[mythtv-users] UK Commercial Detection (Clusters)

Robert Longbottom rongblor at googlemail.com
Tue Mar 12 11:48:38 UTC 2013



On 12 Mar 2013, at 11:01 AM, "Roger Siddons" <dizygotheca at ntlworld.com> wrote:

> On Mon, 11 Mar 2013 23:25:46 -0000, Roger Siddons <dizygotheca at ntlworld.com> wrote:
> 
> I've put a new version (v2.0) of the cluster script on the wiki 
> 
> http://www.mythtv.org/wiki/Commercial_detection_with_silence_for_UK_freeviewHD#Cluster_Detecting_Version
> 
> There's a new silence.py, silence.cpp & Makefile; catagrower.cpp is unchanged.
> 
> It's a re-design that;
> 
> - flags adverts shortly after they end
> - fixes segfaults
> - makes use of Myth headers optional
> - has improved Makefile
> 
> I've made some progress with the python bindings so v3.0, removing the need for headers altogether, shouldn't be far off.
> 
> Just in case anyone is struggling to get their jobs running, note that the wiki command was wrong. I've corrected it now.
> 
> You should be using %VERBOSEMODE% rather than %VERBOSELEVEL%. A quirk of the arg parsing meant that the job would fail silently with the wrong option.
> 
> Apologies to anyone affected.

Ah ha!  Yes I was having that problem but hadn't tracked it down yet. Thanks :-)

Also in the Makefile I had to add -L/usr/lib/qt4 to get it to compile. I'm assuming this is only necessary if you enable the mythtv libs and qt libs by editing silence.cpp (which I have). 

It's found 7 breaks in a 2hour film from the other night which sounds about right to me. I shall try the chase play flagging tonight when I get home.

Thanks for all your hard work. 
Rob. 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.mythtv.org/pipermail/mythtv-users/attachments/20130312/571cc3ec/attachment.html>


More information about the mythtv-users mailing list