I have it fixed and will release a build tomorrow. Thanks for pointing this out.

Email me a screenshot of the DGIndexNV Help/About dialog and attach the license.txt file.

Of course it's fixable.

Thanks for your test results. I confirm two issues with MKV: 1) two frames are missed at the end (480 vs. 478), and 2) failures of random access.

I'll fix this ASAP.

It can't be the same race condition because CUDA is not present.

However, if you can tell me how to duplicate it, I can investigate.

I tried the Matroska version and it was fine too.

Awaiting a stream to duplicate the issue...

I do not see any issues with your M2TS sample and your script. Have you tested with the M2TS? If so, please specify where you think the problems are *with that sample*.

Please test it by opening the script in VirtualDub, not by making an encode.

Hey, wait a minute, your input is MKV and not M2TS!

How did you rip it from the BluRay?

I can't open that MKV with anything. Nevertheless I will try to duplicate this.

That's great because now I know how to fix it. It's actually harmless but the error popup is an irritant that can be fixed.

Thank you for your willingness to perform the debugging steps that were needed.

Thank you for your answers. If possible, can you perform one last experiment? I have an idea that the problem is caused by my stream alignment algorithm for random access in program streams (VOB is a program stream). Can you please demux the M2V (elementary stream) from the VOB, open that in DGIndex...

May I ask for more details on the occurrence of the error? With the one VOB that you were able to make it happen with, does it ever happen if you simply preview from start to finish? Or does it happen only when you navigate on the timeline? If the latter is true, then a) does it seem to happen only ...

It's probably caused by your use of MT(). Didee has posted about this, telling which ways work for multithreading with DGDecNV.

Please try without MT() and report back.

I assume you didn't get the error when serving through your script but only in DGIndexNV, is that correct?

I've just emailed my Nvidia contact about it this morning. Hopefully I'll have a better understanding of it when he replies.

Ah, I see. Thank you.

Sorry, I don't know what you are talking about and I don't recognize those numbers.

Hank, please give me the screenshot. Thanks.

Those are strange driver version numbers!

Can you open the Nvidia control panel, do Help/System Information and post a capture of it? I'm expecting a number like 257.21.

Hank, can you give me details of the crash? DrWatson?

Also, try ConvertToYUY2() at the end of the script.

What version of Avisynth is it?

Version 2.70.01.05 works fine for me here.

You're welcome. Build 2027 is released.

All new builds will be listed in this thread as they are released. Please do not post problems or questions in this thread (they will be deleted). Start a new thread in the DGDecNV forum.

http://rationalqm.us/dgdecnv/dgdecnv.html

Don't worry, I have finished downloading it and fixed the problem. The problem was that the first video packet did not appear for over 30000 packets into the stream and I was aborting if it was not found within the first 10000. DGIndex uses a different method (allows 3 seconds rather than a fixed pa...

10MB would have been enough. It's going to take 2.5 hours just to download this!

Can you please give me the x264 version and command line and your script?

Also, is it possible to test the same encode manually without MEGUI?

Thank you.

You mean just playing in DGIndexNV, serving and playing in VirtualDub, or serving and encoding?

Since this is the support forum for my tools, I would first recommend using DGDecNV together with Avisynth.

A second possibility is the Avisynth filter DSS2() together with one of the following:

CoreAVC
DiAVC
ffdshow