[x264-devel] proposal: nal-hrd and interlacing

Yufei Yuan yfyuan at gmail.com
Wed Jul 20 17:15:08 CEST 2011


The NAL-HRD settings has nothing to do the display order of the
decoded interlaced fields. Assuming that your STB is not the one to
blame (that's probably why you asked on this mailing list), you need
to find some way to analyze the stream, and see if the following
information is present and correct in the stream.

1. The 'delta_pic_order_cnt' related fields in the slice header, in
your case this information alone should be enough to identify which
field, top or bottom, should be displayed first.
2. The 'pic_struct_present_flag' in the VUI, and the corresponding
'pic_struct' in the picture timing SEI message. This would also
provide display order information and it is needed to handle frame
rate conversion cases, like 3:2 pull-down.

If x264 handles these parameters/fields correctly, then it's the STB's fault.

One thing for sure is that NAL-HRD has nothing to do with the issue.

Regards,


On Wed, Jul 20, 2011 at 3:59 AM, Erik Slagter <erik at slagter.name> wrote:
>> Does the same thing happen for progressive material?
>> This should be an issue with general playback. It shouldn't have anything
>> to do with interlaced.
>
> No because it is interlacing specific.
>
> To recap: without using nal-hrd the material is played with the wrong field
> order (bottom field first), although the option -tff has been used. When
> nal-hrd is used, the playback field order is correct. This has been
> suggested by some expert, I didn't see the relation but it clearly works.
>
> So that's why I'd like to have x264 (or the lib) issue a warning when
> interlacing is selected but no nal-hrd, as to prevent others to shoot them
> in their foot as well.
>
> I can't find hardly any information on nal-hrd, so I cannot judge if this is
> expected behaviour. I guess that settopboxes tend to be quite pedantic on
> streams adhering to standards.
>
>
> _______________________________________________
> x264-devel mailing list
> x264-devel at videolan.org
> http://mailman.videolan.org/listinfo/x264-devel
>
>


More information about the x264-devel mailing list