Re: [PATCH] faster DVB-S lock with cards using stb0899 demod
On Mon, Sep 20, 2010 at 8:03 AM, Emmanuel eall...@gmail.com wrote: SE a écrit : hi list v4l-dvb still lacks fast and reliable dvb-s lock for stb08899 chipsets. Can you please try again from here; http://202.88.242.108:8000/hg/var/www/hg/v4l-dvb/ Regards, Manu -- To unsubscribe from this list: send the line unsubscribe linux-media in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH] faster DVB-S lock with cards using stb0899 demod
Hello Mauro, On 19.09.2010 11:46, SE wrote: hi list v4l-dvb still lacks fast and reliable dvb-s lock for stb08899 chipsets. This problem was adressed by Alex Betis two years ago [1]+[2]resulting in a patch [3] that made its way into s2-liplianin, not v4l-dvb. With minor adjustments by me this patch now offers reliable dvb-s/dvb-s2 lock for v4l-dvb, most of them will lock in less than a second. Without the patch many QPSK channels won't lock at all or within a 5-20 second delay. Can you please comment on this patch and tell us if you are considering this patch for integration into the v4l git tree? Regards André -- To unsubscribe from this list: send the line unsubscribe linux-media in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH] faster DVB-S lock with cards using stb0899 demod
Em 18-10-2010 06:27, André Weidemann escreveu: Hello Mauro, On 19.09.2010 11:46, SE wrote: hi list v4l-dvb still lacks fast and reliable dvb-s lock for stb08899 chipsets. This problem was adressed by Alex Betis two years ago [1]+[2]resulting in a patch [3] that made its way into s2-liplianin, not v4l-dvb. With minor adjustments by me this patch now offers reliable dvb-s/dvb-s2 lock for v4l-dvb, most of them will lock in less than a second. Without the patch many QPSK channels won't lock at all or within a 5-20 second delay. Can you please comment on this patch and tell us if you are considering this patch for integration into the v4l git tree? You should ask s2-liplianin and stb08899 maintainers about that, and not me. Ideally, patches at s2-liplianin that fixes bug or improves support should be rebased to upstream and send to us, but, currently, I didn't see any effort from the maintainers to do it. Cheers, Mauro. -- To unsubscribe from this list: send the line unsubscribe linux-media in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH] faster DVB-S lock with cards using stb0899 demod
Sun 19 Sep 2010, at 22:33:16 Emmanuel ALLAUD wrote: I will try this with a TT-S2 3200 when I find some time ;-) Do I need a very recent tree? I have a v4l-dvb tree from a year ago I think. Hello Emmanuel Testers in vdr-portal reported good result with the TT-s2 3200 so it's worth trying. There has not been much going on in that particular file in v4l-dvb, so you might have a chance and use it with your version of v4l-dvb. The majordomo of vger.kernel.org blacklisted the domain name on line 7 of the patch as well (he's thorough, isn't he? ;-) ) so you'll have to replace that hidden domain name inside the patch with the one from the original file and it should apply without rejects. Same goes for line 30 of the szap-s2 patch, there is another domain name to be replaced from the original file, and the patch will apply fine. You had problems two years ago with the patch of Alex at tuning some 8psk channels [1], it might be a good idea to test those channels again with v4l- dvb. happy patching, compiling, testing and tuning! ... Should this terribly fail with the patching, there is a tarball attached inside posting #1 of vdr-portal you can download and use. Most of the text is in German, but feel free to ask questions in English and post your test results in English over here or over there. I'd also like some v4l-dvb maintainers to test and comment on this patch if possible. There are quite a bunch of stb0899 cards out there, so I hope you guys still use one of these cards and can test it on. [1] http://www.linuxtv.org/pipermail/linux-dvb/2008-October/029538.html -- To unsubscribe from this list: send the line unsubscribe linux-media in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH] faster DVB-S lock with cards using stb0899 demod
SE a écrit : hi list v4l-dvb still lacks fast and reliable dvb-s lock for stb08899 chipsets. This problem was adressed by Alex Betis two years ago [1]+[2]resulting in a patch [3] that made its way into s2-liplianin, not v4l-dvb. With minor adjustments by me this patch now offers reliable dvb-s/dvb-s2 lock for v4l-dvb, most of them will lock in less than a second. Without the patch many QPSK channels won't lock at all or within a 5-20 second delay. The algo can be tested with a modified version of szap-s2 [4], introducing: * process a channel list sequentially (-e [number] -n [number]) * DiSEqC repetition (-s [number] - the default is 1 sequence + 1 repetition) * faster status polling (poll instantly after tuning, then poll every 10ms instead of 1 poll per second) * some statistics about the tuning success while processing the list Here are the new features of szap2-s2 explained: ## channel lock with instant status poll [last raw still is 0] using '/dev/dvb/adapter0/frontend0' and '/dev/dvb/adapter0/demux0' status 1f|signal 27948|noise 56032|ber 0|unc -2|tim 0|FE_HAS_LOCK| 0 ## channel lock with the first status poll [last raw is 1] using '/dev/dvb/adapter0/frontend0' and '/dev/dvb/adapter0/demux0' status 0b|signal 23200|noise 40413|ber 0|unc -2|tim 0| status 1b|signal 23200|noise 37136|ber 0|unc -2|tim 1|FE_HAS_LOCK| 1 ## channel lock with the second status poll [last raw is 2] using '/dev/dvb/adapter0/frontend0' and '/dev/dvb/adapter0/demux0' status 00|signal 245|noise21|ber 0|unc -2|tim 0| status 1f|signal 17347|noise 45219|ber 0|unc -2|tim 2|FE_HAS_LOCK| 2 ## no channel lock - try to lock for 10 seconds, then give up and increase lok_errs +1 using '/dev/dvb/adapter0/frontend0' and '/dev/dvb/adapter0/demux0' status 00 | signal 0 | noise 4 | ber 0 | unc -2 | tim0 | status 00 | signal 0 | noise 4 | ber 0 | unc -2 | tim 100 | status 00 | signal 0 | noise 4 | ber 0 | unc -2 | tim 200 | status 00 | signal 0 | noise 4 | ber 0 | unc -2 | tim 300 | status 00 | signal 0 | noise 4 | ber 0 | unc -2 | tim 400 | status 00 | signal 0 | noise 4 | ber 0 | unc -2 | tim 500 | status 00 | signal 0 | noise 4 | ber 0 | unc -2 | tim 600 | status 00 | signal 0 | noise 4 | ber 0 | unc -2 | tim 700 | status 00 | signal 0 | noise 4 | ber 0 | unc -2 | tim 800 | status 00 | signal 0 | noise 4 | ber 0 | unc -2 | tim 900 | ## the tuning statistics look like this: lok_errs =0, runs=3035 of sequ=1207, multi=139, multi_max=2 * lok_errs = amount of lock errors * runs = current channel number while processing the list * sequ = the amount of channels to process you specified with -e [number] * multi = amount of multiple polls * multi_max = the highest status poll of a channel is stored in here Here are the results from ezap2 with an Astra 19.2E list and improved algo: TOT: lok_errs =0, runs=1207 of sequ=1207, multi=48, multi_max=47 real22m52.883s user0m0.004s sys 0m20.297s Here are the results from ezap2 with the same list and v4l-dvb mercurial algo: TOT: lok_errs =233, runs=1207 of sequ=1207, multi=113361, multi_max=987 real135m34.236s user0m0.344s sys 7m52.322s Similar results where reported by testers in vdr-portal.de [5] Feel free to test the improved algo yourself like this: time ./ezap2 -a0 -xHc Astra_only.txt -e 1207 -n 1 zap.log Change adapter to 1 or higher in case stb0899 is a different adapter in your multi card setup. Attachments are stb0899_algo.c.patch, szap-s2-to-ezap2.patch, Astra_only.txt (Astra 19.2E channels list in zap format) Inline posted patches get word wrapped again and again in kmail, even after I followed the suggestions in email-clients.txt [1] http://www.linuxtv.org/pipermail/linux-dvb/2008-September/029361.html [2] http://www.linuxtv.org/pipermail/linux-dvb/2008-October/029455.html [3] http://mercurial.intuxication.org/hg/s2-liplianin/rev/d423b7887ec8 [4] http://mercurial.intuxication.org/hg/szap-s2 [5] http://www.vdr-portal.de/board/thread.php?threadid=99603 Signed-off-by: SE tuxoho...@hotmail.de I will try this with a TT-S2 3200 when I find some time ;-) Do I need a very recent tree? I have a v4l-dvb tree from a year ago I think. Bye Manu -- To unsubscribe from this list: send the line unsubscribe linux-media in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
