DVD2AVIdg Version 1.3.0 with transport inside! [final]

[Guest]

Tom, what an honor to see you here in the DVD2AVI forum.

Yes, your functionality is subsumed and soon will be surpassed with addition of support for PVA and other things I have in mind. But please wait for the release and posting of 1.3.0 on my web site before vectoring there. I'll PM you when that occurs, which shouldn't be long now.

For those that don't know, Tom is a Video and Coding God.

[Dallemon]

well after swapping the mpeg2dec3.dll to the dg version no problems at all ^_^ and i havent noticed anything out of the ordinary (yet) (using 130b6)

[DDogg]

D, good to keep the latest in the first post?, or is that one 'semi-stable' (B4) and the later posted version (B6) considered to be bleeding?

[Guest]

Originally posted by DDogg
D, good to keep the latest in the first post?, or is that one 'semi-stable' (B4) and the later posted version (B6) considered to be bleeding? Just forgot. Actually 6 is the first good version. :)

[Guest]

Originally posted by len0x
I'm looking at the d2v project (1.3.0) created from CLI and GUI and see that iDCT_Algorithm line differs on my PC. I always get 2 (SSEMMX) from CLI and 5 (SSE2MMX) from GUI. Is CPU detection not the same for both cases? Actually, look at the CommandLine.txt file. The CL was not allowing all 5 choices to be set. I've fixed it and it will be in the next release:

-iDCT-Algorithm -IA (1=MMX 2=SSEMMX 3=SSE2MMX 4=Floating point 5=Reference)

[Guest]

Wilbert:
There's one problem. If I open the avs script in wmp6.4, I see the following (it's the second video stream):

http://www.geocities.com/wilbertdijkhof/test.jpg

But it loads fine in vdubmod. So, I guess it's some colorspace conversion error? Strangely, the first stream shows up fine. If needed I can upload the clip somewhere. Yes, please. Make the clip available ASAP because I want to release 1.3.0.

[Guest]

I've reached feature cutoff for 1.3.0 so here is a release candidate:

http://neuron2.net/fixd2v/decodefix130RC1.zip

This versions changes as follows versus beta 6:

1. The bitrate monitor is now numerical (bps).

2. Added start code emulation protection to the transport parser. This makes it more robust.

3. The Info dialog now pops up over the main display instead of beside it. This means there is no need to move it for HD pictures.

4. Fixed the command language -IA option to accept all five supported IDCT algorithms.

5. Added a timeout feature to the transport parser such that if a play, preview, or save project is started with no valid video PID, it will no longer hang DVD2AVIdg.

It would be nice to get this puppy released. :D

[zettai]

Originally posted by neuron2
3. The Info dialog now pops up over the main display instead of beside it. This means there is no need to move it for HD pictures.

Is there any way to make this a little more flexible?

I like to be able to see the footage when I preview with F5 and with this I have to move the info window every time.

Can you set it as an option or maybe only happen for video over a certian frame size or maybe detect being off screen? I dunno...

[Guest]

@zettai

Done. It is now based on the picture size. Thanks for the suggestion.

[DDogg]

DG, I am trying to help dvd2svcd chase a problem. Do you see anything below in this very small d2v that would cause a 100 CPU 'loop' when attempting to open in WMP or VDub? The VOB is only 2 megs created by the preview function of D2S. I have loaded the original D2v as a project in decodefix130RC1, and resaved the d2v below in your new format. I can mail the vob to you if you pm instructions.


LoadPlugin("C:\Program Files\DVD2SVCD\TEMP\MPEG2Dec3dg.dll") #decodefix130RC1 version
MPEG2Source("C:\PROGRA~1\DVD2SVCD\TEMP\preview3.d2v") #created by decodefix130RC1


DVD2AVIProjectFile02
1
42 C:\Program Files\DVD2SVCD\TEMP\preview.vob

Stream_Type=0,0,0
iDCT_Algorithm=2 (1:MMX 2:SSEMMX 3:FPU 4:REF 5:SSE2MMX)
YUVRGB_Scale=1 (0:TVScale 1:PCScale)
Luminance_Filter=0,0 (Gamma, Offset)
Picture_Size=0,0,0,0,0,0 (ClipLeft, ClipRight, ClipTop, ClipBottom)
Aspect_Ratio=16:9
Field_Operation=0 (0:None 1:ForcedFILM 2:SwapOrder 3:RawFrames)
Frame_Rate=29970
Location=0,0,0,3D8

7 0 4295073471 0 0 2 3 10 11 12 13 10 11 12 13 10 11
7 0 4295228094 0 0 2 3 10 11 12 13 10 11 12 12 12 12 12
7 0 4295424968 0 0 12 13 10 11 12 13 10 11 12 13 10 11
7 0 4295608690 0 0 2 3 10 11 12 13 10 11 12 13 10 11
7 0 4295772302 0 0 2 3 10 11 12 13 10 11 12 13 10 11
7 0 4295878972 0 0 2 3 10 11 12 13 10 11 12 13 10 11
7 0 4295987963 0 0 2 3 10 11 12 13 10 11 12 13 10 11
7 0 4296171439 0 0 2 3 10 11 12 13 10 11 12 13 10 11
7 0 4296396571 0 0 2 3 10 11 12 13 10 11 12 13 10 11
7 0 4296618418 0 0 2 3 10 11 12 13 10 11 12 13 10 11
7 0 4296847831 0 0 2 3 10 11 12 13 12 9

FINISHED 95.28% FILM

[Guest]

Sorry, RC1 was hosed. Try RC2:

http://neuron2.net/fixd2v/decodefix130RC2.zip

These stupid __int64 variables just have a mind of their own. They don't behave like normal ints and you have to really watch out.

This also fixes zettai's info popup thing.

Please give me your results tonight, if possible, so I can go to bed.

[DDogg]

RC2

1> Will open preview.vob, save and play via standard script, but throws red on black access violation at very end.

No other problems. That fixed the cpu loop. Thanks very much. You are to coding Gods what Walter is to grumpy. Superior in all ways :)

[Guest]

Violation at end can be due to bad cut leaving incomplete final picture. Trim it away in your script. Please me the VOB fragment with the end intact. That will help me fix the crash at the end issues.

[DDogg]

Please me the VOB fragment with the end intact. And how do I please you? :) Gimme the address and the header codeword?

Yep, frame 124 was the bad seed. trim (0,123) played without error.

[Guest]

I've added some crash-at-the-end protection to DVD2AVIdg but haven't gotten to MPEG2DEC3dg to do the same yet. I figure Trim() will get people by for now as long as DVD2AVIdg can save the project OK.

> Gimme the address and the header codeword?

Huh? I just need your end intact.

[Guest]

I think I'll write my own long long int class that I can trust!

[DDogg]

Huh? I just need your end intact. Oh, me too, but how do I deliver you the intact end of a vob that is 2megs? Perhaps you mean the end of the d2v I guess. Here is the complete d2v. Well hell, it does not throw the error now, but last frame is blocked up.


DVD2AVIProjectFile02
1
42 C:\Program Files\DVD2SVCD\TEMP\preview.vob

Stream_Type=0,0,0
iDCT_Algorithm=2 (1:MMX 2:SSEMMX 3:FPU 4:REF 5:SSE2MMX)
YUVRGB_Scale=1 (0:TVScale 1:PCScale)
Luminance_Filter=0,0 (Gamma, Offset)
Picture_Size=0,0,0,0,0,0 (ClipLeft, ClipRight, ClipTop, ClipBottom)
Aspect_Ratio=16:9
Field_Operation=1
Frame_Rate=23976
Location=0,0,0,3D8

7 0 28756 0 0 2 3 10 11 12 13 10 11 12 13 10 11
7 0 325483 0 0 2 3 10 11 12 13 10 11 12 13 10 11
7 0 549845 0 0 2 3 10 11 12 13 10 11 12 13 10 11
7 0 748319 0 0 2 3 10 11 12 13 10 11 12 13 10 11
7 0 948925 0 0 2 3 10 11 12 13 10 11 12 13 10 11
7 0 1140695 0 0 2 3 10 11 12 13 10 11 12 13 10 11
7 0 1326749 0 0 2 3 10 11 12 13 10 11 12 13 10 11
7 0 1513844 0 0 2 3 10 11 12 13 10 11 12 13 10 11
7 0 1718927 0 0 2 3 10 11 12 13 10 11 12 13 10 11
7 0 1917102 0 0 2 3 10 13 9

FINISHED 99.10% FILM

/Add: Ok, this one throws the violation at the end. I'll keep the matching vob fragment, if you need it let me know.

DVD2AVIProjectFile02
1
42 C:\Program Files\DVD2SVCD\TEMP\preview.vob

Stream_Type=0,0,0
iDCT_Algorithm=2 (1:MMX 2:SSEMMX 3:FPU 4:REF 5:SSE2MMX)
YUVRGB_Scale=1 (0:TVScale 1:PCScale)
Luminance_Filter=0,0 (Gamma, Offset)
Picture_Size=0,0,0,0,0,0 (ClipLeft, ClipRight, ClipTop, ClipBottom)
Aspect_Ratio=16:9
Field_Operation=1
Frame_Rate=23976
Location=0,0,0,3D8

7 0 106175 0 0 2 3 10 11 12 13 10 11 12 13 10 11
7 0 260798 0 0 2 3 10 11 12 13 10 11 12 12 12 12 12
7 0 457672 0 0 12 13 10 11 12 13 10 11 12 13 10 11
7 0 641394 0 0 2 3 10 11 12 13 10 11 12 13 10 11
7 0 805006 0 0 2 3 10 11 12 13 10 11 12 13 10 11
7 0 911676 0 0 2 3 10 11 12 13 10 11 12 13 10 11
7 0 1020667 0 0 2 3 10 11 12 13 10 11 12 13 10 11
7 0 1204143 0 0 2 3 10 11 12 13 10 11 12 13 10 11
7 0 1429275 0 0 2 3 10 11 12 13 10 11 12 13 10 11
7 0 1651122 0 0 2 3 10 11 12 13 10 11 12 13 10 11
7 0 1880535 0 0 2 3 10 11 12 13 12 9

FINISHED 95.28% FILM


I gotta go to bed, night.

[Zep]

Originally posted by neuron2
I've reached feature cutoff for 1.3.0 so here is a release candidate:

http://neuron2.net/fixd2v/decodefix130RC1.zip



5. Added a timeout feature to the transport parser such that if a play, preview, or save project is started with no valid video PID, it will no longer hang DVD2AVIdg.


YEAH #5 got me a few times

oh incase you missed my update edit above here it is again


UPDATE: tried nandub and VDM same test. Same speeds.
So whatever it is, it is on avisynth side and its ability to
frame serve.

I'm using version

AVS 2.5.5 alpha 2004-03-19 15:00
AviSynth_200304.exe

[Zep]

Originally posted by neuron2
Violation at end can be due to bad cut leaving incomplete final picture. Trim it away in your script. Please me the VOB fragment with the end intact. That will help me fix the crash at the end issues.

MPEG2DEC3dg side.

i saw this crash last night on both frist and second pass
when the LAST frame was read it.


VDub just closed with an abort. it was a .ts so it makes me
wonder what was wrong with the very last frame.

[Zep]

Originally posted by neuron2
I've added some crash-at-the-end protection to DVD2AVIdg but haven't gotten to MPEG2DEC3dg to do the same yet. I figure Trim() will get people by for now as long as DVD2AVIdg can save the project OK.

> Gimme the address and the header codeword?

Huh? I just need your end intact.



yeah MPEG2DEC3dg is where i am seeing the problem
and thus a crash in VDub on last frame on each pass.


I will try a trim next chance I get

[Wilbert]

Yes, please. Make the clip available ASAP because I want to release 1.3.0.

http://www.geocities.com/wilbertdijkhof/FaithHill10.ts

a bit more info:

http://forum.doom9.org/showthread.php?s=&postid=491183#post491183

[zettai]

Strange little bug:

If you try and make a d2v file on a hard drive which is *almost* full and there isnt enough space for a d2v, instead of throwing an error the file will be partially written instead!

[len0x]

Originally posted by neuron2
Actually, look at the CommandLine.txt file. The CL was not allowing all 5 choices to be set. I've fixed it and it will be in the next release:

-iDCT-Algorithm -IA (1=MMX 2=SSEMMX 3=SSE2MMX 4=Floating point 5=Reference)

works great now, thanks!
why not add this to the GUI menu as well ?

[Guest]

Originally posted by DDogg
Oh, me too, but how do I deliver you the intact end of a vob that is 2megs? Perhaps you mean the end of the d2v I guess. Here is the complete d2v. If it's only 2Meg, please give me a link to the whole thing! I need to be able to duplicate the crashes to fix them.

[Guest]

Originally posted by len0x
works great now, thanks!
why not add this to the GUI menu as well ? The IDCT menu presents only the options that your processor is capable of supporting.

[len0x]

Originally posted by neuron2
The IDCT menu presents only the options that your processor is capable of supporting.

you mean maximum option? if SSE2 is supported then surely SEE and MMX are supported as well... (but you only see SSE2)

[Guest]

Originally posted by len0x
you mean maximum option? if SSE2 is supported then surely SEE and MMX are supported as well... (but you only see SSE2) See, I haven't an SSE2 so I never could notice that! I'll fix that.

[Guest]

Originally posted by Wilbert
http://www.geocities.com/wilbertdijkhof/FaithHill10.ts

a bit more info:

http://forum.doom9.org/showthread.php?s=&postid=491183#post491183 Wilbert, everything works fine for me with Media Player 6.4 and both streams using RC2. I don't know where to go with this.

[Wilbert]

I followed the advice of Moitah, changing my display to 16 bit works fine. Strange that the main video stream didn't have this problem.

[Guest]

Originally posted by Wilbert
I followed the advice of Moitah, changing my display to 16 bit works fine. Strange that the main video stream didn't have this problem. Maybe you have a bad video card.

[DDogg]

If it's only 2Meg, please give me a link to the whole thing! OK, check your PM

[lineman]

Nice work on Dvd2avidg "neuron2" , I just have a couple of questions.
Will it support DVB-S 4.2.0.transport streams, and also is there a way to have it support DVB-S 4.2.2. mpg transport streams. Nics dvd2avi will work on the 4.2.2. streams but when opened in virtualdub they lack color, and are mostly green. I use VFAPI reader codec to convert the d2v to an false avi and then the image is correct in vitualdub. I will post some sample when I find an suitable site. Thanks for all you hours of hard work...

Lineman

[Moitah]

A friend sent me a small .pva for testing (recorded with ProgDVB), it isn't decoded correctly with RC2 (weird colored and misplaced blocks). It works fine if I use pva2mpg2.exe. It's up on my FTP server.

[Guest]

@lineman

I don't know. Can you make available some test clips.

@Moitah

PVA is not expected to work. I have to add support for it. I have the spec and a sample from CruNcher. I'll get yours too. Maybe I'll start on it tomorrow.

[Guest]

Originally posted by DDogg
OK, check your PM It doesn't crash for me with RC2. What say you?

[DDogg]

Same one I sent you crashes on last frame as mentioned. That really doesn't concern me as this preview.vob is always used with a trim statement to create stills. Since it does not crash for you, it is not a replicatable. That makes it not worth your time, IMO.

[Guest]

I got bored today and decided to write a PID detector. It's a command line utility with the following usage:

>table filename

Obviously, use it on transport streams that carry PAT/PMT.

The zip file includes source code, from which you can see that PID detection is no trivial matter. This will be integrated into DVD2AVIdg.

http://neuron2.net/fixd2v/table101.zip

Feedback will be appreciated.

Here is sample output:

G:\tmp\Zep>table abc.3.video.3.audio.streams.tp
------------------------------------------------
Table 1.0.1 -- PAT/PMT Parser by Donald A. Graft
------------------------------------------------

Program 1
PCR on PID 0x11
Video on PID 0x11
AC3 Audio on PID 0x14
Other on PID 0x1a
Program 2
PCR on PID 0x21
Video on PID 0x21
AC3 Audio on PID 0x24
Program 3
PCR on PID 0x31
Video on PID 0x31
AC3 Audio on PID 0x34

[Guest]

Here is version 1.3.0 RC3:

http://neuron2.net/fixd2v/decodefix130RC3.zip

This version adds integrated PID detection, cleans up the IDCT algorithm and CPU detection menu support, and removes the resize and LB menu items (they were not used by MPEG2DEC3dg).

OK, now we have really reached feature cut-off. :)

[dvd2svcd]

Hi Donald,

For some reason I get an Avisynth Error when trying to use this D2V file:DVD2AVIProjectFile03
1
19 C:\test\preview.vob

Stream_Type=0,0,0
iDCT_Algorithm=2 (1:MMX 2:SSEMMX 3:FPU 4:REF 5:SSE2MMX)
YUVRGB_Scale=1 (0:TVScale 1:PCScale)
Luminance_Filter=0,0 (Gamma, Offset)
Picture_Size=0,0,0,0 (ClipLeft, ClipRight, ClipTop, ClipBottom)
Aspect_Ratio=16:9
Field_Operation=0 (0:None 1:ForcedFILM 2:SwapOrder 3:RawFrames)
Frame_Rate=29970
Location=0,0,0,3DA

7 0 156092 0 0 13 10 11 12 13
7 0 344656 0 0 0 1 12 13 10 11 12 13 10 11 12 13
7 0 601580 0 0 0 1 12 13 10 11 12 13 10 11 12 13
7 0 863760 0 0 0 1 12 13 10 11 12 13 10 11 12 13
7 0 1117468 0 0 0 1 12 13 10 11 12 13 10 11 12 13
7 0 1378036 0 0 0 1 12 13 10 11 12 13 10 11 12 13 10
7 0 1651732 0 0 11 12 13 10 11 12 13 10 11 12 13
7 0 1938552 0 0 12 9

FINISHED 98.70% FILM
My Avisynth file is very simpleLoadPlugin("C:\PROGRAMMER\DVD2SVCD\DVD2AVidg\MPEG2dec3dg.DLL")
MPEG2Source("C:\TEST\preview.d2v")My dvd2avi.ini looks like thisDVD2AVI_Version=DVD2AVIdg 1.3.0 RC3
Window_Position=100,100
iDCT_Algorithm=2
YUVRGB_Scale=1
Field_Operation=0
Output_Method=3
Track_Number=0
DR_Control=2
DS_Downmix=0
SRC_Precision=0
Norm_Ratio=100
Process_Priority=2
Transport_PIDs=11,14
If I use ParseD2V the end lines is65: 81,82 -- 0 [vob/cell=0/0]
[GOP START]
66: 82,83,83 -- 1 [vob/cell=0/0]
67: 84,84 -- 2 [vob/cell=0/0]
68: 85,85,86 -- 3 [vob/cell=0/0]
69: 86,87 -- 0 [vob/cell=0/0]
70: 87,88,88 -- 1 [vob/cell=0/0]
71: 89,89 -- 2 [vob/cell=0/0]
72: 90,90,91 -- 3 [vob/cell=0/0]
73: 91,92 -- 0 [vob/cell=0/0]
74: 92,93,93 -- 1 [vob/cell=0/0]
75: 94,94 -- 2 [vob/cell=0/0]
76: 95,95,96 -- 3 [vob/cell=0/0]
[GOP START]
77: 96,97 -- 2 [vob/cell=0/0] [3:2 pattern break] [ILLEGAL TRANSITION!]
[EOF]
I'm using Avisynth 2.54. I have reproduced this error on 2 computers, both Intel one is P3 and the other a P4. I'll make a PM to you where you can download my test data.
Any other info you need just let me know.

[len0x]

I also have issues with RC3. RC2 worked fine but with RC3 I get "MPEG2Source: couldn't open file".

[Guest]

Originally posted by len0x
I also have issues with RC3. RC2 worked fine but with RC3 I get "MPEG2Source: couldn't open file". It has a new D2V version format; you have to remake your project with the new DVD2AVIdg.

[len0x]

Originally posted by neuron2
It has a new D2V version format; you have to remake your project with the new DVD2AVIdg.

of course I did - now that I'm looking at the dates of DVD2AVI and mpeg2dec3 I see that RC3 archive includes mpeg2dec3dg.dll from RC2!

[Guest]

@Dvd2scvd

As a workaround, you can trim away the last picture in your script. But I agree it would be better to not crash.

[Guest]

Originally posted by len0x
of course I did - now that I'm looking at the dates of DVD2AVI and mpeg2dec3 I see that RC3 archive includes mpeg2dec3dg.dll from RC2! Oh, duh!!! I'll replace it immediately.

EDIT: The archive has been corrected. Please advise if it is OK now for you. The perils of working at 4:00 in the morning. The caffeine level is dangerously low.

[Guest]

Originally posted by dvd2svcd
[GOP START]
77: 96,97 -- 2 [vob/cell=0/0] [3:2 pattern break] [ILLEGAL TRANSITION!]
[EOF] What happens if you use Fixd2v on the D2V and then serve the fixed D2V?

[len0x]

Originally posted by neuron2
EDIT: The archive has been corrected. Please advise if it is OK now for you.


works like a charm now :) thanks!
Another question about archive - some txt/html files there have read only attribute on them. Is that intentional?

Originally posted by neuron2

The perils of working at 4:00 in the morning. The caffeine level is dangerously low.

Wow! I can never program at this time :) Considering that its weekend you'll probably be sleeping all Sunday long :D

[dvd2svcd]

Originally posted by neuron2
What happens if you use Fixd2v on the D2V and then serve the fixed D2V? Seems to be the same problem. I rename preview.d2v.fixed to preview.d2v and tried again, same VirtualDUB error: http://www.dvd2dvd.org/vduberror.jpg

EDIT: Sorry, didn't see the above posts. I'll try with the new mpeg2dec3dg.dll.

[Guest]

Originally posted by len0x
works like a charm now :) thanks!
Another question about archive - some txt/html files there have read only attribute on them. Is that intentional? Nah. They probably spent some time on a CD-ROM before being copied to their current location. If they were write only, that would be a problem. :)

[Guest]

Originally posted by dvd2svcd
EDIT: Sorry, didn't see the above posts. I'll try with the new mpeg2dec3dg.dll. Oh, goodie, less work for me. See you just said Avisynth error, so I assumed you meant a crash.

[dvd2svcd]

Yes, that did it. All works fine again :) Thanks for the speedy fix :)


Originally posted by neuron2
Oh, goodie, less work for me. See you just said Avisynth error, so I assumed you meant a crash. I know, my bad, sorry. I meant to have posted that VirtualDUB error JPG in my first post, but I forgot it seems.