Welcome to Doom9's Forum, THE in-place to be for everyone interested in DVD conversion. Before you start posting please read the forum rules. By posting to this forum you agree to abide by the rules.

 Doom9's Forum Frame corruption with BluRay stream
 Register FAQ Calendar Search Today's Posts Mark Forums Read

 18th May 2012, 20:11 #1  |  Link tormento Acid fr0g     Join Date: May 2002 Location: Italy Posts: 762 Frame corruption with BluRay stream I have a small problem I can't understand. I was trying to encode a BD I own to clean from noise and went thru some frame corruption. I thought it could be DGDecodeNV but the preview window gives me no error. I thought it could be the script but even a very simple Source script gives me the same error while previewing in MeGUI video player. The only thing it could be is AviSynth, even a simple script such as: PHP Code:  LoadPlugin("D:\eseguibili\media\dgdecnv\DGDecodeNV.dll") DGSource("E:\in\_Cielo di fuoco\Twelve O'Clock High.dgi")   Gives me corrupted frames. Any hint? Here's an unmodified sample of the m2ts. __________________ tormento@ircnet
 18th May 2012, 21:42 #2  |  Link Guest Guest   Join Date: Jan 2002 Posts: 21,924 Please give info: Nvidia graphics type DGDecNV version Nvidia driver version OS
 18th May 2012, 21:47 #3  |  Link SEt Registered User   Join Date: Aug 2007 Posts: 374 I see exactly the same corruption with Avisynth 2.6 alpha 3, so it's definitely not my builds or MT related. Please keep this thread discussion for things related to the builds and MT only – general questions should go elsewhere (for example it's best to ask neuron2 directly about this issue and create thread in Avisynth Usage section). PS: neuron2 found it already
 18th May 2012, 21:56 #4  |  Link Guest Guest   Join Date: Jan 2002 Posts: 21,924 Moved posts out to this new thread. I see the issue on a system with DGDecNV 2041 and 8500 GT running 301.24, but not on a system with DGDecNV 2042 and 240 running 296.10. I will try tonight on my 460 and 520 systems. We need to find the relevant variable. EDIT: Just upgraded the 240 system to 301.32 and it still runs fine. EDIT2: This is all irrelevant because I was testing the ES, to "simplify things". Last edited by Guest; 18th May 2012 at 23:52.
 18th May 2012, 22:31 #5  |  Link Boulder Pig on the wing     Join Date: Mar 2002 Location: Hollola, Finland Posts: 4,289 My infamous 520 decodes it fine with DGDecNV 2042 and drivers v296.10. (Apart from the beginning of the stream) __________________ And if the band you're in starts playing different tunes I'll see you on the dark side of the Moon...
 18th May 2012, 23:01 #6  |  Link Guest Guest   Join Date: Jan 2002 Posts: 21,924 Have you tried jumping around on the timeline? And the start should be fine anyway, as it starts with an IDR. @tormento Please try demuxing the ES and then try with that. I predict it will be fine. I'm not sure yet, but it seems that the transport demuxing is getting confused. Standby for more analysis... Last edited by Guest; 18th May 2012 at 23:04.
 18th May 2012, 23:03 #7  |  Link tormento Acid fr0g     Join Date: May 2002 Location: Italy Posts: 762 DGDecodeNV 2042 GTX 460 with Nvidia drivers 301.40_x64 on Win7_x64 AviSynth 2.6_MT latest Tried CUDA and VP rendering, same results. Can't understand why DGDecodeNV window gives a perfect rendering. What the "ES" is? Do you need some other, longer parts where the error appears? __________________ tormento@ircnet Last edited by tormento; 18th May 2012 at 23:06.
 18th May 2012, 23:25 #8  |  Link Guest Guest   Join Date: Jan 2002 Posts: 21,924 ES means elementary stream, i.e., the raw 264. Demux the video and make a project with that. It will work OK. DGIndexNV works fine because it just scans forward in the stream until the next I frame, it doesn't do random access. I know what the problem is. The IDR start code 00 00 01 65 is split between transport packets. I thought I had fixed this once but apparently it wasn't catching all the cases. Fix on the way... Last edited by Guest; 18th May 2012 at 23:29.
 18th May 2012, 23:50 #9  |  Link Guest Guest   Join Date: Jan 2002 Posts: 21,924 The corner case I missed was this: transport packet with beginning of IDR start code 47 .. .. .. 00 00 transport packets with audio 47 .. .. .. (lots of them!) transport packet with remainder of IDR start code 47 .. .. .. 01 65 My fix had assumed that the two video packets would be contiguous. Back to the drawing board for this fix. Last edited by Guest; 19th May 2012 at 03:54.
19th May 2012, 03:55   #10  |  Link
Guest
Guest

Join Date: Jan 2002
Posts: 21,924
Quote:
 Originally Posted by Boulder My infamous 520
Hey, I love the 520 now!

 19th May 2012, 09:14 #11  |  Link Boulder Pig on the wing     Join Date: Mar 2002 Location: Hollola, Finland Posts: 4,289 Edit: read all the posts, fix on the way __________________ And if the band you're in starts playing different tunes I'll see you on the dark side of the Moon...
 19th May 2012, 14:27 #12  |  Link Guest Guest   Join Date: Jan 2002 Posts: 21,924 I have it fixed and working properly. I need to do some regression testing and then I'll slipstream it into 2042.
 19th May 2012, 15:32 #13  |  Link tormento Acid fr0g     Join Date: May 2002 Location: Italy Posts: 762 Here I am for testing __________________ tormento@ircnet
 20th May 2012, 14:44 #14  |  Link Guest Guest   Join Date: Jan 2002 Posts: 21,924 I slipstreamed it into 2042. Re-download and update DGIndexNV.
23rd May 2012, 19:35   #15  |  Link
Guest
Guest

Join Date: Jan 2002
Posts: 21,924
Quote:
 Originally Posted by tormento Here I am for testing
How's the testing going? Thank you.

 23rd May 2012, 19:58 #16  |  Link tormento Acid fr0g     Join Date: May 2002 Location: Italy Posts: 762 Seems fine, thank you! __________________ tormento@ircnet
 24th May 2012, 04:16 #17  |  Link Guest Guest   Join Date: Jan 2002 Posts: 21,924 Sweet! You're welcome and thank you for bringing this to my attention.