Glitches on Recordings

ian_j

Member
Lately we've been getting glitches on nearly every TV recording:

The Secret Life of Sue Townsend
The Story of Skinhead with Don Letts
Arena: The Roundhouse - The People's Palace
Catching History's Criminals
No Such Thing as the News
Sleuths, Spies & Sorcerers Andrew Marr's Paperback Heroes
Gardeners' World
Still Game

All have at least one picture breakup, the one link is that they are all from BBC HD channels.

I've checked the signal strengths and everything looks fine.

I've started a manual recording and tried turning items connected/near the Humax off and on but nothing so far has triggered picture breakup on a recording.

Turned the central heating off/on.

Turned lights off/on

I'm stumped what to try next.
 
Are the glitches on the actual transmission if you watch live, or just on recordings?
 
Is there anything else going on the box at the same time? Parallel recordings, recently started/finished recordings, custom firmware processing...?

Try watching a program live (using the Humax) whilst it is recording - do you see any breakups in the live broadcasts
 
Check the disk. A failing drive or file system causes lots of disk operation retries, which overload the processor and introduce glitches (live and recorded).
 
Are the glitches on the actual transmission if you watch live, or just on recordings?

We very rarely watch anything live, there is a recording for tonight at 22:00 I'll make sure to watch that live.


Is there anything else going on the box at the same time? Parallel recordings, recently started/finished recordings, custom firmware processing...?

I'll keep an eye out while watching a live broadcast.


Check the disk. A failing drive or file system causes lots of disk operation retries, which overload the processor and introduce glitches (live and recorded).

I hope not I only replaced the dead disk last month, but who knows.

Images below of current drive status:

2016 10 25 - Humax HDR Fox T2 (humax) - Disk Information.png 2016 10 25 - Humax HDR Fox T2 (humax) - Attributes.png
 
Well that seemed painless, anything on the following look like it needs further attention?

Code:
Maintenance mode is initialising... Please wait...
...
Menu version 1.20
Enter system PIN: ****

      /---------------------------------------------\
      |  M A I N T E N A N C E   M O D E   M E N U  |
      \---------------------------------------------/

  [ Humax HDR-Fox T2 (humax) 1.03.12/3.10 ]

fixdisk - Check and repair hard disk.
   short - Run short hard-disk self test.
    long - Run long hard-disk self test.
   check - Check self-test progress.
    gptf - Re-format disk using GPT scheme.
     epg - Clear persistent EPG data.
    dlna - Reset DLNA server database.
       x - Leave maintenance mode (Humax will restart).
    diag - Run a diagnostic.
     cli - System command line (advanced users).

Please select option: fixdisk
Any additional options (or press return for none):
Are you sure you wish to run the hard disk checker? [Y/N] y
Running /bin/fix-disk
Custom firmware version 3.10


Checking disk sda (4096 byte sectors)

Unmounted /dev/sda1
Unmounted /dev/sda2
Unmounted /dev/sda3


Running short disk self test

No pending sectors found - skipping sector repair

Checking partition tables...

MBR Status: MBR only
GPT Status: not present

Using superblock 0 on sda1
Using superblock 0 on sda2
Using superblock 0 on sda3


Checking partition /dev/sda3...
e2fsck 1.42.13 (17-May-2015)
Pass 1: Checking inodes, blocks, and sizes
Pass 2: Checking directory structure
Pass 3: Checking directory connectivity
/lost+found not found.  Create? yes

Pass 4: Checking reference counts
Pass 5: Checking group summary information

hmx_int_stor: ***** FILE SYSTEM WAS MODIFIED *****
hmx_int_stor: 14/655776 files (0.0% non-contiguous), 116683/2622611 blocks

Checking partition /dev/sda1...
e2fsck 1.42.13 (17-May-2015)
Pass 1: Checking inodes, blocks, and sizes
Pass 2: Checking directory structure
Pass 3: Checking directory connectivity
/lost+found not found.  Create? yes

Pass 4: Checking reference counts
Pass 5: Checking group summary information

hmx_int_stor: ***** FILE SYSTEM WAS MODIFIED *****
hmx_int_stor: 15/65808 files (13.3% non-contiguous), 16048/263062 blocks

Creating swap file...
Setting up swapspace version 1, size = 1073737728 bytes
UUID=d5161d2a-9939-4eb2-b8a3-77f4dd36efe8

Checking partition /dev/sda2...
e2fsck 1.42.13 (17-May-2015)
Pass 1: Checking inodes, blocks, and sizes
Pass 2: Checking directory structure
Pass 3: Checking directory connectivity
/lost+found not found.  Create? yes

Pass 4: Checking reference counts
Pass 5: Checking group summary information

hmx_int_stor: ***** FILE SYSTEM WAS MODIFIED *****
hmx_int_stor: 5938/60334080 files (1.4% non-contiguous), 17869556/241304331 blocks
Removing extra swap space.
Are you having problems with a delete loop? [Y/N]:
 
Of course it's painless - no harm in checking. Looks fine, presumably you replied "N" at the end.
 
I just meant on this run, if it found something I'm sure it would have taken a while to scan through and correct it.

I replied N to delete loop question.
 
I get glitches in recordings if I have any detect-ads processing going on at the same time (running non-chase & detect but not remove). When I tried to diagnose the problem with the forum's help, no cause was ever identified. Ever since, I have always performed detect-ads operations by manually triggering a sweeper rule when I know that nothing is due for recording.
 
I don't have any autodecrypt or similar auto process running other than on my radio recordings folder.

If I want to take anything off the Humax I manually decrypt & shrink items and always try to do it when the box isn't actively recording.

I don't know if I want it to be electrical interference or something else, the thought of pulling out all the audio/video kit fills me with dread.
 
I recorded 2 programs last night, Ch4 HD 21:55 to 22:40 andBBC4 HD 21:55 to 23:05

I watched the 1 hour program for 30 minutes with no glitches on the live broadcast.

I've just watched the program again and 20 minutes into the recording there is a glitch, I checked the other recording and there is a glitch at 20 minutes.
 
I've unplugged the ethernet cable and will try again tonight with 2x recordings at the same time.

I've also disabled Content Sharing & The Humax FTP Server, just to be sure.
 
Again both recordings last night had glitches.

When watched on the Humax they are in the 7th and 8th minute of each show (both started recording at 19:55), but when I check on the Humax via WebIF the recordings started 30 seconds apart:

Wed Oct 26 19:55:30 2016 BST
Wed Oct 26 19:55:01 2016 BST

So I would say yet again the glitches are at the same time.

What would this point to being the issue? What should I try to check next?
 
Is anything decrypting at 20:00 ? Check the logs. Probably something disk access related.
 
Code:
4936    26/10/2016 23:25:05 -   @ /mod/webif/plugin/sweeper/auto.hook:::sweeper::apply_rules:1139
4935    26/10/2016 23:25:05 -   @ /mod/webif/plugin/sweeper/auto.hook:file:1109
4934    26/10/2016 23:25:05 - Plugin error: could not read "/media/My Video/Little Lifetimes": No such file or directory
4933    26/10/2016 23:25:05 - moveset(/media/My Video/Little Lifetimes/Little Lifetimes_20161026_2255.ts) -> /media/My Video/Radio Recordings
4932    26/10/2016 23:25:05 - move /media/My Video/Little Lifetimes/Little Lifetimes_20161026_2255.ts to Radio Recordings
4931    26/10/2016 17:25:06 - Renaming /media/My Video/Radio Recordings/Man at the Helm_20161026_1125.ts to 2016 10 26 - Man at the Helm
4930    26/10/2016 12:05:07 -   @ /mod/webif/lib/bin/auto:::sweeper::scan:123
4929    26/10/2016 12:05:07 -   @ /mod/webif/plugin/sweeper/auto.hook:::sweeper::apply:1162
4928    26/10/2016 12:05:07 -   @ /mod/webif/plugin/sweeper/auto.hook:::sweeper::apply_rules:1139
4927    26/10/2016 12:05:07 -   @ /mod/webif/plugin/sweeper/auto.hook:file:1109
 
Again both recordings last night had glitches.

When watched on the Humax they are in the 7th and 8th minute of each show (both started recording at 19:55), but when I check on the Humax via WebIF the recordings started 30 seconds apart:

Wed Oct 26 19:55:30 2016 BST
Wed Oct 26 19:55:01 2016 BST

So I would say yet again the glitches are at the same time.

What would this point to being the issue? What should I try to check next?


I suspect this is when the RAM buffer is written to disk and Linux has to allocate some sectors to the file, so the disk is doing some seeking to set up the allocation uints and then write the data, and hence there is an increased current draw on the power supply. My Fox-T2 was doing similar things before it died, even after turning off AR marking and delaying detectadds to the middle of the night. My suspicion is the power supply has electrolytic capacitors that are drying out with age (there is a reason they are known as 'wet electrolytics' :cool:) and that the power supply can no longer sustain the peak current draw required. I have yet to prove this with my unit however.
 
Back
Top