RE: [AVT] VC1 payload last call

Miska,

> What is the guideline for the relation of AU payloads and field
pictures
> - one field picture per AU payload or both field pictures of an
> interlaced frame in an AU payload? Please clarify. 

For interlaced video, a video frame can be coded as two fields.  In that
case, a VC-1 frame contains a picture header followed by a field header
for the first field picture, data for the first field picture (slices,
macroblocks, etc.), and then the field header for the second field
picture, followed by data for the second field picture.

The important thing to note is that the two fields are part of the same
frame.  Since I have written in the Internet-Draft there is one frame
per AU, I think that should already answer your question.  The two
fields will be in the same AU payload, unless the frame is so big that
it needs to be fragmented into multiple RTP packets.

I would prefer to not complicate the Internet-Draft with detailed
discussions about interlaced video.  But I suppose I could add a simple
clarifying sentence, like this:

"Note that in the case of interlaced video, the VC-1 frame consists of
two fields that may be coded as separate pictures.  The two pictures
still belong to the same VC-1 frame."

> In case of field
> pictures/frames, please clarify which headers are allowed in an AU
> payload.

If I add the proposed sentence above, I think it clarifies the relation
between fields and frames, without the need to explain that there is one
picture header and two field headers.  Do you agree?

> How is the presentation time of an interlaced frame defined? Please
> clarify.

The Internet-Draft says that the presentation time of a video frame
differs from the sampling instant of the video frame only by an
arbitrary constant offset.  So, this means that if the frame consists of
two fields, the presentation time of the frame is the sampling instant
of the field that will be presented first (i.e., the field that was
sampled first).

I don't think that there is any ambiguity in the Internet-Draft about
this.  But I don't mind adding a clarifying sentence along the lines of
the paragraph above.

> I suppose the Buffer Fullness (BF)
> syntax element of the picture header could be used for this purpose.
> Could you add some sentences in section 3.3 (e.g. an informative note)
> to clarify this issue. For example, add a new paragraph to the end of
> section 3.3:
> 
> Informative note: The initial buffer fullness for an I-picture in the
> Simple and Main Profile is signaled with the Buffer Fullness (BF)
syntax
> element that is present in the I picture header.

You are right, SMPTE 421M defines a BF syntax element for Simple and
Main profiles, which specifies the initial buffer fullness.  It doesn't
affect the correctness of the RTP payload format specification itself,
but it might be nice to mention BF anyway, for completeness.  

This comment shows that you have been reading SMPTE 421M carefully.
That's great. :-)

Anders

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