[FFmpeg-devel] [PATCH 5/8] avcodec/av1dec: parse dimensions from the sequence header in extradata

[Mark Thompson]

On 25/09/2020 15:43, James Almer wrote:
> Signed-off-by: James Almer <jamrial at gmail.com>
> ---
>   libavcodec/av1dec.c | 10 ++++++++++
>   1 file changed, 10 insertions(+)
> 
> diff --git a/libavcodec/av1dec.c b/libavcodec/av1dec.c
> index 0bb04a3e44..f6b9fbbac3 100644
> --- a/libavcodec/av1dec.c
> +++ b/libavcodec/av1dec.c
> @@ -405,6 +405,9 @@ static av_cold int av1_decode_free(AVCodecContext *avctx)
>   static int set_context_with_sequence(AVCodecContext *avctx,
>                                        const AV1RawSequenceHeader *seq)
>   {
> +    int width = seq->max_frame_width_minus_1 + 1;
> +    int height = seq->max_frame_height_minus_1 + 1;
> +
>       avctx->profile = seq->seq_profile;
>       avctx->level = seq->seq_level_idx[0];
>   
> @@ -423,6 +426,13 @@ static int set_context_with_sequence(AVCodecContext *avctx,
>           break;
>       }
>   
> +    if (avctx->width != width || avctx->height != height) {
> +        int ret = ff_set_dimensions(avctx, width, height);
> +        if (ret < 0)
> +            return ret;
> +    }
> +    avctx->sample_aspect_ratio = (AVRational) { 1, 1 };
> +
>       if (seq->timing_info.num_units_in_display_tick &&
>           seq->timing_info.time_scale) {
>           av_reduce(&avctx->framerate.den, &avctx->framerate.num,
> 

Can you explain more about why this might be wanted?  There is no requirement that the stream actually contains any frames with the size stated in the sequence header.

- Mark