> On Dec 14, 2024, at 17:43, Anton Khirnov <an...@khirnov.net> wrote:
>
> Quoting Zhao Zhili (2024-12-11 05:23:33)
>> From: Zhao Zhili <zhiliz...@tencent.com>
>>
>> ---
>> libavcodec/hevc/hevcdec.c | 75 ++++++++++++++++++++++++++++++++++++++-
>> libavcodec/hevc/hevcdec.h | 2 ++
>> libavcodec/hevc/refs.c | 10 +++++-
>> 3 files changed, 85 insertions(+), 2 deletions(-)
>>
>> diff --git a/libavcodec/hevc/hevcdec.c b/libavcodec/hevc/hevcdec.c
>> index be35a9de82..782c9382bc 100644
>> --- a/libavcodec/hevc/hevcdec.c
>> +++ b/libavcodec/hevc/hevcdec.c
>> @@ -458,6 +458,26 @@ static int export_multilayer(HEVCContext *s, const
>> HEVCVPS *vps)
>> return 0;
>> }
>>
>> +int ff_hevc_is_alpha_video(const HEVCContext *s) {
>> + const HEVCVPS *vps;
>> + int ret = 0;
>> +
>> + if (!s->vps)
>> + return 0;
>
> This seems like something that should not happen.
>
>> + vps = s->vps;
>> + if (vps->nb_layers != 2 || !vps->layer_id_in_nuh[1])
>> + return 0;
>> +
>> + ret = (s->vps->scalability_mask_flag & HEVC_SCALABILITY_AUXILIARY) &&
>> + s->sei.alpha.has_alpha_channel_info;
>
> Is the SEI really necessary? I thought it's enough to have an auxiliary
> picture with AuxId=AUX_ALPHA.
>
V2 dropped the SEI patch.
> Also, I'd prefer the interaction with multiview to be clearer, with e.g.
> a warning message when both are present, and fewer assumptions about
> only one of them being present spread over the code.
V2 added a warning message in vps extension decode.
x265 set both MULTIVIEW and AUXILIARY for alpha video, which should be a
mistake.
I don’t know how to create a multiview alpha video.
>
> --
> Anton Khirnov
> _______________________________________________
> ffmpeg-devel mailing list
> ffmpeg-devel@ffmpeg.org <mailto:ffmpeg-devel@ffmpeg.org>
> https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
>
> To unsubscribe, visit link above, or email
> ffmpeg-devel-requ...@ffmpeg.org <mailto:ffmpeg-devel-requ...@ffmpeg.org> with
> subject "unsubscribe".
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".
