[FFmpeg-devel] setting vps/sps/pps id via hevc_metadata

[Eran Gonen]

Would it make more sense if I'll change it to advance sps id? offering no
risk for multiple ps

On Tue, Jan 7, 2020 at 3:54 PM Eran Gonen <eran.gonen at cloudinary.com> wrote:

> Hi,
> This filter allows editing hevc headers and change metadata. Most
> modifications doesn't make sense for all multiple PS streams (even if they
> are decodable): same crop for different resolutions, the same tick_rate for
> different FR. Changing the PS id allows a wide range of stream compositions
> and offer no decoding risk for most streams.
>
> On Mon, Jan 6, 2020 at 11:48 PM Mark Thompson <sw at jkqxz.net> wrote:
>
>> On 06/01/2020 19:24, Eran Gonen wrote:> On Mon, Jan 6, 2020 at 9:05 PM
>> Andreas Rheinhardt <
>> > andreas.rheinhardt at gmail.com> wrote:
>> >
>> >> On Mon, Jan 6, 2020 at 7:54 PM Eran Gonen <eran.gonen at cloudinary.com>
>> >> wrote:
>> >>
>> >>> Hi,
>> >>>
>> >>> The attached patch allows setting vps/sps/pps id
>> >>>
>> >>> Usage:
>> >>> /ffmpeg -i source -c:v libx265 -bsf:v
>> >>> hevc_metadata=sps_id=8:pps_id=12:vps_id=3 output
>> >>>
>> >>>
>> >> There is a problem with this: Input files with parameter sets of the
>> same
>> >> kind with different ids can be broken by this (because they would all
>> share
>> >> the same id in the output and hence overwrite each other upon
>> decoding).
>> >>
>> >> - Andreas
>> >>
>> > Thanks Andreas. I also thought about this point.
>> > I guess the same goes to the rest of parameters in hevc_metadata.
>> > They all affect all sets in the file. Maybe that's the intention of the
>> > user for such files.
>>
>> I'm not sure that comparison is really right, because the existing
>> elements which can be modified are the metadata which don't affect the
>> decoding process.  Changing the PS IDs will break most streams with
>> multiple parameter sets (try PPS_A_qualcomm_7.bit from the conformance test
>> suite).
>>
>> What is your actual use-case here?  If it's fixing some particular set of
>> broken streams then perhaps it could be a separate BSF like
>> h264_redundant_pps.
>>
>> - Mark
>> _______________________________________________
>> ffmpeg-devel mailing list
>> ffmpeg-devel at ffmpeg.org
>> https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
>>
>> To unsubscribe, visit link above, or email
>> ffmpeg-devel-request at ffmpeg.org with subject "unsubscribe".
>
>