[Mp4-tech] [H.264] Use of SPS/PPS in H.264

[Gary Sullivan]

It is certainly legitimate for particular application environments to
choose among the profiles and levels of a standard.  And sometimes we do
see other application-specific constraints being imposed (e.g., in the
DVD format specs).  But from an MPEG point of view, this sort of thing
is somewhat discouraged -- it should at least be kept to a minimum.  The
desire is to have the standard work across different application domains
as much as possible.  If each application domain imposes a bunch of
picky detailed special constraints, then we may end up with
interoperability problems.  For example, transferring video from your
DVD storage device or from broadcast television or from your camcorder
or webcam or videoconferencing system to your DVR or PC or mobile video
player (after making sure you have the proper legal right to do that, of
course) could become an unnecessarily bothersome, confusing, and
unpredictable hassle.  This is particularly an issue with
application-domain constraints that would be difficult to detect (e.g.,
at the sequence level --  Do you really think it is OK to need to scan
through the bitstream and count the number of PPSs?  Do you really want
to need a "Paris" scene detector in a video copying program?).
Best Regards,
Gary Sullivan
________________________________
	From: mp4-tech-bounces lists.mpegif.org
[mailto:mp4-tech-bounces lists.mpegif.org] On Behalf Of Dave Singer
	Sent: Thursday, June 22, 2023 9:56 AM
	To: arun menon; Steinberg Erez-RM96322
	Cc: mp4-tech lists.mpegif.org
	Subject: Re: [Mp4-tech] [H.264] Use of SPS/PPS in H.264
	Don't forget that although the AVC core specification doesn't
place limits on the number of SPS/PPS, how often they change etc., it's
perfectly legitimate for an environment that *uses* AVC to limit
	* profiles
	* levels
	* changes in visual size
	* pixel aspect ratios, color spaces, etc.
	* numbers of SPS/PPS
	* whether updates of SPS/PPS are out of band or in band in the
video stream, if they are allowed
	* whether 'Paris' appears in the video sequence
	and so on.
	At 9:58  +0000 22/06/06, arun menon wrote:
		Hi Erez,
		Please my comments below.
		Thanks,
		Arun
		On Wed, 21 Jun 2023 Steinberg Erez-RM96322 wrote :
		>Hi,
		>
		>H.264 allows up to 32 sequence parameter sets (SPS) and
256 picture
		>parameter sets (PPS) in a stream.
		>
		>An SPS includes information on the picture resolution,
number of
		>reference frame, profile and level. These are cardinal
for decoder
		>initialization and setup.
		>
		>My questions:
		>
		>(1) Is a decoder expected to handle spontaneous changes
in these
		>parameters? I.e., if an IDR picture is received with a
different
		>PPS/SPS, is it supposed to reallocate all its memory
buffers? (This is
		>effectively resetting the decoder)
---------------------------------------------------------------------
		An example I can think of when the decoder will receive
SPS and PPS information with different picture resolution could be
during playing of live content. If at any time the encoder finds that
the network bandwidth is reduced due to some reason, it could encode and
send smaller frame size. In this case you should reallocate the memory
buffers used for the frames. Alternatively you can allocate your memory
buffers with the max size supported to prevent the freeing and
reallocation.
----------------------------------------------------------------------
		>
		>(2) For practical implementations -- are there any
known applications
		>that use different SPSs in the same stream? Or, can a
decoder assume
		>only one SPS?
		>
		  It would be wrong if the decoder assumes that there
would be only one SPS or PPS because you do get multiple SPS and PPS in
some streams. But there is also the possibility that sometimes the same
information in SPS and PPS is sent again, instead of some thing
different (I have encountered a few such clips).
		>
		>Thanks for your help!
		>
		>Regards,
		>Erez
		>
		>_______________________________________________
		>NOTE: Please use clear subject lines for your posts.
Include [audio, [video], [systems], [general] or another apppropriate
identifier to indicate the type of question you have.
		>
		>Note: Conduct on the mailing list is subject to the
Antitrust guidelines found at
http://www.mpegif.org/public/documents/vault/mp-out-30042-Antitrust.php
<http://adworks.rediff.com/cgi-bin/AdWorks/sigclick.cgi/www.rediff.com/s
ignature-home.htm/1507191490 Middle5?PARTNER=3> 
		_______________________________________________
		NOTE: Please use clear subject lines for your posts.
Include [audio, [video], [systems], [general] or another apppropriate
identifier to indicate the type of question you have.
		Note: Conduct on the mailing list is subject to the
Antitrust guidelines found at
http://www.mpegif.org/public/documents/vault/mp-out-30042-Antitrust.php
	-- 
	David Singer
	Apple Computer/QuickTime
-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/mp4-tech/attachments/20060623/8cd6a0c7/attachment-0001.html