[FFmpeg-devel] [PATCH 1/2] ffmpeg: handling copyts use case in forced key frames functionality

[Dixit, Vishwanath]

On 4/27/18 6:24 AM, Michael Niedermayer wrote:
> On Thu, Apr 26, 2018 at 10:33:59AM +0530, vdixit at akamai.com wrote:
>> From: Vishwanath Dixit <vdixit at akamai.com>
>>
>> Forced key frames creation functionality was assuming the first PTS
>> value to be zero, but, when 'copyts' is enalbed, the first PTS can
>> be any big number. This was eventually forcing all the frames as
>> key frames. To overcome this issue, the actual first PTS value has
>> to be considered.
>> ---
>>  fftools/ffmpeg.c     | 5 ++++-
>>  fftools/ffmpeg_opt.c | 1 +
>>  2 files changed, 5 insertions(+), 1 deletion(-)
>
> breaks fate
> --- ./tests/ref/fate/pva-demux	2018-04-26 01:45:19.589231218 +0200
> +++ tests/data/fate/pva-demux	2018-04-27 02:50:51.828844428 +0200
> @@ -34,3 +34,18 @@
>  1,      47520,      47520,     2160,      384, 0xfddbb7df
>  1,      49680,      49680,     2160,      384, 0xbbb2af0c
>  1,      51840,      51840,     2160,      384, 0x8f03b5fc
> +0,         26,         26,        1,   470016, 0xead57e3f
> +0,         27,         27,        1,   470016, 0x787085d0
> +0,         28,         28,        1,   470016, 0x3d529159
> +0,         29,         29,        1,   470016, 0xed32986b
> +0,         30,         30,        1,   470016, 0xcdd2c5a7
> +0,         31,         31,        1,   470016, 0xb271d212
> +0,         32,         32,        1,   470016, 0x838cad86
> +0,         33,         33,        1,   470016, 0xdb992f3f
> +0,         34,         34,        1,   470016, 0x88779a1e
> +0,         35,         35,        1,   470016, 0x6a73db7b
> +0,         36,         36,        1,   470016, 0x7b4e6ace
> +0,         37,         37,        1,   470016, 0x72d14489
> +0,         38,         38,        1,   470016, 0xd5b17c4d
> +0,         39,         39,        1,   470016, 0x1f62ed78
> +0,         40,         40,        1,   470016, 0x97f98fff
> Test pva-demux failed. Look at tests/data/fate/pva-demux.err for details.
> make: *** [fate-pva-demux] Error 1
>
The patch was affecting the 'recording time' (-t) functionality. Sorry for not conducting fate tests before submitting the patch. I have updated the patch and have run the fate tests. Apart from the issue fix updates, I have modified the commit name and message for better clarity. Please find the revised version here https://patchwork.ffmpeg.org/patch/8814/ .

>
> [...]
>
>
> _______________________________________________
> ffmpeg-devel mailing list
> ffmpeg-devel at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/ffmpeg-devel