My Humax Forum » Freeview HD » FVP 4000T, 5000T

Record live stops after a minute or two

(72 posts)
  1. User has not uploaded an avatar

    RichHumax

    member
    Joined: Feb '21
    Posts: 13

    offline

    Hi all, an update on my Instant Record issue on my 5000T, which despite reporting that it’ll record until the end of the EPG time, actually only records what’s in the buffer plus approx 1 minute. The machine also refuses to reformat the hard drive, reporting the failure to do so as ‘due to an unknown error’.

    Sadly, no amount of unplugging it from the mains or factory resetting made any difference to either issue.

    I removed the hard drive and installed it in my W10 PC, as I suggested I would. The SMART data showed both the drive as whole and its individual parameters to be in good health. It had 4 healthy active partitions and a small unallocated area of around 34MB. I decided to delete all 4 so that the drive effectively became new/blank with just one large unallocated area.

    I reinstalled it in the 5000T and the machine immediately formatted the drive, even before I then told it to! It now happily reformats when requested to, so did another factory reset and reformat. On removing the drive and installing it in the PC once more, it revealed that it was still healthy and had been formatted exactly back to the partitions as they originally were (but without the error which was stopping the 5000T from reformatting it I assume, whatever that was!).

    OK, now the bad news! The machine works perfectly, including being able to reformat but the Instant Record function remains broken exactly as before.

    To me, the narrowness of the fault still seems to point to a Jan 21 update bug which seems to affect some versions of hardware. Any minor flaw in its actual loading would surely cause far more catastrophic failures in the machine’s operation, as would a hardware failure?

    I can sadly go no further but hope my investigations may help someone else to crack the problem!

    | Tue 2 Mar 2021 16:34:39 #31 |
  2. User has not uploaded an avatar

    Martin Liddle

    special member
    Joined: Feb '11
    Posts: 4,619

    offline

    RichHumax - 2 hours ago  » 
    Hi all, an update on my Instant Record issue on my 5000T, which despite reporting that it’ll record until the end of the EPG time, actually only records what’s in the buffer plus approx 1 minute.

    I tried an instant recording and it worked perfectly (Box on Channel 4, about 45 minutes into a hour programme). It produced a perfect 1 hour recording. Can you provide a precise recipe for triggering the problem please.

    Did you record the Values of the individual SMART parameters? I am interested in the Raw values of the following attributes: 05 reallocated sector count, 197 Current pending sector count and 198 Offline uncorrectable count.

    To me, the narrowness of the fault still seems to point to a Jan 21 update bug which seems to affect some versions of hardware. Any minor flaw in its actual loading would surely cause far more catastrophic failures in the machine’s operation, as would a hardware failure?

    According to what Barry has said the change in the latest update was a fix for the Youtube search function which involved updating the version of the Opera browser that the Apps use. It doesn't immediately seem to me that it is the sort of change to introduce a problem with instant recording but who knows. Is your box an FVP-4000T or an FVP-5000T?

    | Tue 2 Mar 2021 19:08:37 #32 |
  3. User has not uploaded an avatar

    RichHumax

    member
    Joined: Feb '21
    Posts: 13

    offline

    Thank you Martin. The faulty behaviour occurs in all situations the Instant Record is used and 100% of the time. In the situation you describe, I would have ended up with approx a 46 min recording, 45 from buffer plus about another 1 min. I have uploaded two JPEGs, one of my reported SMART values and one of the partitioning on the disk (Disk 1). My ID values are slightly different to yours but the data you request is there I think. Humax also made me aware of the nature of the update, however I do not share their belief that it couldn't have effected other elements of the software. I accept that it may not have done but such coding changes often have unexpected consequences or accidentally inject errors. It's a 5000T.

    Attachments

    1. Disk_Man.jpg (643.6 KB, 5 downloads) 3 years old
    2. SMART.jpg (518.9 KB, 0 downloads) 3 years old
    | Wed 3 Mar 2021 10:45:09 #33 |
  4. User has not uploaded an avatar

    Martin Liddle

    special member
    Joined: Feb '11
    Posts: 4,619

    offline

    RichHumax - 1 hour ago  » 
    The faulty behaviour occurs in all situations the Instant Record is used and 100% of the time. In the situation you describe, I would have ended up with approx a 46 min recording, 45 from buffer plus about another 1 min. I have uploaded two JPEGs, one of my reported SMART values and one of the partitioning on the disk (Disk 1).

    Certainly to my eyes the hard drive SMART data looks absolutely fine and you have now reformatted the file system so that should also be fine so my initial theory wasn't correct.

    Humax also made me aware of the nature of the update, however I do not share their belief that it couldn't have effected other elements of the software. I accept that it may not have done but such coding changes often have unexpected consequences or accidentally inject errors. It's a 5000T.

    I am still doubtful about your software theory (why does my box do perfect instant records and your box doesn't; why aren't there more than two reports of this problem) but to test it have you considered disconnecting the box from the Internet and installing a previous version of the firmware? If that works it would be strong evidence to present to Humax. I have a copy of a previous version of the software if you need it. I am running out of other ideas about this but my one other thought is that it just might be a problem with the file system on the NVRAM; you could ask Humax if they have system flush update file for the FVP-5000T; they exist for some Humax boxes and are designed to clean up issues with the NVRAM (usually most useful for problems with boxes not booting properly but have been known to fix other obscure problems).

    | Wed 3 Mar 2021 12:39:59 #34 |
  5. User has not uploaded an avatar

    stevem1960

    member
    Joined: Jan '21
    Posts: 12

    offline

    My problem sounds almost exactly the same as RichHumax describes, on pressing the instant record button it would record what was in the buffer plus a variable but short additional amount of 1-5 minutes (and I may be getting confused with how much EXTRA was recorded after pressing the button and how much had already been buffered, it may have been always only 1 minute extra). This was happening for me before the recent software update so I would discount the YouTube changes. Also my problem has now gone away after a simple power off/on. I know this has not happened for RichHumax but in my experience as a software developer this screams uninitialised variable. For example, for me it powered up in a low or zero value that allows it to work up until such time as it is left with a high value whereas for Rich its power up value in ram is already too high to allow the function to work properly. Its a simple fault in software, easily fixed once you find the right variable.
    Also it seems to me to be more linked with the program guide (or whatever gives the program information) rather than the delay buffer because it always captures what has already been watched, just not what is yet to come for that program

    | Wed 3 Mar 2021 13:37:42 #35 |
  6. User has not uploaded an avatar

    RichHumax

    member
    Joined: Feb '21
    Posts: 13

    offline

    Hi Martin and Steve, thank you once again for your contributions. Yet another update as I think we’re closely circling the issue now!

    Before receiving your suggestions, I decided this morning to see if I could reinstall the Jan 21 software using an USB. I believe the original update must have automatically occurred via LAN as OTA no longer exists I assume? As asking the machine to check for new software with the internet connected resulted in a ‘nothing new available’ response, I decided to unplug the LAN connection (WLAN wasn’t setup on the machine at that point) before trying the USB. (I later learned that this was unnecessary as the USB seems to take precedence.) Much to my surprise, the machine reported new software on the USB and happily installed it, reporting the various stages on screen as it went. After completion I checked all the software details and of course they were exactly the same as before but the machine did seem to actually reinstall the software correctly, rather than just pretending to!

    OK, all round joy, the Instant Record now keeps going, problem over, it must have been something to do with the NVRAM being slightly corrupted perhaps, although for the machine to perform as well as it did under such circumstances felt wrong.

    However, long story short, my joy was short lived, I soon discovered that it actually only keeps on recording if the internet is not reconnected (via LAN or WLAN). Once the internet is re-established, the recording stops after approx 60 secs, even when its been happily recording for 30 or more mins up to that point. The behaviour is perfectly reproduceable 100% of the time and once connected to the internet, the original faulty behaviour remains until it is once again removed.

    It would seem to me that something associated with real time EPG data is being inappropriately addressed and acted on. I sadly don’t have any current experience of the protocols but assume once again that the Jan 21 update must have a bug but then again how do you explain that so few people seem to have it? Perhaps a NVRAM flush is needed or an earlier Software Version to see if the issue goes away?

    | Wed 3 Mar 2021 15:29:12 #36 |
  7. User has not uploaded an avatar

    Martin Liddle

    special member
    Joined: Feb '11
    Posts: 4,619

    offline

    RichHumax - 18 mins ago  » 
    OK, all round joy, the Instant Record now keeps going, problem over, it must have been something to do with the NVRAM being slightly corrupted perhaps, although for the machine to perform as well as it did under such circumstances felt wrong.
    However, long story short, my joy was short lived, I soon discovered that it actually only keeps on recording if the internet is not reconnected (via LAN or WLAN). Once the internet is re-established, the recording stops after approx 60 secs, even when its been happily recording for 30 or more mins up to that point. The behaviour is perfectly reproduceable 100% of the time and once connected to the internet, the original faulty behaviour remains until it is once again removed.

    That is a very interesting discovery but I don't understand what is happening. For what little it is worth my Internet is permanently connected.

    I sadly don’t have any current experience of the protocols but assume once again that the Jan 21 update must have a bug but then again how do you explain that so few people seem to have it? Perhaps a NVRAM flush is needed or an earlier Software Version to see if the issue goes away?

    I just don't know. It isn't unknown for one person to have a very reproducible problem with their Humax that nobody else can replicate; often the problem eventually goes away after a while.

    | Wed 3 Mar 2021 15:57:07 #37 |
  8. User has not uploaded an avatar

    Fyl

    member
    Joined: Nov '16
    Posts: 17

    offline

    Hi Rich,

    I'll have to try that.

    I tried a long power-down - that didn't fix it on my 4000T.

    Then I did Graham's suggestion from post #8 and left it paused for a while. It then created a recording similar in length (I didn't check the exact length relative to the pause).

    Now it seems to be creating a recording the length(-ish) of the buffer.

    | Wed 3 Mar 2021 17:49:08 #38 |
  9. grahamlthompson

    grahamlthompson

    special member
    Joined: Feb '11
    Posts: 14,442

    offline

    Fyl - 1 hour ago  » 
    Hi Rich,
    I'll have to try that.
    I tried a long power-down - that didn't fix it on my 4000T.
    Then I did Graham's suggestion from post #8 and left it paused for a while. It then created a recording similar in length (I didn't check the exact length relative to the pause).
    Now it seems to be creating a recording the length(-ish) of the buffer.

    If I remember correctly a second push of the instant record button allows you to extend the recording time. So you actually get some of the next programme on the end of the recording.

    At the moment using the Aura as the primary recorder.

    | Wed 3 Mar 2021 19:49:01 #39 |
  10. User has not uploaded an avatar

    Fyl

    member
    Joined: Nov '16
    Posts: 17

    offline

    Reapplying the update hasn't fixed it for me.

    | Thu 4 Mar 2021 11:58:13 #40 |

RSS feed for this topic

Reply »

You must log in to post.