Gonna crank a new release of DGMPGDec

[Guest]

I'm planning to crank a new release of DGMPGDec. I know of a bug with still handling that I want to fix. Can someone give me a list of known bugs that you'd like to see fixed? After the fix release I'll address new features.

Thank you.

[Cyberia]

Read the stickies :)

[Guest]

That's what I was hoping to avoid, but if you insist...

[len0x]

One thing that I hacked myself already but would be nice to have in official version: to always decode LPCM audio tracks even if just demuxing was selected (so that if you don't know the type of audio beforehand DGIndex always processes selected audio). This doesn't have to be for "Demux All", just "Demux" mode.

[Mouse]

neruon2, any chance you can read in the thread about AutoGK (start from page 165 -> (part about audio getting fuxored ;) in MPEG2 DVD-compliant streams, captured from DTV).

[Leak]

Originally posted by neuron2
I'm planning to crank a new release of DGMPGDec. I know of a bug with still handling that I want to fix. Can someone give me a list of known bugs that you'd like to see fixed? After the fix release I'll address new features.

Well, there's at least one DVD (Ghost In The Shell: Stand Alone Complex Vol. 1 as released in the US) that gets misdecoded frames upon seeking with the current version of DGIndex/DGDecode:

http://forum.doom9.org/showthread.php?s=&threadid=85209

With TIVTC and some cache wonkiness, I've managed to have it seek to every 5th frame in the file while processing it, and the result was horrible.

With a fixed version of TIVTC, it reads linearly through the file without seeking (thanks to AviSynth's cache), and all is well. Using a D2V file made with Nic's DVD2AVI_Nic still has some frames that get b0rked when seeked to, but it works a lot better; as I wrote in the thread above, the D2V file produced by his version contains totally different offsets... :confused:

np: Pan Sonic - Selittämätön / Inexplicable (Kesto (234.48:4) (Disc 3))

[Nic]

The offsets aren't totally different, but they are different (they aren't as accurate as DGIndex AFAIK though).

The weird thing is that your sample works with DVD2AVI 1.76 and MPEG2Dec3 Ok. So I believe it is something that can be fixed in DGDecode.DLL

-Nic

[blahh]

There's a little problem with mono-LPCM, which I wrongly reported over here already...

http://forum.doom9.org/showthread.php?s=&threadid=87883

[len0x]

Originally posted by blahh
There's a little problem with mono-LPCM

Don, basically its not about mono, but about 22KHz LPCM streams that are not supported in DGIndex and get outputted as 44KHz (and therefore are played at the speed of 2x).

[blahh]

Hmm... Looks like the bug is still present. :(

[Guest]

If you want a fix, you'll need to provide a test file for me.

[blahh]

Ok, but I won't be able to upload it until friday, because my traffic-limit is almost reached.

[Prettz]

Heh, I'd like some closure to my audio desync problems with the Akira DVD, but raymod2 disappeared and never came back.

[Guest]

It's not always easy to identify the cause of a desync, because there are so many possible causes. It's even harder when one doesn't possess the DVD in question. I can't afford to buy every DVD that posters mention.

If you know that the delay is off by 100ms, why don't you just make that adjustment and be happy?

[Prettz]

Originally posted by neuron2
It's not always easy to identify the cause of a desync, because there are so many possible causes. It's even harder when one doesn't possess the DVD in question. I can't afford to buy every DVD that posters mention.

If you know that the delay is off by 100ms, why don't you just make that adjustment and be happy?
The exact delay is unknown (it's in the neighborhood of 100ms in a particular scene). It only becomes desynced somewhere in the middle of the film, where something in the VOB seems to confuse most programs. It's not worth the trouble for me unless I can get a process that's guaranteed to give a correct result. So my problem isn't an absolute priority, but I'll still be waiting and watching for a solution to be found.

[Guest]

Originally posted by Prettz
The exact delay is unknown (it's in the neighborhood of 100ms in a particular scene). It only becomes desynced somewhere in the middle of the film, where something in the VOB seems to confuse most programs. It's not worth the trouble for me unless I can get a process that's guaranteed to give a correct result. So my problem isn't an absolute priority, but I'll still be waiting and watching for a solution to be found. That's new information. I don't recall you saying that the desync started only halfway through the file. As you may know, DGIndex determines the delay only once at the beginning of the clip. It would be interesting to try to determine the cause of the issue, but I do not own this DVD. And even if DGIndex could report multiple delays and their starting points, what could you do with that information?

[nnigam]

How about having it create seperate files when the delay/rate changes. It could be used to split up the video to seperate files, each with correct audio synching, or to drop sections if feesable.

[Napoleon]

How about the fact that DGIndex doesn't do Forcefilm from Robot4Rip like the old DVD2AVI did? I'm not sure if this is exactly where to post this issue...

[Guest]

Originally posted by Napoleon
How about the fact that DGIndex doesn't do Forcefilm from Robot4Rip like the old DVD2AVI did? I'm not sure if this is exactly where to post this issue... Is there something that has changed in DGIndex such that this third party application has a problem? If so, what is it? I've never heard of "Robot4Rip", so if you want a solution, you'll have to tell me what needs to be changed in DGIndex. If you can't do that, I suggest contacting the author of "Robot4Rip". Perhaps she/he can explain things so that I can make any change that might be required.

[Napoleon]

Robot4Rip is a program that initiates DVDecrypter, DVD2AVI (now DGIndex) and finally BeSweet to simplify the ripping from DVD and setting up the required files for use with Gordian Knot. I'm surprised you've never heard of it! I think this issue was mentioned in the latest Gordian Knot guide, and it was supposed to being worked on. Or suggested to be worked on. You can download it for yourself, it's included in the latest Gordian Knot rippack.

If the movie's not ForcedFilmed by DVD2AVI/DGIndex, then it has to be IVTC'd by Gordian Knot later, which doesn't seem to result in as clean an image. (I've considered going back to a previous version of Gordian Knot because of this issue with DGIndex- it doesn't seem to be an "upgrade" of DVD2AVI if it doesn't work right with Robot4Rip.)
Somehow they'd work together to determine if a movie film had 95% film and if so then would "forcefilm" it.

[len0x]

Originally posted by Napoleon
I think this issue was mentioned in the latest Gordian Knot guide, and it was supposed to being worked on. Or suggested to be worked on.

Where did you get that? Development and support for R4R ended long time ago and its clearly indicated that some features of it will no longer work on info tab of GK (because of new structure of d2v, parsing inside R4R has to be rewritten). That's it.

[Napoleon]

It says on the GK info tab:

ALSO IMPORTANT NOTE ABOUT ROBOT4RIP: its been put back into the pack but make sure that you're not using DVDDecrypter/VStrip/internal routines to demux audio - use only DGIndex with it (otherwise you'll get out of synch audio). Also disable Auto Force Film option as it won't work properly with DGIndex D2V files.

All I'm wondering is, why doesn't DGIndex "ForceFilm" (Convert the FPS)like the old DVD2AVI in R4R? Why was R4R discontinued? I don't know any other way to get the files for GK. Am I right to think that having DVD2AVI "forcefilm" during the rip process resulted in a better rip than having Gordian Knot IVTC it later?

I know I read in a guide somewhere that the writer was going to bring this issue up, but I can't find where I read that now.

[len0x]

R4R was discontinued because the developer dropped it and is no longer around. D2V structure between old DVD2AVI and new DGindex is changed and incompatible, so all programs that uses D2V files directly have to be updated.

P.S. you should learn how to demux audio and video manually via DGIndex - there are plenty of guides around.

[Guest]

Originally posted by Napoleon
All I'm wondering is, why doesn't DGIndex "ForceFilm" (Convert the FPS)like the old DVD2AVI in R4R? There are two ways to get FF out of DGIndex. You can set the appropriate option (either through the menus or through the CLI) before saving your project, or you can edit two lines of the D2V file after the project was saved as non-FF. If R4R was using the latter method, the change of D2V file format may have rendered R4R's algorithm for finding the right lines incorrect.

I try to avoid changes to the D2V format unless absolutely necessary, but sometimes progress requires breaking with the past.

[Napoleon]

If I knew anything about programming, I'd revive Robot4Rip.

Rest In Peace - Robot 4 Rip

[lark]

it's never too late to start learning.
regards
t :)