On Sun, 2009-05-03 at 17:59 -0700, Matt Beadon wrote: > >After you've tested it and are happy the settings are good enough > for > >general user use, I can add a patch and issue a PULL request. I > usually > >issue a PULL request on Sunday nights. The v4l-dvb maintainer > usually > >gets to them within a week, unless he's really busy. > > > I'm happy that analog tuning is working now. Do you have some > suggestions on what testing I should do beyond watching it? :)
Test all the channels. Especially ones near the VHF-Low/VHF-High boundary and the VHF-High/UHF boundary. The channel/frequency tables in the ivtv-tune source can tell you the mapping of cable channels to frequency. The file linux/drivers/media/common/tuners/tuner-types.c shows that for the FM1236 MK3 tuner, the VHF-Low/VHF-High boundary is around 160 MHz and the VHF-High/UHF boundary is 442 MHz. There is only one other analog/only NTSC tuner in the tuner-types.c file with a TDA9887: tuner type 46, the TUNER_PANASONIC_VP27. It likely won't work. Please test tuner type 46 (I suspect it won't work or work as well). > As a follow up I'm having some issues with the ATSC channels on this > card (beyond the listings problem I still have posted on the > mythtv-users list). Occasionally the signal breaks up and/or stutters > along with audio glitches and pops. > I have a Core2 E6600 and when I'm playing live ATSC the cpu usage is > around 44% but it's actually about 2% higher when I'm playing live > NTSC so I don't think the system is too slow. That seems high for NTSC - you (MythTV) must be doing some software scaling somewhere. mplayer doesn't do software scaling unless you tell it to. You can load the cx18 module with the debug=15 module parameter and look for "Possibly falling behind" messages to see how respinsive you system is to CX23418 interrupts. If you get a lot of those messages in the logs that come with sequence numbers that are close together, you've got some system level inefficiencies. > I thought it might be signal strength so I've cut out as many > splitters as possible and this card is getting the 2nd cleanest cable > I have (best goes to cable modem). You whole cable plant in your home affects the S/N ratio (e.g. voltage wave reflections due to impedenace mismatches come back up from the far end of the cable and can look like noise). Please review: http://www.ivtvdriver.org/index.php/Howto:Improve_signal_quality > > When I tune an ATSC channel with mythtv it says "Signal 0% | SNR 2.5dB > | BEXXX | LAM(lock)". The 0% part is what has me worried that I still > don't have good enough SNR but I haven't poked around enough to figure > what other people are getting for signal % and SNR? Meh. The signal 0% is a known deficiency. I can't remember if it's the mxl5005s driver or the s5h1409 driver that's reporting junk. The SNR of about 2.3 dB is what I get for OTA ATSC. LAM(lock) is good. BE (block errors) should not be increasing, but a fixed non-zero number is OK. Regards, Andy > There seem to be a ton of cx18 threads. Is there a guide for > optimizing settings for cx18? Or some general tips that might help? > > _______________________________________________ ivtv-users mailing list [email protected] http://ivtvdriver.org/mailman/listinfo/ivtv-users
