DGMPGDec 1.2.0 RC5

[moonwatcher]

ok uploading now, used DVDDecrypter's "split by chapter".
uploading first chapter (+-100mb), let me know if you need another one, or its not a good way to cut.

i checked that the problem occures on it and it does.

[Guest]

Thank you, moonwatcher. Your help in fixing bugs is greatly appreciated.

[moonwatcher]

delighted to be paret of the fight. And besides i want it to work too :)

using your stuff for quite a while, you really do an incredible job.

[muppets]

originally posted by moonwatcher
checked with a diffrent PAL stream (no force film) and it works fine.
checked with the last vob (78mb):
when opened with DGIndex i get the followi8ng message:

WARNING! Opening GOP is not closed.
The first few frames may not be decoded correctly.

(i guess because its not the first VOB)

Did you try to process more than one file in one dgindex session?
I had the same warning and it went away when i closed dgindex after each file and then opened a new dgindex for every file that i wanted to process. Maybe that works for you also.

muppets

[Guest]

@moonwalker

I got a file about 71 MB in size.

I used DGIndex 1.2.0 RC2 to generate a D2V file with and without Force Film. I used this script to serve to VirtualDub:

loadplugin("dgdecode.dll")
mpeg2source("movie.d2v")

...using DGDecode 1.2.0 RC2, of course.

Everything operated perfectly normally, without any exception as you reported.

Can you please explain your exact steps to generate the issue? Also, please try on a differnt machine if possible. What version of VirtualDub did you serve to?

[moonwatcher]

First: i am still uploading, only 82% complete, sorry but i have a slow uplink.

second, tried serving to Latest virtualdub (1.6.3), virtualdubmod (latest), and media player classic, with exactly the same error.

regarding the steps.
ripped with DVDDecrypter.
open DGIndex, Load VOB file (single).
set force film.
save project.

will try later on the laptop...

edit: sorry for not checking before, doesnt happen when i dont force film.

[Guest]

I thought it was done because the filesize stopped getting bigger. I'll try again with the complete file.

[loni_blues]

I seem to be having a similar problem with 1.2.0 RC2, though in my case with or without forced film is the same.

Using this simple script:
LoadPlugin("C:\ARCHIV~1\GORDIA~1\AviSynthPlugins\dgdecode.dll")
mpeg2source("C:\DVD\Projects\Confessions\Deleted no ff.d2v")

Checked with Vdubmod 1.5.4.1 and 1.5.10.1 (build 2439). I get the message: "MPEG2Source: couldn't open source file, or absolete D2V file"
My PC is AMD Athlon 1.20 GHz, OS Win XP SP1.

Regards

[moonwatcher]

done uploading...

[Guest]

Originally posted by loni_blues
I seem to be having a similar problem with 1.2.0 RC2, though in my case with or without forced film is the same.

Using this simple script:
LoadPlugin("C:\ARCHIV~1\GORDIA~1\AviSynthPlugins\dgdecode.dll")
mpeg2source("C:\DVD\Projects\Confessions\Deleted no ff.d2v")

Checked with Vdubmod 1.5.4.1 and 1.5.10.1 (build 2439). I get the message: "MPEG2Source: couldn't open source file, or absolete D2V file"
My PC is AMD Athlon 1.20 GHz, OS Win XP SP1.

Regards Your DGIndex and DGDecode are mismatched. Please be sure you are executing the RC2 versions and remake the D2V file. You cannot load previously made D2V files.

[Guest]

@moonwatcher

Got the file and duplicated the problem. My source code is at home so when I get home this evening I will determine the cause.

[loni_blues]

neuron2,

Thanks a lot and sorry for the last post. I'm afraid I made some mess with uninstalling and reinstalling things.

Now I am getting the same "Evaluate: Unrecognized exception!" in Vdubmod as moonwatcher reports and only when I apply Forced Film.
All I can say is that it it invariably happens with every vob I've tried: No FF: success, FF: failure.
Kind regards,
loni_blues

[Guest]

@loni_blues

Yes, me too. I'll fix it and make an RC3 tonight.

[loni_blues]

neuron2,

Thanks!
In case it might bring some extra info for RC3 I've just found out this: the audio delays I get are different between the tracks demuxed with ff and those without ff (using RC2, of course). Those which have 0ms delay are the tracks demuxed with no ff applied. It seems all the problem goes to ff in RC2, doesn't it?
Regards.

[Guest]

Here's RC3. It's important to get the FF crash fixed.

http://neuron2.net/fixd2v/dgmpgdec120rc3.zip

I'll look into the delay issue now, loni_blues. How big of a difference did you see? It should never be more than one frame time (33ms for NTSC, 40ms for PAL). I know why it is happening and it will be very HARD to change. I was planning to write a journal entry about it. This will motivate me to do so.

Oh, RC3 has the DGParse and DGFix for the new D2V format, and it gives len0x what he was looking for in regard to the old 0,1,2,3 thingie. :)

[moonwatcher]

what delay issues?
i dont think the DVD has any angles in it.
it does look like a poor mastering (shame for such a greate movie).

will not using force film and using decomb to do the job help?
i am not sure i understand what could cause the behaviour you described.
any chance it has to do with the way i configured DVDDecrypter?
(i asked to leave only the first angle, even though i dont think there are any). any further info you want about the media?

[Guest]

@moonwatcher

I was all wrong. It was a straight bug in allocation of memory in the FF case. I hate being wrong, but there it is. :)

[moonwatcher]

so are we waiting for another version or is it already fixed on RC3?
BTW i would love to help programing if needed, i have quite some background in programing, but mainly enterprise, never did any A/V stuff. though i am quite good with algorithms in general.

[loni_blues]

You're right neuron2!

I have only tested NTSC (due to availability) and the differences in audio delay range from 16 to 34 ms (0 and 66ms with no ff and 16 and 100ms with ff, accordingly).

FF d2vs can be opened now with RC3!

Thanks!

[Guest]

Originally posted by moonwatcher
so are we waiting for another version or is it already fixed on RC3?
BTW i would love to help programing if needed, i have quite some background in programing, but mainly enterprise, never did any A/V stuff. though i am quite good with algorithms in general. It's fixed, moonwatcher. :D

Thank you for your offer of assistance. When I release 1.2.0 final, you can grab the source code and have a look at the development list sticky thread. If you feel like tackling something, just let me know.

[len0x]

Originally posted by neuron2
Oh, RC3 has the DGParse and DGFix for the new D2V format, and it gives len0x what he was looking for in regard to the old 0,1,2,3 thingie. :)

Thanks! :) Not that I mind having different schema (as long as I know how to interpret it), but strings are easier to compare than hex numbers =)

I was also surprised to see matrix info in RC2 when I managed to use debug info from RC1 already (I have custom build avisynth script debugger built for my programs), but its definitely faster to parse D2V then do special analysis pass :)

[cw_uk]

Firstly, nice work on these new builds of DGindex.
Secondly, I feel kind of stupid for asking this but what do we do with the output of dgparse.
Sample from last run,

[GOP START]
95463: 95463,95463..............2
95464: 95464,95464..............2
95465: 95465,95465..............2
95466: 95466,95466..............2
95467: 95467,95467..............2
95468: 95468,95468..............2
95469: 95469,95469..............2
95470: 95470,95470..............2
95471: 95471,95471..............2
95472: 95472,95472..............2
95473: 95473,95473..............2
95474: 95474,95474..............2
95475: 95475,95475..............2
95476: 95476,95476..............2
95477: 95477,95477..............2
[EOF]

[Guest]

You can use the output to determine whether pulldown is present and what patterns it has. The numbers on the right are the pulldown digits. They can be 0, 1, 2, or 3. See here for a technical discussion:

http://www.doom9.org/ivtc-tut.htm#TechnicalGuideToDVD2AVI

For example, I can see from the frame numbers to the left that there has not been any pulldown up to that point. I can conclude that you have a top-field first movie. If it is NTSC, it is probably interlaced but may be hard telecined. To distinguish the two, you'd need to step through the output and see if you get a pattern of 3 progressive frames followed by 2 interlaced frames.

[hippoth]

Hi neuron2,

can you bring back the autoincrement functionality? I mentioned it in another thread too. I know that we have a drag & drop function but it´s not the same like...you know what I mean ;)

Please!

[Guest]

I agree that it is not the same but, no, I don't know what you mean. You have to justify your request. Is there a problem for you with the current operation? If so, what is it?

[hippoth]

The problem I see: If I open a folder with a lot of vobs (e.g. a DVD folder)...I know whats the first vob but I don´t know whats the last needed vob...with "autoincrement open" I never had to think about because the old DGIndex former DVD2AVI took care of this

...you know now what I mean?

[Cyberia]

Here's the full difference between the methods:

You must draw a box around the files you want rather than double-clicking on the first one.

That's it.

Plus you gain the ability to selectively choose your files. Seems like a good trade off to me.

[hippoth]

Originally posted by Cyberia
You must draw a box around the files you want
Thats the problem...you must know the needed files...do you know always the main movie vobs?
Drag & drop is fine but I miss the good old double-click-on-the-first-file-method

[Guest]

Sorry, but the time to have brought this up is when the feature change was discussed here. If you were not around at that time, or not interested in the discussion, well, tough luck.

[Cyberia]

Originally posted by hippoth
Thats the problem...you must know the needed files...do you know always the main movie vobs?
Drag & drop is fine but I miss the good old double-click-on-the-first-file-method

Yep, I always know what vobs are main movie vobs. I could offer help in this regard...

How are you ripping the vobs? If you are using DVDDecrypter you can configure it to rip only the main movie vobs and not to split vobs. Would that help?

[hippoth]

Originally posted by neuron2
If you were not around at that time
Yes...thats my problem...I didn´t know it.

Originally posted by Cyberia
I could offer help in this regard...
No...I know how to handle a backup of a DVD...sure ;) (look at my registration date)
It´s just laziness to look for the right vobs because sometimes I backup the whole DVD to HD...not just the main movie files.

[Guest]

If I get some spare coding bandwidth, I'll put it back in.

[hippoth]

Oh great...be sure of my thankfulness! :)

[cw_uk]

Originally posted by neuron2
You can use the output to determine whether pulldown is present and what patterns it has. The numbers on the right are the pulldown digits. They can be 0, 1, 2, or 3. See here for a technical discussion:

http://www.doom9.org/ivtc-tut.htm#TechnicalGuideToDVD2AVI

For example, I can see from the frame numbers to the left that there has not been any pulldown up to that point. I can conclude that you have a top-field first movie. If it is NTSC, it is probably interlaced but may be hard telecined. To distinguish the two, you'd need to step through the output and see if you get a pattern of 3 progressive frames followed by 2 interlaced frames.

Thank you for that very clear description, it just knocked me off a bit where the day before i had used RC2 where i just told it the path and it spat it all out for me then the next day with RC3 i got that lol.

[Guest]

Originally posted by cw_uk
it just knocked me off a bit where the day before i had used RC2 where i just told it the path and it spat it all out for me then the next day with RC3 i got that lol. Now you're confusing me.
Spat all what out? There wasn't a compatible DGParse released until RC3.

You know what, if you don't know what DGParse is for, then there is no reason for you to be using it. :D

[Bluedan]

Neuron, I totally disagree with the current agreement!

hippoth, the "new" way is not even a click more when drawing a rectangle around the main movie VOBs for selection and if you were really comfortable with DVD structures you'd easily spot the main movie VOBs which lie anyway lined up next to each other in your folder because of the sequential numbering, even if you simply put a copy of the whole DVD on HD!
I cannot help thinking that you've become a bit rusty, he?
I feel ashamed that you dare nagging a busy coder like Donald just for your personal selfish comfort. :mad:

I don't want to sound too harsh, but the discussion about this point is already over, Neuron was right in his first reply to your request and I don't understand why he gives in...

The big part of the board members agreed on this, why should he waste his valuable time for this on a single users requests who is - as he admits himself - obviously lazy. And it' sooo obvious! I feel like this really doesn't deserve to be a reason to change the opening dialogue again !

I'm so glad that these people spend their time for writing free programmes while I didn't advance since doing assembler for the c=64!
Truly, I find this demanding! Period.

[Cyberia]

@Bluedan - I couldn't agree more. Well said.

[aaar9800]

yeah, come on...

If you know that the first file of the main movie is in the 3rd VTS (for example)- VTS_03_1.vob in other words, then I have no idea how someone could not understand that the end of the main movie would be VTS_03_X.vob, where X is the largest number it gets too in the 3rd VTS. Also its the one before the VTS_04_0.vob(If there is a 4th VTS) Anyway, I don't think that we should be saying whether or not neuron2 should add what hippoth asked for - it's for neuron2 to decide. I personally don't care if the autoincrement will be back, but do believe that it is completely unnecessary.

[Guest]

Please don't be hard on hippoth! I always encourage users to speak their minds. He's been polite. He has strong views; so do I! It's not a crime. I certainly wouldn't want him to be discouraged from making suggestions down the line.

So, I appreciate everybody's positions but we can all be a little tolerant, right? I knew you could. :)

[Bluedan]

Originally posted by neuron2
I always encourage users to speak their minds. :)

Well, then I'd like to have automatic AVS creation with, let't say...ehh... best settings, in the next major release, ok? Kindly-wise. ;)

No more anger, but sometimes it seems like people are forgetting, that this is supposed to be a community, where singles must stand back... couples welcome! ...Mhh, now it's getting the wrong way... :o

BTW, Neuron, by looking at your postage times I cannot figure out when you might have put on your night cup in the meantime? You're not actually 66 years old and feel refreshed after 4 hours of sleep?

[Guest]

@Bluedan

Was that the night I didn't sleep at all?

I'm 52 and look like this:

http://neuron2.net/board/viewtopic.php?t=619

Everybody is welcome to post their pic in that thread.

Please define "best settings". You know, I hope, that you can make your AVS template generate any default settings that you like.

[fccHandler]

You're 52??? :eek:

Is that human years or cat years?

Truthfully man, I would never have guessed it from your pic. I just hope I look that good when I'm 52! (13 years from now...)

[Guest]

Ya gotta be a swimmer to look that good. :)

[MacAddict]

Originally posted by fccHandler
You're 52??? :eek:

Is that human years or cat years?

Truthfully man, I would never have guessed it from your pic. I just hope I look that good when I'm 52! (13 years from now...) Totally agree! I should think even the celebrity types in Hollywood would be jealous of you Donald:) So the cold waters of Lake Michigan is the secret ay:D Personally I'm more jealous of where you live, a great city and certainly my favorite!

Anyways, no problems with delay or anything else here using RC3. Thanks again for the continued effort!

[zeus163]

I've been having some problems with the delay value that is being reported for audio by rc3 in my .ts files. Normally I open up my .ts files in Project X, cut out commercials, and then demux using project x. Next I would open up the created m2v in DGIndex. That way I can use the original audio. But some of the HD stations that I cap don't open up in Project X. If that is the case, I then edit out commercials with HDTV2MPEG2. I needed to do this for a recent cap and ran rc3 on it. The delay value it reported back was nothing on the ac3 file. Which surprised me greatly. When I muxed the two back together after encoding the file, the audio was out of sync. I then tried to run the edited .ts file through autoGordionKnot(?) v. 1.89 and the delay that was reported back was -365. When I ran that offset on the .ac3 file and muxed it with the encoded file it was perfectly in sync.

I then did another cap from the same channel and this time rc3 reported back a positive delay of 4880. In the meantime, I had upgraded AGK to 1.90 and it also reported back 4880, but when I tried this value with the encoded file it was super out of sync. I then used an older version of DGIndex and got a delay value of -357. When I used that on the .ac3 file and muxed back together, it was in sync.

I'm just curious to why there would be such significant differences and why the older versions seem to have the correct values. When I get home I can edit a short clip and send it off--if the cut clip also shows what is happening.

[len0x]

FYI AGK 1.89 was using DGIndex 1.1.0 (1.90 is using RC3), so that's what delay issues should be compared to.

[Guest]

@zeus163

You have to give me a clip that shows a significant difference when processed directly with 1.2.0RC3 and some earlier version that you think is correct. I don't want to hear anything about this cutter or that cutter, GK, etc. When you give me the file, tell me which version you are comparing 1.2.0RC3 to. Thank you.

[bob0r]

Now that DGMPGDec can open .PVA files. i got a question/request.

I have BBC.2.England.20.Jan.2005.pva which i open and i make a .d2v file. for example:

cap MPA DELAY -300ms.mpa
cap.d2v

Now notice the -300ms, when i encode like a xvid sample, and i merge the video/audio, ofcourse the audio is desynced.

I fix this by delaying the audio with -300ms in nandub.

This works...but,

Is it possible to sync the .mpa file with DGMPGDec?

PVAStrumento.exe (2.0.25) demuxes a .pva into video/audio with an option to sync the audio, so it is possible i guess.

.. is it? :confused:

[zeus163]

@neuron2

Thanks for your response.

In order for me to send you a file that is small enough, I would have to cut it. If that is OK then I will do it. If not, I will cap a few minutes tonight when I am at home (if that is acceptable).

How big can the file size be?

Thanks for helping out. I appreciate all of your hard work as I know that I wouldn't be able to do it.

[cw_uk]

Originally posted by neuron2
Now you're confusing me.
Spat all what out? There wasn't a compatible DGParse released until RC3.

You know what, if you don't know what DGParse is for, then there is no reason for you to be using it. :D

Didnt realise i came off sounding like a complete n00b. i know exactly what DGParse is for. i just didnt know what to do with the latest version as it had no readme with it to say what had changed. i just happened to mention the incorect version i was using at the time. :(