DGMPGDec 1.5.7

[Guest]

* Fixed postprocessing pitch code in DGDecode that caused problems in Avisynth 2.6 (contributed by IanB).

* Added indexing status reporting to stdout (contributed by vlada).

http://neuron2.net/dgmpgdec/dgmpgdec157.zip

[Atak_Snajpera]

* Added indexing status reporting to stdout (contributed by vlada).
will you be able to do the same in dgavcdec?

[Guest]

Do you need the NV or non-NV version?

[Atak_Snajpera]

non-NV (based on libavcodec)

[Guest]

That one is withdrawn due to GPL issues. The latest libavcodec.dll also appears unusable for accurate random access and after the "Hall of Shame" episode, I'm disinclined to work on it anymore.

[stax76]

Thanks vlada, I have to check how other GUIs use DGIndex and if this new output can be useful for my GUI.

I'm not sure how to read the output, probably the same way than from a console application I guess.

[Atak_Snajpera]

That one is withdrawn due to GPL issues.
is this because you did not include stupid sources?

[Guest]

I modified libavcodec in a very minor way and released code for that. But a stringent interpretation of GPL (disputed by some) is that if the DLL is modified then the invoking app must be open.

I was on the verge of releasing all the source anyway when, instead of emailing me about the minor "violation", some idiot put me in a "Hall of Shame". Now my response is they can go pound sand, and I wash my hands of it all. I hope they are happy with what they have wrought.

My efforts will be restricted to the NV line, unless CoreAVC releases a suitable API, which they continue to say they will do. But we all know about CoreAVC promises.

[Atak_Snajpera]

Did you think about DiAVC?

[Guest]

There's no source code or published API, but let's take this to a separate thread, please, as we are getting OT.

[olex99]

Do you need the NV or non-NV version?

It would be great if you could get it into a future release of the NV version

[Brazil2]

That one is withdrawn due to GPL issues. The latest libavcodec.dll also appears unusable for accurate random access and after the "Hall of Shame" episode, I'm disinclined to work on it anymore.
Sad news :(


I was on the verge of releasing all the source anyway when, instead of emailing me about the minor "violation", some idiot put me in a "Hall of Shame". Now my response is they can go pound sand, and I wash my hands of it all. I hope they are happy with what they have wrought.
Well, the point is that by doing that you're not punishing the idiot in any way. The only ones to be punished are us, the users. And especially the ATI users.


Back to the topic: is there any reason why a MPEG PS file is not showing any bitrate information when previewed ? I've encoded the file with the goal of making a NTSC DVD so it's 23.976 FPS with a soft 3:2 pulldown. However I have some sample files with the same characteristics which are showing a bitrate so I wonder what is going wrong.

[Guest]

is there any reason why a MPEG PS file is not showing any bitrate information when previewed ? I've encoded the file with the goal of making a NTSC DVD so it's 23.976 FPS with a soft 3:2 pulldown. However I have some sample files with the same characteristics which are showing a bitrate so I wonder what is going wrong. No reason that I know of. Can you post a link to sample that does it?

[Brazil2]

Can you post a link to sample that does it?
Here is a short sample:
http://www.megaupload.com/?d=CZ4Y3XWU

Notice that I've already made the DVD several months ago and that it's playing fine in every player I've tried. I just recently noticed the bitrate issue when doing some comparison tests.
And also, if it can give a hint, when I demux the final PS file of the project I'm getting different file sizes (about 800 kB) whether I'm using DGindex or ProjectX using the default settings with no error/warning message shown.

[Guest]

Ah, "Baraka". Cool movie. But you should get the BluRay. :)

I see what the problem is. I'll update the 1.5.7 executable and repost in a few minutes to tell you when it's ready. Thanks for pointing this out.

The differing demux sizes is a different issue and is expected.

[Guest]

OK, redownload 1.5.7 and all will be well.

[Guest]

Well, the point is that by doing that you're not punishing the idiot in any way. The only ones to be punished are us, the users. And especially the ATI users. Here's a thought for you:

Install a cheapo Nvidia card as a secondary graphics card. My tools can use it and you'll still have the ATI card for your gaming or whatever. You can get a workable Nvidia card for $50 bucks. That way you'll get VC1 support and GPU deinterlacing/resizing as well.

[Brazil2]

The differing demux sizes is a different issue and is expected.
May I ask why ?
Are you suggesting that ProjectX is not to be trusted when it comes to demuxing ?


OK, redownload 1.5.7 and all will be well.
OK it's now working fine with the new version, thanks :)


Here's a thought for you:

Install a cheapo Nvidia card as a secondary graphics card. My tools can use it and you'll still have the ATI card for your gaming or whatever. You can get a workable Nvidia card for $50 bucks. That way you'll get VC1 support and GPU deinterlacing/resizing as well.
I'm not sure sure it's a good idea to have both brands at the same time on the same system. I have the feeling it may lead to more troubles than solutions. Not to mention that I'm not intensively doing video stuff, it's only a hobby so I think it's not worth it, sorry.
But here is a thought for you: just ignore the idiot and keep on doing the good work! :D

[Guest]

May I ask why ?
Are you suggesting that ProjectX is not to be trusted when it comes to demuxing? Not at all, just that different strategies are used. For example, should the demuxer discard any video data before the first sequence header, or should it assume that the first sequence header can be retroactively applied to the leading video data? Should detected errors lead to discarding data? How much data to discard? Etc.

But here is a thought for you: just ignore the idiot and keep on doing the good work! Just for the record, I'm not calling anyone an idiot, that is your characterization.

[Brazil2]

Not at all, just that different strategies are used. For example, should the demuxer discard any video data before the first sequence header, or should it assume that the first sequence header can be retroactively applied to the leading video data? Should detected errors lead to discarding data? How much data to discard? Etc.
Well, the video demuxed by ProjectX is actually a little bit bigger than the one demuxed by DGindex.
That's just for my information anyway since the final DVD is playing fine as I said.


Just for the record, I'm not calling anyone an idiot, that is your characterization.
Nope, it's yours:
some idiot put me in a "Hall of Shame".

[Guest]

Oh, you meant that idiot! Agreed. :)

I thought we were talking about the libavcodec developers, who are certainly not idiots.

[SilaSurfer]

Hey neuron2 me again, I've been having some weird issues with some encodings from DVD sources. So if you could please take look here

http://85.230.118.136/showthread.php?t=119500&page=12

because it is not allowed to ask same questions on many threads. I want to exclude Dgindex, or maybe that I'm doing something wrong with it. Thanks again

[Guest]

I don't have time to get involved in that unless you can demonstrate an issue with DGMPGDec.

[SilaSurfer]

Ok sorry for taking your time. Thanks

[alfixdvd]

With version 1.5.7 and 1.5.6 the option -preview -exit don't works well.

It loops over first 100 frames continuosly and don't exit.

Version 1.5.5 dos well the job.

[Guest]

Redownload 1.5.7 to get the fix. Thank you for pointing it out.

[alfixdvd]

Redownload 1.5.7 to get the fix. Thank you for pointing it out.

Quasi works well.

After 100 frames the Information Window closes, but the process Dgindex not ends, it is present in the Task list of windows, I have to terminate it manually :mad:.

Another surprise is that I have two log files.
I examine VTS_01_2.VOB, and I get VTS_01_2.log, but also get VTS_01.log with the same contents.

best regards

[Guest]

Please redownload and try again. I have tested it this time. :)

[alfixdvd]

Please redownload and try again. I have tested it this time. :)

All's ok. :):):)

Thank you.

[Monamona]

The below linked TS file contains an AAC audio stream.
However, DGIndex demuxs as MPA with 'File -> Save Project.'
With 'Preview', DGIndex correctly recognizes it as AAC.
I appreciate it if you take a look.

http://rapidshare.com/files/319530870/Sample.zip.html

[Guest]

Is that the beginning of the stream, or did you cut it?

[stax76]

Here is a VB.NET code snipped to use the new stdout feature:


Dim x = <x>-i "E:\Samples\DVD\Lost\VTS_01_1.VOB" -o "D:\Unsortiert\aaa\bbb" -hide -exit</x>
Dim p As New Process
AddHandler p.OutputDataReceived, Sub(e2, args) Invoke(Sub() Text = args.Data)
p.StartInfo.UseShellExecute = False
p.StartInfo.RedirectStandardOutput = True
p.StartInfo.FileName = "D:\Unsortiert\dgmpgdec157\DGIndex.exe"
p.StartInfo.Arguments = x.Value
p.Start()
p.BeginOutputReadLine()

[Guest]

The below linked TS file contains an AAC audio stream. Can you please post a longer version of the stream? That one does not have an AAC audio header in it and it only has an emulated MPEG audio header. I need a longer sample to resolve this.

[Guest]

Never mind about the longer stream. It's just simply getting stuck on an emulated MPEG audio header. I need to add the ability to force it either way as is done in DGDecNV. May take me a while. For now as a workaround, you can just let it go and rename it to AAC because the same demuxing is used for both.

[TinTime]

I've been experimenting with the new CLI syntax and run into a problem.

The old syntax...

"D:\Vtemp\programs\dgindex\DGIndex.exe" -IF=[E:\Video\test\test.m2v] -IA=5 -FO=0 -OM=0 -OF=[D:\Vtemp\demux\old] -AT=[] -hide -exit

...works fine, but this...

"D:\Vtemp\programs\dgindex\DGIndex.exe" -i "E:\Video\test\test.m2v" -ia 5 -fo 0 -om 0 -o "D:\Vtemp\demux\new" -at "" -hide -exit


...doesn't exit. It hangs after producing the d2v and I have to terminate the process. The resulting d2v is identical to the first method though.

Thanks for these changes by the way - comma support in filenames is very useful to me.

[Guest]

Please redownload 1.5.7 and try again. I silently fixed a problem like that some time ago.

[TinTime]

I spotted the similar problem (are you talking about this post? (http://forum.doom9.org/showpost.php?p=1348382&postcount=25)) and redownloaded / retested just before I posted.

[stax76]

@neuron2

The stdio status report feature works great, it would be very helpful to have this on all your index applications.

[Monamona]

Originally Posted by neuron2
Can you please post a longer version of the stream? That one does not have an AAC audio header in it and it only has an emulated MPEG audio header. I need a longer sample to resolve this.

Sorry for late response.
Please find attached.

http://rapidshare.com/files/322746444/Sample.zip.html

This is the beginning of the stream.

[Guest]

I've been experimenting with the new CLI syntax and run into a problem. Correct syntax for specifying no AVS template is:

-at

i.e., omit the "".

I'll update the help file.

[Monamona]

Originally Posted by neuron2
Never mind about the longer stream...
For now as a workaround, you can just let it go and rename it to AAC because the same demuxing is used for both.

Well, somehow I did not realize this post...

Anyway, thank you for your attention.
Actually, I renamed mp1 to aac, but it seems that it is not recognized as AAC by other programs.
The reason might be the audio header as you wrote.

Again, thank you for your continuous effort to your great DGIndex.

[TinTime]

Correct syntax for specifying no AVS template is:

-at

i.e., omit the "".

I'll update the help file.


That's done the trick - thanks very much for your help.

[BuHHunyx]

File: DGIndex.exe
Function: AttachConsole
Requirements: Client Requires Windows Vista or Windows XP.
DLL: Kernel32.dll

[Guest]

Please redownload 1.5.7 and try again. I added code so that the AttachConsole functionality is invoked only if the OS is XP or higher. So things should be OK for Win2000. Please advise if it is working. Thank you.

[BuHHunyx]

Please redownload 1.5.7 and try again. I added code so that the AttachConsole functionality is invoked only if the OS is XP or higher. So things should be OK for Win2000. Please advise if it is working. Thank you.

DGIndex.exe from 29.12.2009 12:00 - still _statically_ depends from AttachConsole.
Anyway - thanks for a quick reply!

[Guest]

So you get an error message? As I do not have Win2K, you'll have to give me more details about exactly what is happening and when.

[BuHHunyx]

So you get an error message? As I do not have Win2K, you'll have to give me more details about exactly what is happening and when.

When I try to launch DGIndex.exe I get following MB:

---------------------------
DGIndex.exe - Entry Point Not Found
---------------------------
The procedure entry point AttachConsole could not be located in the dynamic link library KERNEL32.dll.
---------------------------
OK
---------------------------

[Tima]

Seems like 'stdout reporting' patch breaks starting DGIndex from console applications, such as Far Manager.

DGI 1.5.7 attaches to a parent console, and interfers with it in non-desired way (in case of Far, it writes progress data to running session of Far).

Quoting one of Far developers, it can be fixed by merely changing 'AttachConsole' to 'AllocConsole' in DGI.

[Wilbert]

Would it be possible to put the source on your webpage?

[Guest]

Quoting one of Far developers, it can be fixed by merely changing 'AttachConsole' to 'AllocConsole' in DGI. The problem with that is that when you open DGIndex, it pops open another separate console window.