown does not necessarily require HRD info as far as I can tell. Are you aware of such a requirement?

Trahald
6th February 2009, 12:16
i havent (although honestly i cant find much on pulldown and h264). but you can just call a buffering period message with both queries within set to 0 (if you didnt already know).

Trahald
6th February 2009, 12:35
It seems hrd_parameters are not needed for HDMV bluray streams. Really without hd-dvd around there should be no need for HRD (and the x264 hrd patch for that matter.) HRD complicated the process unnecessarily... good riddance.

lexor
6th February 2009, 17:45
It seems hrd_parameters are not needed for HDMV bluray streams. Really without hd-dvd around there should be no need for HRD (and the x264 hrd patch for that matter.) HRD complicated the process unnecessarily... good riddance.

Than does the HDR patch do something else? Because without it we don't get Blu-ray compatibility.

Trahald
6th February 2009, 18:50
Than does the HDR patch do something else? Because without it we don't get Blu-ray compatibility.

it adds timing info . the other hypothetical reference decoder parameters (cpb dpb params etc) are supposedly not needed.