[mythtvnz] Pixellation Every ~ 20 minutes

Fri Oct 5 09:05:54 BST 2012

On Fri, 05 Oct 2012 12:15:07 +1300, Curtis Walker <sultanoswing at gmail.com>  
wrote:

> On 5 October 2012 09:41, Paulgir <paulgir at gmail.com> wrote:
>>
>> On Fri, 05 Oct 2012 05:49:30 +1300, Pieter De Wit  
>> <pieter at insync.za.net> wrote:
>>
>> On 4/10/2012 22:34, Nick Rout wrote:
>>
>>
>> On Oct 4, 2012 8:07 PM, "Curtis Walker" <sultanoswing at gmail.com> wrote:
>> >
>> > Odd problem this.
>> >
>> > On all TV recordings, after around 20 minutes, the picture starts  
>> becoming pixellated. Over the next 30 or so seconds it becomes more and  
>> more pixellated until it looks like an impressionist painting. The  
>> sound is generally OK, but does have the odd, brief "glitch" (a noise  
>> like a modem negotiating). After no more than a minute, the picture and  
>> sound suddenly return to normal.
>> >
>> > I'm fairly sure this is not a reception issue, since it seems to  
>> occur so regularly, and at roughly the same time in each recording, and  
>> for the rest of the recording picture and sound are perfect. Signal  
>> strength and SNR are fine (98% and 2.3dB respectively, and this is  
>> representative). Same thing occurs on all channels.
>> >
>> > It doesn't appear to be related to the playback, as the same  
>> pixellation occurs in the same place whether viewing the recording on  
>> desktop (nvidia) or Laptop (intel integrated), although it sounds  
>> similar to symptoms reported by others on mythtv <0.25 which people  
>> ascribed to vdpau and buffer settings.
>> >
>> > Could be some scheduled process is running every 20 minutes or so on  
>> my disks? Although why it occurs ~20 minutes into a recording, and  
>> never at other times, doesn't seem to fit with this either.
>> >
>> > Hmmmm ....
>>
>> I have seen similar stuff. I thought it was to do with the networking  
>> between my HDHR and my backend. I should indicate further.
>>
>>
>> Quite correct Nick. If you drop a few packets from the h264 stream, it  
>> does this, the more your drop, the worst the result (logical, I know :)  
>> ). I used to see this with program changes (I am sure most of you have  
>> seen my rants on the main list), but since then I took a certain  
>> supplier to task on a $99 deal for a 120gig SSD, problem gone :)
>>
>> Why this happens at the 20 min mark is hard to say, can you tell us  
>> more about your setup ? Perhaps more details around  
>> CPU,memory,disk,mount points ?
>>
>> Cheers,
>>
>> Pieter
>>
>>
>> In my case I have a combined BE/FE wit an Intel G620 2.6 GHz CPU,4GB  
>> RAM OS is on a separate drive to storage and is not performing comm  
>> flagging or transcoding.
>>
>> Cheers
>>
>> Paul
>>
>> --
>> Using Opera's revolutionary email client: http://www.opera.com/mail/
>>
>> _______________________________________________
>> mythtvnz mailing list
>> mythtvnz at lists.linuxnut.co.nz
>> http://lists.ourshack.com/mailman/listinfo/mythtvnz
>> Archives http://www.gossamer-threads.com/lists/mythtv/mythtvnz/
>>
>
>
> Ahhhhh. Glad (in some ways) to see I'm not the only one then.
>
> Specs (Desktop):
> ASUS P7P55D-E-Pro
> Core 2 Duo i750
> 8GB RAM
> Geforce 460 GTX (drivers 304.51)
> Hauppage HVR-2210 (built-in kernel drivers + firmware  
> NXP7164-2010-03-10.1.fw)
> System drive: Samsung 830 (ext4); media drive (/mnt/media) Seagate
> Barracuda 7200.12 SATA 1TB (ext4) w/470GB free space
> Arch Linux (kernel 3.5.4, Gnome 3.2)
> This is a combined local FE/BE machine on taniwha-net using mythtv,
> arch pkg 0.25.2-1, so no over-the-network issues
>
> Specs (Laptop):
> HP Folio 13, core i5
> Integrated Intel Graphics
> 128GB SSD
> Arch Linux (kernel 3.5.4, Gnome 3.2)
> FE machine connected to BE via wireless G (WRT54GL, Tomato 1.28)
>
> Log from day and time from an example of when pixellation occurred.
> Program is "Go Diego Go" (Channel Four, H.264, SD 480, 2.0 sound)
> scheduled from 0820 to 0850. Pixellation started 21:10 minutes into
> the recording i.e. around 0841, got slowly worse until completely
> smeared at 22:10, then resolved suddenly at 22:55. Although it's a 4:3
> aspect recording, the pixel smears extend into the black side bars of
> the picture.
>
> Here's the log:
>
> 2012-09-26 08:19:55.248911 I [992/1307] TVRecEvent tv_rec.cpp:3989
> (TuningNewRecorder) - TVRec(1): rec->GetPathname():
> '/mnt/media/mythtv/4301_20120926082000.mpg'
> 2012-09-26 08:20:13.078247 W [992/1745] TFWWrite
> ThreadedFileWriter.cpp:499 (DiskLoop) -
> TFW(/mnt/media/mythtv/4301_20120926082000.mpg:56): write(61100) cnt 15
> total 818552 -- took a long time, 1663 ms
> 2012-09-26 08:23:31.214100 I [992/1337] HouseKeeping
> housekeeper.cpp:225 (RunHouseKeeping) - Running housekeeping thread
> 2012-09-26 08:24:57.583625 N [992/1338] Expire autoexpire.cpp:263
> (CalcParams) - AutoExpire: CalcParams(): Max required Free Space:
> 102.0 GB w/freq: 14 min
> 2012-09-26 08:28:36.218878 I [992/1337] HouseKeeping
> housekeeper.cpp:225 (RunHouseKeeping) - Running housekeeping thread
> 2012-09-26 08:33:38.224655 I [992/1337] HouseKeeping
> housekeeper.cpp:225 (RunHouseKeeping) - Running housekeeping thread
> 2012-09-26 08:38:44.231152 I [992/1337] HouseKeeping
> housekeeper.cpp:225 (RunHouseKeeping) - Running housekeeping thread
> 2012-09-26 08:38:57.589742 N [992/1338] Expire autoexpire.cpp:263
> (CalcParams) - AutoExpire: CalcParams(): Max required Free Space:
> 102.0 GB w/freq: 14 min
> 2012-09-26 08:43:48.237017 I [992/1337] HouseKeeping
> housekeeper.cpp:225 (RunHouseKeeping) - Running housekeeping thread
> 2012-09-26 08:45:49.350493 I [992/1793] Metadata_39 jobqueue.cpp:2151
> (DoMetadataLookupThread) - JobQueue: Metadata Lookup Starting for "Go
> Diego Go" recorded from channel 4301 at 2012-09-26T08:20:00
> 2012-09-26 08:45:49.350941 I [992/1795] SystemManager
> system-unix.cpp:263 (run) - Starting process manager
> 2012-09-26 08:45:49.351130 I [992/1796] SystemSignalManager
> system-unix.cpp:485 (run) - Starting process signal handler
> 2012-09-26 08:45:49.355944 I [992/1797] SystemIOHandlerR
> system-unix.cpp:90 (run) - Starting IO manager (read)
> 2012-09-26 08:45:49.356029 I [992/1798] SystemIOHandlerW
> system-unix.cpp:90 (run) - Starting IO manager (write)
> 2012-09-26 08:45:50.286250 I [992] ProcessRequest mainserver.cpp:1360
> (HandleAnnounce) - MainServer::ANN Monitor
> 2012-09-26 08:45:50.286265 I [992] ProcessRequest mainserver.cpp:1362
> (HandleAnnounce) - adding: taniwha-desktop as a client (events: 0)
> 2012-09-26 08:45:50.286901 I [992] ProcessRequest mainserver.cpp:1360
> (HandleAnnounce) - MainServer::ANN Monitor
> 2012-09-26 08:45:50.286913 I [992] ProcessRequest mainserver.cpp:1362
> (HandleAnnounce) - adding: taniwha-desktop as a client (events: 1)
> 2012-09-26 08:48:52.243113 I [992/1337] HouseKeeping
> housekeeper.cpp:225 (RunHouseKeeping) - Running housekeeping thread
> 2012-09-26 08:50:00.004754 I [992/1307] TVRecEvent tv_rec.cpp:1029
> (HandleStateChange) - TVRec(1): Changing from RecordingOnly to None
> 2012-09-26 08:50:00.005061 E [992/1748] DVBRead
> dvbstreamhandler.cpp:214 (RunTS) - DVBSH(/dev/dvb/adapter0/frontend0):
> Device EOF detected
> 2012-09-26 08:50:00.435109 I [992/992] CoreContext scheduler.cpp:638
> (UpdateRecStatus) - Updating status for "Go Diego Go" on cardid 1
> (Recording => Recorded)
> 2012-09-26 08:50:00.458742 I [992/1335] Scheduler scheduler.cpp:2034
> (HandleReschedule) - Reschedule requested for id 0.
> 2012-09-26 08:50:00.479037 I [992/1307] TVRecEvent
> recordinginfo.cpp:1113 (FinishedRecording) - Finished recording Go
> Diego Go: channel 4301
> 2012-09-26 08:50:00.525543 I [992/1335] Scheduler scheduler.cpp:2081
> (HandleReschedule) - Reschedule interrupted, will retry
> 2012-09-26 08:50:00.525568 I [992/1335] Scheduler scheduler.cpp:2034
> (HandleReschedule) - Reschedule requested for id 0.
> 2012-09-26 08:50:00.573833 I [992/1335] Scheduler scheduler.cpp:2094
> (HandleReschedule) - Scheduled 0 items in 0.0 = 0.00 match + 0.05
> place
> 2012-09-26 08:50:00.724612 I [992/1854] ProcessRequest
> mainserver.cpp:1360 (HandleAnnounce) - MainServer::ANN Monitor
> 2012-09-26 08:50:00.724618 I [992/1854] ProcessRequest
> mainserver.cpp:1362 (HandleAnnounce) - adding: taniwha-desktop as a
> client (events: 0)
> 2012-09-26 08:50:00.724853 I [992] ProcessRequest mainserver.cpp:1360
> (HandleAnnounce) - MainServer::ANN Monitor
> 2012-09-26 08:50:00.724868 I [992] ProcessRequest mainserver.cpp:1362
> (HandleAnnounce) - adding: taniwha-desktop as a client (events: 1)
> 2012-09-26 08:52:57.592815 N [992/1338] Expire autoexpire.cpp:263
> (CalcParams) - AutoExpire: CalcParams(): Max required Free Space:
> 100.0 GB w/freq: 15 min
> 2012-09-26 08:53:54.248818 I [992/1337] HouseKeeping
> housekeeper.cpp:225 (RunHouseKeeping) - Running housekeeping thread
> 2012-09-26 08:58:58.255214 I [992/1337] HouseKeeping
> housekeeper.cpp:225 (RunHouseKeeping) - Running housekeeping thread
> 2012-09-26 09:04:04.261442 I [992/1337] HouseKeeping
> housekeeper.cpp:225 (RunHouseKeeping) - Running housekeeping thread
> 2012-09-26 09:07:57.596617 N [992/1338] Expire autoexpire.cpp:263
> (CalcParams) - AutoExpire: CalcParams(): Max required Free Space:
> 100.0 GB w/freq: 15 min
> 2012-09-26 09:09:07.267769 I [992/1337] HouseKeeping
> housekeeper.cpp:225 (RunHouseKeeping) - Running housekeeping thread
> 2012-09-26 09:14:13.274213 I [992/1337] HouseKeeping
> housekeeper.cpp:225 (RunHouseKeeping) - Running housekeeping thread
> 2012-09-26 09:19:14.280495 I [992/1337] HouseKeeping
> housekeeper.cpp:225 (RunHouseKeeping) - Running housekeeping thread
> 2012-09-26 09:22:57.599553 N [992/1338] Expire autoexpire.cpp:263
> (CalcParams) - AutoExpire: CalcParams(): Max required Free Space:
> 100.0 GB w/freq: 15 min
>
>
> Looks like some housekeeping, metadata, SystemManager I/O process is
> happening around the pixellation time (0843-0845).
>
I also have a seagate 1TB drive approx 50% full; Haupage 2200; O/S  
Mythbuntu 11.10
No autoexpires are happening.
I will check my logs tomorrow if I get time.
I also see the skip back effect where the pixellation clears,but this only  
works after the "peak" of the distortion is passed and the picture is  
still showing a low level of distortion rather than becoming totally clear.

-Paul