Nova-t 500 recording issue (short or missing time)

[bgowland]

I also experienced this problem last week. I have a scheduled recording set from 22:00-23:00. I did a quick record of the brit awards from 21:00-23:00 but this recording ended at 22:00.

I'm confused - which recording ended at 22:00? The Brit Awards ran from 20:00 - 22:00. That's what it says in my 'Ready' list.

Cheers,
Brian

[gEd]

I made the times up (based on when I thought they were on) merely to illustrate the problem. The brits show (quick record) ended an hour early, when the scheduled recording started.

maybe gbpvr detected that the brits was crap and just switched it off....

[Bobins]

Set up my test last night BBC1 Holby City 01:05-02:05 and CH4 Brat Camp 01:15-02:15.

Strange thing happened, when the BBC1 recording started it started on the 2nd Nova-t tuner (Dev1 Path1) not the 1st. The CH4 recording then started on the Leadtek tuner.

Unfortunately, the EPG update cut in at just gone 2 o'clock and swamped the log. The EPG kicked in on the 1st Nova-T tuner (Dev1 Path0)

Both recordings finished ok so I didn't manage to reproduce the fault (probably because the recordings were not both on the Nova-T).

I'll try again tonight and will move the EPG update time till later.

[Bobins]

Moved EPG time to 05:00 and restarted recording service last night . Setup overlapping BBC1 and CH4 recordings and bugger me, both recorded perfectly from the Nova-T .

There were also no errors in the EPG update log unlike before.

This is very frustrating. The past 3 Saturdays Primeval has recorded from BBC1 and ER on CH4 has truncated when the Primeval recording stops. This has been consistent.

The only difference between last night and Saturdays is that Primeval & ER are automatically set season recordings whereas last nights were manually selected from the EGP list.

I'll try again tonight.

[Bobins]

Ok finally making some progress.

Cleared logs & restarted recording service. Set CH4 to record "Imagine" 22:35-23:30 and BBC1 to record "Eunuchs" 23:05-00:05.

When CH4 stopped recording the BBC1 recording was truncated. The logs are attached.

What was different from the previous nights? Well this time I started PVRX2 on the PC to programme the recordings instead of using the MVP. When I exit PVRX2, I get an error message although there is nothing in the PVRX2 logs.

The error is "PVRX2 has encountered a problem and needs to close"

Event ID 5000 Source .NET Runtime 2.0 Error
EventType : clr20r3 P1 : pvrx2.exe P2 : 1.0.0.0 P3 : 29
P4 : mscorlib P5 : 2.0.0.0 P6 : 461eee3d P7 : 3313 P8 : 10
P9 : system.objectdisposedexception

Whilst everything else seems to work ok afterwards, I'm beginning to suspect there's a correlation here. This error occurs every time I exit the PVRX2 screen.

I've also noticed the EPG update error has returned in the logs as well.

[nj2112]

I've tried a few things to try to resolve this...

Changed from Via usb drivers to standard usb driver all functioned correctly but the problem remained.

Changed to earlier nova-t-500 drivers CD3.3b version, problem remained.

Tried changing filter names for the nova-t-500, I thought maybe the similar filter names for each tuner might be causing this (bit of a long shot) but problem remained.

I have found out through testing that the issue can be triggered by manual cancellation.

Start a recording on one tuner then repeatedly start and cancel a recording on the other tuner while monitoring the recording service disk I/O (Process Explorer is handy for this). If the recoding service IO drops to virtually nothing then the problem has been triggered.

Unfortunately there is still a seemingly random aspect to this so it might take 5 or 10 or 25 or more cancellation cycles to to trigger the issue.

Still at least it gives us a test that doesn't take several hours of overlapping recordings.

For my set-up I have moved the nova-t-500 tuners down in priority below my nova-t-stick to mitigate this issue. ie times when both 500 tuners are active is reduced and so the chances of getting an affected recording are reduced.

I've tried everything I can think of without success and just don't know what else I could do to diagnose or resolve this.

I just though I'd add my input in case it might be of help to someone.

[gEd]

Sub will really need a set of log files (debug level) from the time when you were able to reproduce the problem.

I'll see if I can repro the problem using the method you have described.

[nj2112]

Sub will really need a set of log files (debug level) from the time when you were able to reproduce the problem.

I'll see if I can repro the problem using the method you have described.

I posted a set of debug (flushed) logs for this issue when I started this thread.

[nj2112]

Well it seems the nova-t stick also has a issue which may or may not be related.

In this case it will sometimes drop a chunk of data of random length and random timing during recordings, it's happened several times now and just as with the 500 issue the logs carry on as normal, nothing to indicate any problem.

[gEd]

well thanks to nj2112's tip, I managed to reproduce the problem on my nova-t 500.

Season recording of moving wallpaper in progress from 21:00-21:30.
I started and stopped ~8-10 "quick recordings" of "Walliams and Lucas New Heroes of Comedy" and after stopping each recording, I checked to see if the moving wallpaper recording file was still growing. Eventually at 21:13 it stopped growing although gbpvrtray still showed that the recording was in progress.

Logs attached.