Re: [FFmpeg-devel] AVPixelFormat for YUYV 10 bit.

2019-03-25 Thread Yufei He
It's a standard format. I suggest that it should be in AVPixelFormat.  
the codec can do the conversion.

Yufei.


On 03/25/2019 09:55 AM, Kieran Kunhya wrote:
> On Mon, 25 Mar 2019 at 13:53 Yufei He  wrote:
>
>> Hi
>>
>> I think there may be one format missing in AVPixelFormat.
>>
>> AV_PIX_FMT_YUYV210
>>
> For various reasons v210 and friends are implemented as a codec in FFmpeg.
>
> Kieran
> ___
> 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".
___
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".

Re: [FFmpeg-devel] AVPixelFormat for YUYV 10 bit.

2019-03-25 Thread Devin Heitmueller
Hello Yufei,


> On Mar 25, 2019, at 9:53 AM, Yufei He  wrote:
> 
> Hi
> 
> I think there may be one format missing in AVPixelFormat.
> 
> AV_PIX_FMT_YUYV210
> 
> 
> [cid:part1.002B7405.3138ACFB@matrox.com]
> 
> 
> The start of each line in the V210 video buffer format must be
> aligned to a multiple of 48 pixels (128 bytes). This means that if a line is 
> not a
> multiple of 48, each line of video data must be padded out to the nearest 48 
> pixel
> boundary. For example, a 1280 × 720 video buffer is 1280 pixels wide, which 
> is not a
> multiple of 48. Each line of the video buffer must be padded to 1296 pixels 
> (3456
> bytes) in order to make each line a multiple of 48.
> 
> It's popular when video is in 10bit.
> 

So V210 video is supported, but not as a pixel format.  It’s treated as a 
packet format.  Hence in order to use v210 video, you have to pass the video 
frame through the v210enc or v210 decoder module, at which point you end up 
with video in AV_PIX_FMT_YUV422P10 format.

We can argue about whether that was the right approach, but that’s what is done 
in ffmpeg today, and we use it regularly with the Blackmagic cards (i.e. you 
can look at libavdevice/decklink_enc.cpp, decklink_dec.cpp for usage).

Devin

---
Devin Heitmueller - LTN Global Communications
dheitmuel...@ltnglobal.com

___
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".

Re: [FFmpeg-devel] AVPixelFormat for YUYV 10 bit.

2019-03-25 Thread Kieran Kunhya
On Mon, 25 Mar 2019 at 13:53 Yufei He  wrote:

> Hi
>
> I think there may be one format missing in AVPixelFormat.
>
> AV_PIX_FMT_YUYV210
>

For various reasons v210 and friends are implemented as a codec in FFmpeg.

Kieran
___
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".

[FFmpeg-devel] AVPixelFormat for YUYV 10 bit.

2019-03-25 Thread Yufei He
Hi

I think there may be one format missing in AVPixelFormat.

AV_PIX_FMT_YUYV210


[cid:part1.002B7405.3138ACFB@matrox.com]


The start of each line in the V210 video buffer format must be
aligned to a multiple of 48 pixels (128 bytes). This means that if a line is 
not a
multiple of 48, each line of video data must be padded out to the nearest 48 
pixel
boundary. For example, a 1280 × 720 video buffer is 1280 pixels wide, which is 
not a
multiple of 48. Each line of the video buffer must be padded to 1296 pixels 
(3456
bytes) in order to make each line a multiple of 48.

It's popular when video is in 10bit.

Yufei.





___
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".