On Tue, 23 Aug 2005 16:40:58 +0300 Kimmo Taskinen <kimmo.taskinen@xxxxxxxxxxxxxxx> wrote: > The patch didn't help. I'm not familiar with the vdr source code and Well, I'm not suprised, but never underestimate the placebo effect :) > Below is some vdr output during record watching. Somehow it seems to me > that as vdr is scanning through different channels during record > watching (EPG scan?) the ttxtsubs-plugin "attaches" to that channel and > shows the subtitles from it. So, something went really broke while updating from 1.3.27 to 1.3.28 and a quick look at VDR's HISTORY reveals these changes: 2005-08-07: Version 1.3.28 - Reactivated cVideoRepacker in remux.c after some fixes (thanks to Reinhard Nissl). - Modified cEITScanner::Process() so that it uses the primary device if it is replaying and is the only device that provides the given transponder, and that a forced EPG scan works even if EPG scan timeout is set to 0 (thanks to Bernhard Stegmaier for reporting a problem with this). Could you revert the changes above one by one and check if one or another triggers the actual problem? You can easily deactivate the cVideoRepacker by commenting the "#define TEST_cVideoRepacker" line in remux.c and the eitscan.c changes are well documented in: ftp://ftp.cadsoft.de/vdr/Developer/vdr-1.3.27-28.diff BR, -- rofa
