Hi Thomas, Am Samstag, den 27.02.2010, 01:31 +0100 schrieb thomas schorpp: > Hello, Hermann, > > hermann pitton wrote: > > Hi Thomas, > > > > Am Samstag, den 27.02.2010, 00:34 +0100 schrieb thomas schorpp: > >> Looks like fixed by linux 2.6.33 just in time, BIG Thank You guys ;-) > >> > >> Even at higher BER: > >> > >> Current parameters: > >> Frequency: 1945.320 MHz > >> Inversion: OFF > >> Symbol rate: 22.000154 MSym/s > >> FEC: FEC 5/6 > >> > >> cycle: 1 d_time: 0.001 s Sig: 18504 SNR: 39578 BER: 168 UBLK: 0 Stat: 0x1f [SIG CARR VIT SYNC LOCK ] > >> cycle: 2 d_time: 0.073 s Sig: 18247 SNR: 39578 BER: 225 UBLK: 0 Stat: 0x1f [SIG CARR VIT SYNC LOCK ] > >> cycle: 3 d_time: 0.079 s Sig: 18504 SNR: 37779 BER: 140 UBLK: 0 Stat: 0x1f [SIG CARR VIT SYNC LOCK ] > >> cycle: 4 d_time: 0.072 s Sig: 18504 SNR: 39835 BER: 198 UBLK: 0 Stat: 0x1f [SIG CARR VIT SYNC LOCK ] > >> cycle: 5 d_time: 0.071 s Sig: 18504 SNR: 39835 BER: 221 UBLK: 0 Stat: 0x1f [SIG CARR VIT SYNC LOCK ] > >> cycle: 6 d_time: 0.072 s Sig: 18247 SNR: 39578 BER: 249 UBLK: 0 Stat: 0x1f [SIG CARR VIT SYNC LOCK ] > >> cycle: 7 d_time: 0.072 s Sig: 18504 SNR: 39835 BER: 191 UBLK: 0 Stat: 0x1f [SIG CARR VIT SYNC LOCK ] > >> cycle: 8 d_time: 0.072 s Sig: 18504 SNR: 39578 BER: 185 UBLK: 0 Stat: 0x1f [SIG CARR VIT SYNC LOCK ] > >> cycle: 9 d_time: 0.072 s Sig: 18761 SNR: 39578 BER: 137 UBLK: 0 Stat: 0x1f [SIG CARR VIT SYNC LOCK ] > >> > >> I'll report if issue reoccurs and try to finetune crystal based tuner/demod parameters, then. > >> > >> y > >> tom > > > > I just started to try to look it up, but don't have ground yet. > > Look for tda10086 changesets in the stable branch git repository at kernel.org 2.6.32.7...33 > and linux-media repository ? > > If there's no applicable change then I've misinterpreted the fix for the clear sky tonight :D > but I'm pretty sure the issue occured at any weather with hours of clear sky periods last week, > there's not been a minute without TS errors in VDR as long as the card has been in use. > > > > > I reported unexpected bad performance under GNU/Linux for that card > > previously. > > On this list? Give weblink pls. No, but here is http://www.mail-archive.com/linux-media@xxxxxxxxxxxxxxx/msg15335.html I know for sure, that I was totally baffled, when Manu, Oliver and Hartmut did start talking about big head room for improvements on the tda826x, given the results I had on the tda8261 compared on m$ for both. The flaw on linux was on the tda8261. I'm sure I did not shut up, but don't have a link for now. Thanks to Julian in the first place I would expect. Cheers, Hermann We always will have bugs. > > > > Can you point me to the fix? > > > > Cheers, > > Hermann > > y > tom > > > > >> thomas schorpp wrote: > >>> Hi, > >>> Issue is already confirmed here: > >>> http://www.vdr-portal.de/board/thread.php?threadid=93268 > >>> > >>> Linux 2.6.32.8, 80cm dish. > >>> > >>> Do we have any Tuner/Decoder optimization points in the FE code? > >>> > >>> This is not OK: > >>> > >>> lspci -s 00:08.0 -v 00:08.0 Multimedia controller: Philips > >>> Semiconductors SAA7134/SAA7135HL Video Broadcast Decoder (rev 01) > >>> Subsystem: Creatix Polymedia GmbH Device 0005 Flags: bus master, medium > >>> devsel, latency 32, IRQ 19 Memory at fbeff400 (32-bit, non-prefetchable) > >>> [size=1K] Capabilities: [40] Power Management version 1 Kernel driver in > >>> use: saa7134 > >>> > >>> grep cTS2PES /var/log/syslog > >>> Feb 26 13:46:59 tom1 vdr: [4082] cTS2PES got 7 TS errors, 113 TS > >>> continuity errors > >>> Feb 26 13:46:59 tom1 vdr: [4082] cTS2PES got 0 TS errors, 29 TS > >>> continuity errors > >>> Feb 26 13:47:52 tom1 vdr: [4082] cTS2PES got 17 TS errors, 5 TS > >>> continuity errors > >>> Feb 26 14:03:03 tom1 vdr: [4082] cTS2PES got 2 TS errors, 136 TS > >>> continuity errors > >>> Feb 26 14:03:03 tom1 vdr: [4082] cTS2PES got 0 TS errors, 32 TS > >>> continuity errors > >>> Feb 26 14:41:42 tom1 vdr: [4082] cTS2PES got 1 TS errors, 853 TS > >>> continuity errors > >>> Feb 26 14:41:42 tom1 vdr: [4082] cTS2PES got 0 TS errors, 194 TS > >>> continuity errors > >>> Feb 26 14:52:58 tom1 vdr: [4082] cTS2PES got 2 TS errors, 196 TS > >>> continuity errors > >>> Feb 26 14:52:58 tom1 vdr: [4082] cTS2PES got 0 TS errors, 52 TS > >>> continuity errors > >>> Feb 26 14:59:34 tom1 vdr: [4082] cTS2PES got 0 TS errors, 137 TS > >>> continuity errors > >>> Feb 26 14:59:34 tom1 vdr: [4082] cTS2PES got 0 TS errors, 43 TS > >>> continuity errors > >>> Feb 26 14:59:34 tom1 vdr: [4082] cTS2PES got 0 TS errors, 16 TS > >>> continuity errors > >>> Feb 26 14:59:34 tom1 vdr: [4082] cTS2PES got 0 TS errors, 57 TS > >>> continuity errors > >>> Feb 26 14:59:54 tom1 vdr: [4082] cTS2PES got 0 TS errors, 3 TS > >>> continuity errors > >>> Feb 26 14:59:54 tom1 vdr: [4082] cTS2PES got 0 TS errors, 2 TS > >>> continuity errors > >>> > >>> dvbsnoop -s feinfo -adapter 2 > >>> Current parameters: > >>> Frequency: 1236.253 MHz > >>> Inversion: OFF > >>> Symbol rate: 31.794142 MSym/s > >>> FEC: FEC 3/4 > >>> > >>> dvbsnoop -s signal -adapter 2 > >>> cycle: 1 d_time: 0.001 s Sig: 26471 SNR: 49858 BER: 0 UBLK: 0 Stat: 0x1f > >>> [SIG CARR VIT SYNC LOCK ] > >>> cycle: 2 d_time: 0.072 s Sig: 26471 SNR: 50115 BER: 0 UBLK: 0 Stat: 0x1f > >>> [SIG CARR VIT SYNC LOCK ] > >>> cycle: 3 d_time: 0.072 s Sig: 26728 SNR: 50115 BER: 0 UBLK: 0 Stat: 0x1f > >>> [SIG CARR VIT SYNC LOCK ] > >>> cycle: 4 d_time: 0.088 s Sig: 26728 SNR: 50115 BER: 0 UBLK: 0 Stat: 0x1f > >>> [SIG CARR VIT SYNC LOCK ] > >>> cycle: 5 d_time: 0.072 s Sig: 26471 SNR: 50115 BER: 0 UBLK: 0 Stat: 0x1f > >>> [SIG CARR VIT SYNC LOCK ] > >>> cycle: 6 d_time: 0.072 s Sig: 26471 SNR: 50115 BER: 0 UBLK: 0 Stat: 0x1f > >>> [SIG CARR VIT SYNC LOCK ] > >>> cycle: 7 d_time: 0.072 s Sig: 26471 SNR: 50115 BER: 0 UBLK: 0 Stat: 0x1f > >>> [SIG CARR VIT SYNC LOCK ] > >>> cycle: 8 d_time: 0.072 s Sig: 26471 SNR: 50115 BER: 0 UBLK: 0 Stat: 0x1f > >>> [SIG CARR VIT SYNC LOCK ] > >>> > >>> Low signal strength values are AGC-loop misinterpretation as usual? > >>> > >>> y > >>> tom > >>> > >>> > > > > > > -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
