[FFmpeg-devel] [PATCH v2 1/4] avcodec/h2645_vui: Ensure color primaries/trc/space isn't reserved value

[Zhao Zhili]

> On May 27, 2025, at 11:07, Zhao Zhili <quinkblack-at-foxmail.com at ffmpeg.org> wrote:
> 
> 
> 
>> On May 27, 2025, at 02:48, Kieran Kunhya via ffmpeg-devel <ffmpeg-devel at ffmpeg.org> wrote:
>> 
>> On Mon, May 26, 2025 at 8:13 PM Zhao Zhili
>> <quinkblack-at-foxmail.com at ffmpeg.org> wrote:
>>> 
>>> 
>>> 
>>>> On May 22, 2025, at 13:06, Zhao Zhili <quinkblack-at-foxmail.com at ffmpeg.org> wrote:
>>>> 
>>>> From: Zhao Zhili <zhilizhao at tencent.com>
>>>> 
>>>> Fix error reported by swscaler:
>>>> Unsupported input (Operation not supported): fmt:yuv420p csp:unknown prim:reserved trc:bt709 -> fmt:yuv420p csp:bt709 prim:reserved trc:bt709
>>>> —
>>> 
>>> 
>>> Will apply soon.
>> 
>> Surely this should be filtered in swscale?
>> I think we need to pass through future reserved values of csp and stuff.
> 
> Both of H.264 and H.265 spec says:
> 
>> Decoders shall interpret reserved values of colour_primaries as equivalent to the value 2.
> 
> The code matches the spec.
> 
> There is another similar topic in 2024
> 
> https://ffmpeg.org/pipermail/ffmpeg-devel/2024-December/337160.html
> 
> cc haasn. This is a regression. Should the fixes go into swscale/avfilter, or the decoder, or both?
> 
> Pass through reserved value of csp has its use case. However, it’s more common that those samples are generated by buggy software.

Will apply if no more comments.

> 
>> 
>> Kieran
>> _______________________________________________
>> 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".
> 
> _______________________________________________
> 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".