Re: [FFmpeg-devel] [PATCH] ffmpeg: add 100ms delay for first stats

2020-12-23 Thread Gyan Doshi
On 23-12-2020 04:03 pm, Nicolas George wrote: Gyan Doshi (12020-12-23): It's    void av_dump_format(...) Seems too trivial to change it , or add + initialize yet another global variable just for this. Exactly. Dumping the output is a global affair, it can warrant a global variable.

Re: [FFmpeg-devel] [PATCH] ffmpeg: add 100ms delay for first stats

2020-12-23 Thread Nicolas George
Gyan Doshi (12020-12-23): > It's    void av_dump_format(...) > > Seems too trivial to change it , or add + initialize yet another global > variable just for this. Exactly. Dumping the output is a global affair, it can warrant a global variable. Something like this: of->header_written = 1;

Re: [FFmpeg-devel] [PATCH] ffmpeg: add 100ms delay for first stats

2020-12-23 Thread Gyan Doshi
On 23-12-2020 03:53 pm, Nicolas George wrote: Gyan Doshi (12020-12-23): That's not the issue. It's the mangled line - the output from lavf/dump.c. mangled with stats line AV_NOPTS_VALUE is an acceptable value and already checked for at line 1815. As is <=0 output size. Indicating to the

Re: [FFmpeg-devel] [PATCH] ffmpeg: add 100ms delay for first stats

2020-12-23 Thread Nicolas George
Gyan Doshi (12020-12-23): > That's not the issue. It's the mangled line - the output from lavf/dump.c. > mangled with stats line > > AV_NOPTS_VALUE is an acceptable value and already checked for at line 1815. > As is <=0 output size. > > Indicating to the user that ffmpeg processing has stalled

Re: [FFmpeg-devel] [PATCH] ffmpeg: add 100ms delay for first stats

2020-12-23 Thread Gyan Doshi
On 23-12-2020 03:35 pm, Nicolas George wrote: Gyan Doshi (12020-12-23): Which piece of info or object would indicate that? At the stage of this check, I don't see any available. Read Michael's mail: Output #0, avi, to 'file.avi':= 0kB time=-577014:32:22.77 bitrate= -0.0kbits/s

Re: [FFmpeg-devel] [PATCH] ffmpeg: add 100ms delay for first stats

2020-12-23 Thread Nicolas George
Gyan Doshi (12020-12-23): > Which piece of info or object would indicate that? At the stage of this > check, I don't see any available. Read Michael's mail: Output #0, avi, to 'file.avi':= 0kB time=-577014:32:22.77 bitrate= -0.0kbits/s speed=N/A The problem is obviously the time. The

Re: [FFmpeg-devel] [PATCH] ffmpeg: add 100ms delay for first stats

2020-12-23 Thread Gyan Doshi
On 23-12-2020 03:12 pm, Nicolas George wrote: Gyan Doshi (12020-12-23): Avoids breaking output file dump report. --- fftools/ffmpeg.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) This looks broken. Why 100ms? Because on your setup, 80ms is not enough but 100ms is? On a slower

Re: [FFmpeg-devel] [PATCH] ffmpeg: add 100ms delay for first stats

2020-12-23 Thread Nicolas George
Gyan Doshi (12020-12-23): > Avoids breaking output file dump report. > --- > fftools/ffmpeg.c | 3 ++- > 1 file changed, 2 insertions(+), 1 deletion(-) This looks broken. Why 100ms? Because on your setup, 80ms is not enough but 100ms is? On a slower system, it would need to be 120 or 200. Test

Re: [FFmpeg-devel] [PATCH] ffmpeg: add 100ms delay for first stats

2020-12-23 Thread Gyan Doshi
On 23-12-2020 10:28 am, Gyan Doshi wrote: Avoids breaking output file dump report. --- fftools/ffmpeg.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/fftools/ffmpeg.c b/fftools/ffmpeg.c index 2c0820aacf..449da484d9 100644 --- a/fftools/ffmpeg.c +++ b/fftools/ffmpeg.c

[FFmpeg-devel] [PATCH] ffmpeg: add 100ms delay for first stats

2020-12-22 Thread Gyan Doshi
Avoids breaking output file dump report. --- fftools/ffmpeg.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/fftools/ffmpeg.c b/fftools/ffmpeg.c index 2c0820aacf..449da484d9 100644 --- a/fftools/ffmpeg.c +++ b/fftools/ffmpeg.c @@ -1699,7 +1699,8 @@ static void