GES hangs on MPEG2 program stream files?

Chris Wine chriswine at gmail.com
Wed Feb 10 01:25:31 UTC 2021 

Thibault,

I wrote an element that wraps MainConcept's MPEGPS demuxer, and set the
rank high enough so it gets chosen by uridecodebin; I'm getting the same
messages in the log about wrong seqnums, so I'm starting to thing that the
problem isn't with the decoder or the demuxer. I'm kind of at a loss of
what to look at next. Any suggestions?

--Chris

On Mon, Feb 8, 2021 at 8:13 AM Chris Wine <chriswine at gmail.com> wrote:

> Hello Thibault,
>
> I see this same message in the log for both <video_nlecomposition0> and
> <audio_nlecomposition1>, so it seems like it would stand to reason that the
> demuxer is responsible for the bad sequence numbers. Is there any way I
> could validate that though setting GST_DEBUG to output more detailed
> logging?
>
> --Chris
>
> On Sat, Feb 6, 2021 at 2:42 PM Thibault Saunier <tsaunier at gnome.org>
> wrote:
>
>> Hi,
>>
>> > *0:00:11.110380900 92428 000000000386C840 INFO          nlecomposition
>> nlecomposition.c:2320:_is_ready_to_restart_task:<video_nlecomposition0>
>> WARNING: eos seqnum 1287 != wanted 652*
>>
>> Looks like a very good hint, we should figure out who is pushing an EOS
>> with the wrong seqnum.
>>
>> - Thibault
>>
>> On Fri, Feb 5, 2021 at 7:54 PM Chris Wine <chriswine at gmail.com> wrote:
>>
>>>  Hi Thibault,
>>>
>>> I'm seeing what looks like the issue right at the moment the video hangs
>>> up during playback:
>>>
>>> 0:00:10.893906600 92428 0000000007AFBD80 INFO            videodecoder
>>> gstvideodecoder.c:1310:gst_video_decoder_sink_event_default:<avdec_mpeg2video1>
>>> upstream tags: taglist, video-codec=(string)"MPEG-2\ Video\ \(Main\
>>> Profile\)", minimum-bitrate=(uint)5235884, maximum-bitrate=(uint)58581100,
>>> bitrate=(uint)24890275;
>>> 0:00:10.894304800 92428 0000000007AFBA00 INFO                    task
>>> gsttask.c:312:gst_task_func:<mpegpsdemux2:sink> Task going to paused
>>> *0:00:10.894364200 92428 0000000007AFBD80 WARN                validate
>>> gst-validate-reporter.c:198:gst_validate_report_valist: <multiqueue2:src_0>
>>> 3668 (warning) : event: a serialized event received should be pushed in the
>>> same 'time' as it was received : Serialized event eos wasn't pushed before
>>> expected timestamp 22:11:32.195722222 on pad multiqueue2:src_0*
>>> 0:00:10.894563000 92428 0000000007AFBD80 WARN            videodecoder
>>> gstvideodecoder.c:2762:gst_video_decoder_prepare_finish_frame:<avdec_mpeg2video1>
>>> decreasing timestamp (22:11:31.878733333 < 22:11:31.912100000)
>>> 0:00:10.910185800 92428 0000000007B07D80 INFO            d3dvideosink
>>> d3dhelpers.c:1953:d3d_render_buffer:<videosink-actual-sink-d3dvideo> Render
>>> 0:00:10.193516667
>>> 0:00:10.926574600 92428 0000000007B07D80 INFO            d3dvideosink
>>> d3dhelpers.c:1953:d3d_render_buffer:<videosink-actual-sink-d3dvideo> Render
>>> 0:00:10.210200000
>>> 0:00:10.943754500 92428 0000000007B07D80 INFO            d3dvideosink
>>> d3dhelpers.c:1953:d3d_render_buffer:<videosink-actual-sink-d3dvideo> Render
>>> 0:00:10.226883333
>>> 0:00:10.959988900 92428 0000000007B07D80 INFO            d3dvideosink
>>> d3dhelpers.c:1953:d3d_render_buffer:<videosink-actual-sink-d3dvideo> Render
>>> 0:00:10.243566667
>>> 0:00:10.976805500 92428 0000000007B07D80 INFO            d3dvideosink
>>> d3dhelpers.c:1953:d3d_render_buffer:<videosink-actual-sink-d3dvideo> Render
>>> 0:00:10.260250000
>>> 0:00:10.993280500 92428 0000000007B07D80 INFO            d3dvideosink
>>> d3dhelpers.c:1953:d3d_render_buffer:<videosink-actual-sink-d3dvideo> Render
>>> 0:00:10.276933333
>>> 0:00:11.010221900 92428 0000000007B07D80 INFO            d3dvideosink
>>> d3dhelpers.c:1953:d3d_render_buffer:<videosink-actual-sink-d3dvideo> Render
>>> 0:00:10.293616667
>>> 0:00:11.026865700 92428 0000000007B07D80 INFO            d3dvideosink
>>> d3dhelpers.c:1953:d3d_render_buffer:<videosink-actual-sink-d3dvideo> Render
>>> 0:00:10.310300000
>>> 0:00:11.043658900 92428 0000000007B07D80 INFO            d3dvideosink
>>> d3dhelpers.c:1953:d3d_render_buffer:<videosink-actual-sink-d3dvideo> Render
>>> 0:00:10.326983333
>>> 0:00:11.060436300 92428 0000000007B07D80 INFO            d3dvideosink
>>> d3dhelpers.c:1953:d3d_render_buffer:<videosink-actual-sink-d3dvideo> Render
>>> 0:00:10.343666667
>>> 0:00:11.076939700 92428 0000000007B07D80 INFO            d3dvideosink
>>> d3dhelpers.c:1953:d3d_render_buffer:<videosink-actual-sink-d3dvideo> Render
>>> 0:00:10.360350000
>>> 0:00:11.077507800 92428 000000000386C080 WARN                validate
>>> gst-validate-reporter.c:198:gst_validate_report_valist:
>>> <track-element-videorate:src> 3668 (warning) : event: a serialized event
>>> received should be pushed in the same 'time' as it was received :
>>> Serialized event eos wasn't pushed before expected timestamp
>>> 22:11:31.917133333 on pad track-element-videorate:src
>>> 0:00:11.093741900 92428 0000000007B07D80 INFO            d3dvideosink
>>> d3dhelpers.c:1953:d3d_render_buffer:<videosink-actual-sink-d3dvideo> Render
>>> 0:00:10.377033333
>>> 0:00:11.094189300 92428 000000000386C080 INFO                    task
>>> gsttask.c:312:gst_task_func:<queue0:src> Task going to paused
>>> 0:00:11.110188200 92428 0000000007B07D80 INFO            d3dvideosink
>>> d3dhelpers.c:1953:d3d_render_buffer:<videosink-actual-sink-d3dvideo> Render
>>> 0:00:10.393716667
>>>
>>>
>>> *0:00:11.110354600 92428 000000000386C840 INFO          nlecomposition
>>> nlecomposition.c:1320:ghost_event_probe_handler:<video_nlecomposition0> Got
>>> EOS, last EOS seqnum id : 0 current seq num is: 12870:00:11.110380900 92428
>>> 000000000386C840 INFO          nlecomposition
>>> nlecomposition.c:2320:_is_ready_to_restart_task:<video_nlecomposition0>
>>> WARNING: eos seqnum 1287 != wanted 6520:00:11.110398000 92428
>>> 000000000386C840 INFO          nlecomposition
>>> nlecomposition.c:1344:ghost_event_probe_handler:<video_nlecomposition0> Got
>>> an EOS but it seqnum 1287 != next eos seqnum 652*
>>> 0:00:11.110418300 92428 000000000386C840 INFO                    task
>>> gsttask.c:312:gst_task_func:<gessmartmixer0-compositor:src> Task going to
>>> paused
>>>
>>> It seems like it is something with the EOS event perhaps getting ignored
>>> by the compositor, but I don't understand the particular error messages
>>> about EOS seqnums. Do you have any further ideas?
>>>
>>> Thanks,
>>> --Chris
>>>
>>> On Fri, Jan 22, 2021 at 6:45 PM Thibault Saunier <tsaunier at gnome.org>
>>> wrote:
>>>
>>>> Hello,
>>>>
>>>> There indeed is an issue using that file with GES I do not know what is
>>>> going on and some debugging would be needed to determine it. Check some
>>>> GST_DEBUG logs to figure it out?
>>>>
>>>> - Thibault
>>>>
>>>> On Fri, Jan 8, 2021 at 2:30 PM Chris Wine <chriswine at gmail.com> wrote:
>>>>
>>>>> Hi, I'm using GStreamer/GES 1.18 on Windows, and having an issue where
>>>>> my GESTimeline hangs indefinitely when one of my input clips is a MPEG
>>>>> Program Stream file. I can also reproduce the issue with ges-launch-1.0.exe:
>>>>>
>>>>> ges-launch-1.0.exe +clip SampleMpegProgramStream.mpg
>>>>>
>>>>> And here is a link to the file:
>>>>> https://www.dropbox.com/s/vlno2iwsml71obr/SampleMpegProgramStream.mpg?dl=0
>>>>>
>>>>> With gst-play-1.0.exe, it is able to play the file all the way through
>>>>> successfully. Does anyone have any insight they could share with me?
>>>>>
>>>>> Thanks,
>>>>> --Chris
>>>>> _______________________________________________
>>>>> gstreamer-devel mailing list
>>>>> gstreamer-devel at lists.freedesktop.org
>>>>> https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel
>>>>>
>>>> _______________________________________________
>>>> gstreamer-devel mailing list
>>>> gstreamer-devel at lists.freedesktop.org
>>>> https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel
>>>>
>>> _______________________________________________
>>> gstreamer-devel mailing list
>>> gstreamer-devel at lists.freedesktop.org
>>> https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel
>>>
>> _______________________________________________
>> gstreamer-devel mailing list
>> gstreamer-devel at lists.freedesktop.org
>> https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/gstreamer-devel/attachments/20210209/ff42e841/attachment-0001.htm>

More information about the gstreamer-devel mailing list