Random Megui error trying to encode audio

[Hotpocketdeath]

I keep getting the error shown in the image but I can not figure out what is causing it or how to fix it.

http://www.hotpocketdeath.org/tmp/Megui%20error.jpg
As you can see, similar jobs run fine, while others error out randomly.

Now I know for a fact my system is fine and stable.

Does anyone know if there is a fix for this? I hate it if I set up a bunch of jobs to run while I'm asleep or at work and come to find only half the jobs have been completed because of this error instead of being completed while I am not using my computer.

Edit: Oh, and this error only occurs at the start of the encoding audio jobs. Nowhere else. It is the exact same every time it occurs.

[Sharktooth]

what version of nero encoder are you using?

[Hotpocketdeath]

The latest.

1.3.3.0.

If you are saying it might be the cause, then I'll downgrade a version and see if it makes a difference.

If I can even find a previous version.

[Sharktooth]

are you using the standard .exe or the SSEx one?

[Hotpocketdeath]

There is no SSE version for the 1.3.3. They unified the two version in this latest release.

I have been trying out the previous version before that 1.1.34, out of about 30-40 encodes, I got 2 errors, One was the memory error, the other was an exception error.

So it still seems to happen with previous versions, just not nearly as often.

[tebasuna51]

Seems the error is at decoder time (dimzon_avs_getaframe).

Ramdomly?
Occurs always with the same Ac3 input file?
If yes:
Please test the conflictive Ac3 files with DelayCut and put the log errors, if any, to isolate the problem.

[Hotpocketdeath]

Seems the error is at decoder time (dimzon_avs_getaframe).

Ramdomly?
Occurs always with the same Ac3 input file?
If yes:
Please test the conflictive Ac3 files with DelayCut and put the log errors, if any, to isolate the problem.

The problem is it's too random for that. There's no way to tell which Ac3 files would be affected. When I mean the error is the same, I mean it's always the same cause of the error and the place where it occurs. The jobs the errors occur on are totally random.

If a job errors out, I simply go back to the queue and change it's status to waiting from error and try the encode again, most of the time, it will work on the second attempt with no change other change to the job itself.

[Sharktooth]

have you tested your hardware for possible faults (expecially the RAM)?
is your hardware (CPU/RAM/Videocard) overclocked (if so remove ANY overclocking and try again)?

[Hotpocketdeath]

I use Linx and Prime95 for check system stability. At my current settings, I get no errors from those stress tests.

I am overclocking an i7 to 4.1ghz. My Memory is actually underclocked, rated for 1600 7,7,7,24. It's currently running 1550 8,8,8,24, and passes Memtest. Of all the apps I run, MeGui is the only one that I get an error on.

[Doom9]

Linx? Prime95 pushes the CPU.. but for memory I'm running memtest86

[Hotpocketdeath]

Linx? Prime95 pushes the CPU.. but for memory I'm running memtest86

Large FFT's in prime95 pushes the memory subsystem. Small FFT's only push the CPU.

LinX (http://www.xtremesystems.org/forums/showthread.php?t=201670) is a linpack stress test similar to Intel Burn Test, only more reliable and informative. IBT however, does not work correctly on i7's, which is another reason why I use LinX, which pushes an i7 to 100% load.

And I use Memtest86 and Memtest (http://hcidesign.com/memtest/) for windows.

What's really odd is I have just done some 40-50 tests just now after switching back to the latest version of Nero AAC Encoder and not a single error occurred. And that's with 4 separate audio encodes going at once.