Paul Sijben wrote: > I think that before this code is included in the main distribution a > couple of loose ends need fixing: > > - how to check for the number of (ivtv) capture cards without naively > checking against the length of VIDEO_GROUPS (input needed, is there a > function to call for that???)
I think you would need to scan the video devices using tv.v4l2 it queries the devices capabilities. If you can read C check the ivtv-detect program in the ivtv-0.8.1 driver. > - I wrote:/ I am pretty sure I broke some system that handles > interaction between TV viewing and recording. /This is because I changed > the name of the lockfile to deal with multiple cards. So this needs to > be aligned with other parts of freevo that look at this lockfile. IIRC, I changed the lockfile to use the video device name as a suffix, but you may need to fix upsoon as this checks this device and sends a stop to the player if the tv card's tuner is in use. > - I think conflict resolution needs a bit more refinement > - is there an elegant way (or even a need) to allow parallel recording > over multiple card types? BTTV and saa713x cards use a raw video stream, which in turn needs encoding in real-time (The raw files are massive, 27Mb/sec). Real-time software encoding on more than one channel would put a very high load on the system, but I've not tried it. Duncan > Duncan Webb wrote: >> Paul Sijben wrote: >> >>> OK here are the 1.6 files as patched. >>> >>> The patch allows recording multiple shows concurrently. It assumes >>> that all TV channels are available from the same card type. (in this >>> case ivtv, you can make a similar patch against other record types, >>> just patch ????_record.py). >>> >>> I am using VIDEO_GROUPS as follows. (note that I also assume that this >>> is the only kind of entries there are in video groups.I simply check >>> for the number of cards by checking the length of VIDEO_GROUPS, there >>> must be a more elegant way to do this...) >>> >>> VIDEO_GROUPS = [ >>> VideoGroup(vdev="/dev/video0", >>> adev=None, >>> input_type='tuner', >>> # input_num=0, >>> tuner_norm=CONF.tv, >>> tuner_chanlist=CONF.chanlist, >>> desc='Cable', >>> group_type='ivtv', >>> record_group=None), >>> VideoGroup(vdev="/dev/video1", >>> adev=None, >>> input_type='tuner', >>> # input_num=0, >>> tuner_norm=CONF.tv, >>> tuner_chanlist=CONF.chanlist, >>> desc='Cable', >>> group_type='ivtv', >>> record_group=None) >>> ] >>> >>> conflict resolution is now as follows: >>> - are there unused capture cards, pick one of those, highest number >>> first (as TV viewing is always done on video0) >>> - if all are busy find a card where the show is to end now >>> - otherwise pick an active card and use the normal conflict resolution >>> code. (indeed this could use with some refining) >>> >>> Note: I am pretty sure I broke some system that handles interaction >>> between TV viewing and recording. >>> >> >> I'm interested to get some feedback from anybody who has tried Paul's >> changes, if some of you are using this and others want to use it then >> I'm more than happy to add the changes into rel-1. >> >> Duncan ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV _______________________________________________ Freevo-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/freevo-users
