[mythtv-users] silence detection blocks jobqueue [ mythcommflag; silence.py ]

Stefan Pappalardo stefan.pappalardo at gmx.de
Mon Oct 20 15:49:38 UTC 2014


Am 20.10.2014 um 16:29 schrieb mythtv at phipps-hutton.freeserve.co.uk:
>
> Quoting Roger Siddons <dizygotheca at ntlworld.com>:
>> On Mon, 20 Oct 2014 14:49:00 +0100, Stefan Pappalardo <stefan.pappalardo at gmx.de> wrote:
>>
>>> I wonder if mythffmpeg should switch to the "true" audio stream by itself, or if this is the job of mythfrontend...
>>>
>>
>> Good point. Strange we haven't stumbled on it before...
>>
>> To be clear: mythfrontend does handle this (somehow). This is an external 3rd party script and not condoned, sanctioned or supported by the devs - they're getting enough flak for other people's f**k-ups at the moment!
>>
>
> This is an ffmpeg bug. The program mythffmepg is used in the script as it is included with the mythtv packages.  Could you try replacing mythffmpeg with avconv? If you don't have avconv you probably need to install it. It's in libav-tools on Debian.
>
> avconv is likely to be more up-to-date than mythffmpeg and so may already have this bug fixed.
>
> There is code in silence.cpp to stop after a timeout but all that dies is to kill the first process in the pipeline. mythffmpeg/avconv is the second and it will normally quit when the first process is killed but in this case it is hung. We could change the silence.py to pass the pid of the first and second pipeline processes to silence.cpp and have that kill both processes.
>
> Cheers,
> Tim.
>
>
> _______________________________________________
> mythtv-users mailing list
> mythtv-users at mythtv.org
> http://www.mythtv.org/mailman/listinfo/mythtv-users
> http://wiki.mythtv.org/Mailing_List_etiquette
> MythTV Forums: https://forum.mythtv.org
>

OK, using avconv instead of mythffpeg works fine. Here are the steps for people with similar problem:
1.) install libav-tools (and libavcodec-extra-54, only necessary for encoding mp3)
2.) change silence.py only substituting "mythffmpeg" with "avconv"
=> avconv can handle recordings with variable count of audio streams and channels (mythffmpeg does not)

(I will also place a hint in the discussion of the wiki (http://www.mythtv.org/wiki/Talk:Commercial_detection_with_silences)

THANK YOU VERY MUCH FOR YOUR HELP :-)

-- 
Stefan Pappalardo


More information about the mythtv-users mailing list