Burning DL media with PgcEdit

[jinjin_jp]

@jinjin_jp
Read the link in my post above Building and Burning Dual-Layer DVD (http://www.emedialive.com/Articles/PrintArticle.aspx?ArticleID=8421)
Dual-layer DVD-R is fully compliant to DVD-9!
Is it mean not necessary to set L0=2,092,896sectors for -R/DL?

Your example L0=L1 is standard compliant, but L0=L1 is not a MUST! There are no new problems! :rolleyes:
Please understand I don't mean L0=L1 is MUST. Almost which I have seen were L0>L1. I want to report this example(L0=L1) as rare one.

[frank]

Is it mean not necessary to set L0=2,092,896sectors for -R/DL?No, that is the maximum on L0 of this media. As you can see in your second example, when the user data < 8.5 GB then the L0 sectors are decreased.

But it is much simpler.
Use PgcEdit, and look at the shown values before and after burning!

[blutach]

If you want to try the last beta, here is beta 16.A beta a day keeps the doctor away.

Regards

[jinjin_jp]

No, that is the maximum on L0 of this media. As you can see in your second example, when the user data < 8.5 GB then the L0 sectors are decreased.

But it is much simpler.
Use PgcEdit, and look at the shown values before and after burning!
Sorry, I can't understand so well.
In my first, second, and third example, I think all are L0=2,092,896sectors, and L0 sectors aren't decreased in second example.
I think this L0(shown in DVDDecryoter) doesn't mean the maximum on L0, and means to where burned on L0, so this value differs each DVD.
I think what is decreased is sectors from VIDEO_TS.IFO.

[add]the Image I have is like below.
http://img251.echo.cx/img251/4929/dltestresults21ba.jpg (http://www.imageshack.us)

[r0lZ]

Well, I really don't understand why this value of 2,092,896 sectors should have something magic. Seems it's aproximatively the size of a -R layer. But I don't think it's a good idea to push the data to the outer edge of the disc, as this part is more often subject to errors.
Also, I agree with frank. All informations available on the net are the same: it is not needed to fill L0 completely.
If it is really needed to do strange things like that to burn a DVD-R DL, I think I will not support -Rs anymore! I don't want to encourage people to use the -R technology while the +R is available, and is obviously superior.
Of course, I may change my mind later, when -R DLs will be available here to test both methods.

Anyway, jinjin_jp, thanks for your investigations, and for your translations in english!

[nwg]

From what I have read. DVD-R DL is less supported than DVD+R DL (Verbatim in the test). It was in the review of my Pioneer 109 writer. IT was found that not mnay players or even drives could read the discs. Even, my LG 4040 which doesn't support reading and writing of a DL+R or DL-R can read a DVD ROM DL+R.

Even though my Pioneer 109 supports -R DL. I won't be using it. I went from DVD+ to DVD- with single layer discs but will stick with DVD+R DL.

[Tobii]

short question at the beta 16...

In the Burn DVD dialog -> Cell list window (setting the LB), the preview doesn't work for a non-seamless cell any more.
With me, the preview works only at the seamless cells.
Changes the non-seamless cell into a seamless cell, the preview doesn't work either.

Something was changed here or does anybody have an idea?

[r0lZ]

No, Tobi. Nothing has changed since beta 15, except the default number of sectors in DVD-R DL. Obviously, it's not the problem.
Have you tried to launch the preview on the same cell from the PGC Editor's cell list? Do you have the same problem with another DVD?

[Tobii]

I have tried four different DVDs now. I have the problem at everybody.

Under Preview Menu -> Automatically start Preview Playback...is switched on.
The preview for the cell in question, doesn't work in the PGC Editor -> cell list.

The preview works for any cell, which has a seamless flag(8).
I start the preview for cell 19 (cell type flag=2 and LB), preview window opens -> no preview playback here and the functions above are inactive. (look at the snapshot)

http://img76.echo.cx/img76/8053/previewfenster2hm.th.png (http://img76.echo.cx/my.php?image=previewfenster2hm.png)

In the Burn DVD dialog, it is just the same. Preview window opens -> no preview playback here and the functions above are inactive.

I go backward tomorrow ... beta 15 -> beta14->beta13 etc.. and then wanting to test.

[r0lZ]

Hum, strange.

I've just found a similar problem with Matrix (Z2 Fr), but an error message appear: "Couldn't locate sector 1343176 !".
But note that my original rip of Matrix has some realloc problems with Nero, even if I remove the empty VOBs.

Note that the problem CANNOT come from a modification in one of the 0.6.0 betas. The PgcEditPreview.exe has not been modified, nor the method to launch it, at least from the PGC Editor's cells list.
Maybe it's something you have installed recently that cause the problem?

[r0lZ]

Well, I have mock stripped Matrix. Same problem.
Seems it is exactly the same problem as yours (the error message appear only when running PgcEdit in debug mode.)
I have emailed jeanl to let him know. It's probably a bug in PgcEditPreview.exe.

[r0lZ]

The preview works for any cell, which has a seamless flag(8).
I start the preview for cell 19 (cell type flag=2 and LB), preview window opens -> no preview playback here and the functions above are inactive.Matrix has a STC Discontinuity on the layer break's cell. When there is no STC discontinuity in the VOB, seems there is no problem with the Preview. So, the STC discontinuity could be the problem. Tobi, can you test if this flag is present in the DVDs you have tested so far? And try to find one without this flag and report if it works?
(Note that removing the STC falg in the IFO is not a solution: the Preview exe don't use the IFOs at all.)

[frank]

1) I've tested a backup with 6 titles.
In every title set the first cell has STC discontinuity but no seamless flag.
No problem with preview!

May be PgcEditPreview.exe has an issue.
If Jeanl used DVD2AVI as source then I remember there is a GOP bug that produces frame lost. It's better to change to DGIndex. Neuron2 can help when some CLI changes needed. More in the DGIndex f.a.q.

2) PgcEdit blocks the Burn DVD Menu when you test a DVD from drive and not from HD backup.
Usually comes the message that it can't create the backup folder, ok.

When you start the Burn DVD / Create ISO menu you'll get an application error:

http://img299.imageshack.us/img299/1781/err7om.jpg ([URL=http://www.imageshack.us)

Tested DVD (PCgo 4/2005) also has some DVD-ROM content. Strange.

[r0lZ]

1) I've tested a backup with 6 titles.
In every title set the first cell has STC discontinuity but no seamless flag.
No problem with preview!

May be PgcEditPreview.exe has an issue.
If Jeanl used DVD2AVI as source then I remember there is a GOP bug that produces frame lost. It's better to change to DGIndex. Neuron2 can help when some CLI changes needed. More in the DGIndex f.a.q.OK, but the problem may be caused by the STC discontinuity inside the VOB, and not when the first cell of the VOB is played. I'm not sure, though.

2) PgcEdit blocks the Burn DVD Menu when you test a DVD from drive and not from HD backup.
Usually comes the message that it can't create the backup folder, ok.

When you start the Burn DVD / Create ISO menu you'll get an application error:

http://img299.imageshack.us/img299/1781/err7om.jpg ([URL=http://www.imageshack.us)

Tested DVD (PCgo 4/2005) also has some DVD-ROM content. Strange.OK, the error message comes from a bug in the routine that calculates the size of DVD-ROM subdirs, with empty directories, and is not related to the fact that you opened the DVD from your drive. Fixed now.

What do you mean exactly with "PgcEdit blocks the Burn DVD Menu"? Is it the same problem as the error message? I have just tried to launch the Burn function from a DVD drive, and it worked, although slowly. The Layer Break selection dialog opened successfully.
Anyway, you cannot burn directly from a read-only filesystem, since PgcEdit needs to write the new VTS sector pointers in the IFOs before launching mkisofs. So, if you accept the dialogs, you will have an error message, and the burn process will stop.

[frank]

It was the same problem. The File -> Burn DVD / Create ISO windows didn't start, only the error message was shown.
The DVD with "Message in a Bottle" comes from a computer magazine and has much DVD-ROM directories. I never had this error. Does the DVD-ROM content disturb the routine dvd_rom_size_subdir? I'm waiting for your update, or post the corrected line.

I thought Tobii tested his 4 DVDs from drive, because there may be strange behaviour like that problem. I will test other DVDs with STC in VOB at home.

Yes, I know that we cannot burn from a read-only system.
But for searching of original LB settings it is very useful to read from drive. :)

[Tobii]

I have changed nothing in my system. I use the original files, without any changes. I just don't try out of a disk drive.

The STC flag, is in the VOB at the DVDs in question.
Unfortunately, I haven't found any DVD without STC flag yet.

Sorry, r0lZ...
I haven't tested matrix. Till now, I also haven't noticed the problem.


Waiting what jeanl says to it.

[jeanl]

mmm I don't have much to say unfortunately! I can't replicate that problem, I would need a short test DVD to do that, then I'll figure out what the problem is. I'm not sure about the STC discontinuity, as far as I can tell, DVD2AVI does not seem to use the STC, but maybe I'm wrong (I haven't looked that hard). Tobii or r0lZ, could you possibly prepare a small version of the DVD and send it to me via www.yousendit.com? Make it as small as possible, just to let me see the problem.
About using DGIndex instead of DVD2AVI, it would have been a good idea when I started, but unfortunately, I didn't do it. Now I've added enough stuff to DVD2AVI that it would be quite a bit of work to replicate all these changes in DGIndex. Besides, I'm not sure how neuron2 would feel about having someone adding stuff to the code he now maintains....

jeanl

[r0lZ]

I'm waiting for your update, or post the corrected line.In both dvd_rom_size and dvd_rom_size_subdir, you have to change the first "return 0" to "return [list 0 0]":
proc dvd_rom_size {rootdir} {
if {[catch {set allfiles [glob -tails -directory $rootdir *]}]} {
return [list 0 0] ;# BUG fixed; was: return 0
}
...Let me know if it works...

[r0lZ]

Tobii or r0lZ, could you possibly prepare a small version of the DVD and send it to me via www.yousendit.com? Make it as small as possible, just to let me see the problem.Unfortunately, if I keep only the problematic cell with VobBlanker, the problem dissapear. Will try another method, but I'm not sure I will be able to give you a good example.

[frank]

@rOlZ
Bug fixed. Compiled and works! :D


What Tcl editor you are using?

[Tobii]

@ jeanl

Difficult, I can solve out the cell with VobBlanker or PgcDemux. The size of the cell is 36 MB (Nevertheless sending?)
But, I am of the same opinion like r0lZ. No good example. I search after another solution.

BTW, if the preview is going on in the trace, the problematic cell is skipped directly.
I start the preview alone for the cell 19, the preview hangs and I see a picture from the cell 24. :confused:

[jeanl]

Tobii, can you tell me which DVDs this happens with, maybe I have one of them. Which region are you in (I'm R1)? Also, you can sent up to 1GB with yousendit.com.
Try to find a DVD where this happens for a cell that in the first file of the VTS, for example, VTS_01_1.VOB or VTS_02_1.VOB etc. THen cut anything after the cell that causes the problem with cutfile (one of jsoto's tools here (http://jsoto.posunplugged.com/others.htm)). Don't modify the IFOs, make sure the problem occurs and send me all the IFOs, and that vob using yousendit.com (use my address: jeanldvd at free dot fr). I should then be able to reproduce the problem.
jeanl

[r0lZ]

Also interesting to note: the Layer Break cell on Matrix is just at the beginning of a VOB file. Is it the same for your DVDs, Tobi?

[Tobii]

The DVD is Constantine(R2).
I am not sure whether the cell is at the beginning of a VOB. I noticed the same problem also at Blade Trinity (R2).
I will send the files in question (tonight).

[r0lZ]

You can check if the cell is at the beginning of the VOB with VobEdit.
If the first pack is a nav pack, look at the value of "this->lba" at offset 002d or 040b.
If this LBA number is the Entry VOBU Sector of the cell, then the VOB begins with your cell.

[frank]

Loading the VOB into DVD2AVI is much easier. ;)

TITANIC (PAL RC2)
Layer Break at the beginning of VTS_01_5.VOB: VOB/CELL ID 5/1.
STC discontinuity, timer restarts with 0.

IfoEdit cell preview freezes, showing a grey window!
PgcEdit freezes (PGC Editor preview), showing a still from 0:19:56 in VOB/ID 5/3. If you change the cell type flags the behaviour doesn't change.
A full copy (CloneDVD) without any stripping sorted the VOBs new (there were some < 1 GB). Now the LB is not at beginning of VOB - and all things work perfectly!

The Burn Dual Layer window of PgcEdit works well in both cases, just the preview has the bug.

There must be something wrong in preview for ages. Bug in nav pack analyzing stuff...
Strange.

[frank]

@rOlZ
I recommend to rename Burn Dual Layer into Burn Double Layer (= DL)!
Dual Layer is the reserved -R name of DVD Forum.
And we like +R! :D :D

[r0lZ]

OK! Sure! I'll do it asap!

[r0lZ]

Seems I have understood the problem with the LB cell and the preview. There is a wrong calculation of the relative LBA offset in the VOB when a cell starts at the beginning of a VOB.
So, when a cell starts at sector 0 of a VOB, the sector is still assumed to be at the end of the previous cell, and the relative LBA number is wrong: it should be 0, but is computed from the start sector of the previous VOB.

Tobi, can you confirm that the bugged cells are all at the beginning of a VOB?

[frank]

Yeah, confirmed!! You are right rOlZ!

In TITANIC VTS_01_4.VOB has 884016 KB = 84.3 % of 1048576 kB.
That distance added to the begin of VTS_01_5.VOB gives the shown still (~10s difference because of VBR).

The preview doesn't notice that the VOB before LB can be < 1 GB!!
Then the end is crossed and playing stops.

[jinjin_jp]

These two have the same problem.
(1)Matrix(RC2,NTSC)
.....VTS_02_1.VOB 1,048,574KB
.....VTS_02_2.VOB 1,048,574KB
.....VTS_02_3.VOB 910,276KB
.....VTS_02_4.VOB 1,048,574KB===>FirstCell(V/C-ID=16/1) is LayerBreakCell
.....VTS_02_5.VOB 1,048,574KB
.....VTS_02_6.VOB 956,676KB
(2)Animatrix(RC2,NTSC)
.....VTS_01_1.VOB 1,048,574KB
.....VTS_01_2.VOB 1,048,574KB
.....VTS_01_3.VOB 1,048,574KB
.....VTS_01_4.VOB 633,396KB
.....VTS_01_5.VOB 1,048,574KB===>FirstCell(V/C-ID=15/1) is LayerBreakCell
.....VTS_01_6.VOB 27,822KB

Only above two cell "play" button of preview is inactive(gray out).
In example(2) there are many cells of type 2, but only this is inactive, but others are not.

I think these example have characteristic which previous VOB file is smaller than about 1GB inspite of last VOB of VTS_?? .

[jinjin_jp]

I found one more example.
(3)Ocean's 12(RC2,NTSC)
.....VTS_01_1.VOB 1,048,574KB
.....VTS_01_2.VOB 1,048,574KB
.....VTS_01_3.VOB 1,048,574KB
.....VTS_01_4.VOB 61,798KB
.....VTS_01_5.VOB 1,048,574KB===>FirstCell(V/C-ID=3/1) is LayerBreakCell
.....VTS_01_6.VOB 1,048,574KB
.....VTS_01_7.VOB 900,072KB

[r0lZ]

Seems this way of authoring the LB at the beginning of a VOB is present on all Warner Bros DVDs. Jeanl, you should find an example easily. Now, it's up to you!

[blutach]

Yes, I have seen that all the time on Warner - they start a new VOBID at the LB. Pretty convenient too if you ask me cos the audio buffer will have run down. Natural non-seamless position.

Regards

[frank]

I believe Nero makes the same at the LB. They split a VOB automaticly when needed.
All my tests with Nero produced coasters. :devil:

[r0lZ]

...except when doing a 1:1 direct copy. In this case, Nero works well.

You should try RecordNow or GEAR... if you want an alternative to PgcEdit/DVD Decrypter.

[frank]

Hey, since PgcEdit can burn DL I'm happy with it! :D

Some ideas to clean up menus:
The Burn DVD Setup menu should be moved to Options-Install as submenu.
The Preview Options and Virtual Player Setup should be moved to Options-Install as submenus.

Start in Trace mode works only if Launch Open DVD.. ist activated.
Why not start without conditions?

Suggestion: submenus
Launch Open DVD dialog?
DVD Start...
-Trace
-BOV Finder
-Create Backup at first time
Preview...
Virtual Player...

[jeanl]

OK guys, I found a small bug in DVD2AVI which was the cause of all the problems you've seen, when trying to go straight to the first LBA of a middle VOB (i.e. vts_02_3.vob). The problem was not related to the fact that the previous VOB was not 1GB. It was related to the fact that the cell started exactly at the first lba of a middle vob, which is normally exceedingly rare, except when it's done on purpose!!!
Anyway. The problem is now fixed, and r0lZ has the new versions. I suspect he'll soon post a new beta...
jeanl
P.S. I will let neuron2 know.

[frank]

Then IfoEdit's VOB cell preview apparently has the same bug. [edit]
:thanks:

[jeanl]

That's surprising, from what I gather, IFOEdit uses the direct-show engine to play dvds. It must be another problem...
jeanl

[r0lZ]

Well, I don't make a new PgcEdit beta, because I have not changed the code today (except renaming Dual Layer to Double Layer).
If you want to test the new Preview, just download it (http://www.videohelp.com/~r0lZ/pgcedit/beta/PgcEditPreview20.zip) and copy the executable in the PgcEdit's install dir\bin subdirectory.
Seems all known bugs are now fixed. Thanks jeanl!

[frank]

@jeanl
I mean the VOB preview of IfoEdit, starting when you click on a cell in the main window. In the buggy case the window is grey and no preview starts.

[jinjin_jp]

Well, I don't make a new PgcEdit beta, because I have not changed the code today (except renaming Dual Layer to Double Layer).
If you want to test the new Preview, just download it (http://www.videohelp.com/~r0lZ/pgcedit/beta/PgcEditPreview20.zip) and copy the executable in the PgcEdit's install dir\bin subdirectory.
Seems all known bugs are now fixed. Thanks jeanl!


I comfirmed "PgcEditPreview20" fixed the problem that "play" button of preview is inactive(gray out) in example
(1)Matrix(RC2,NTSC)
(2)Animatrix(RC2,NTSC)
(3)Ocean's 12(RC2,NTSC)
Thanks jeanl.

[jeanl]

@jeanl
I mean the VOB preview of IfoEdit, starting when you click on a cell in the main window. In the buggy case the window is grey and no preview starts.
That also uses direct-show (I asked derrow if I remember correctly). It's somewhat surprising that errors crop up on the same files in the two apps! But I really think there's no real link between the two...
jeanl

[jinjin_jp]

More information about burning -R/DL was added in the bottom comparison table here (http://yaki2fan.hp.infoseek.co.jp/-r_dl/-r_dl.html).
And below is the table I tried to English.

In the table the value "L0" of burned -R/DL is always 2,092,896sectors and LayerBreakCell is diffent from original by ISO-Writing(1)&(2) of DVDDecrypter.
(1)Using MDS file which is output when ISO-Reading from original.
(2)Setting "sectors in L0" of "User Specified" as same as original.
===>Both Layer Break information seems to be ignored about -R/DL.
http://img210.imageshack.us/img210/8905/image05070310fl.jpg (http://www.imageshack.us)

Considering about above things, here (http://yaki2fan.hp.infoseek.co.jp/-r_dl/dec_-dl.html) is explained the process for padding necessary Sectors and adjusting Layer Break Position.
It is comlicated,but this process is equivalent to the process of PgcEdit like below Figure quoted,
1. set A(Final position layer break) to 2,092,896.
2. set B(Offset in L0) to the corresponding to A.
.....in this case
.........2,092,896(new A)-1,791,232(old A)+12(old B)=301,676(new B)
http://img210.imageshack.us/img210/6664/image05070328ah.jpg (http://www.imageshack.us)

[r0lZ]

Well, there is still something I don't unserstand.

If a DVD-R DL is minimum 4,150,400 sectors long, it means that a single layer is 2,075,200 sectors minimum.
In this case, it is NOT possible to force the layer break at sector 2,092,896, which is higher than the number of sectors per layer!
So, I imagine that the total size of the discs used in the tests is 2 x 2,092,896 = 4,185,792 sectors. As the DVD-R specs are not precise on the size of the DVD, I cannot assume this number is the right one.
Also, why is it needed to fill the entire disc with padding sectors? Does it mean that the -R technology is not compatible with the DVD-Video specs? In this case, you should not be able to do a 1:1 copy.

Anyway, I will make a special beta version (for the jap market ;)), with a field to input the layer break sector, to be able to force it at this position instead of the optimal position computed by PgcEdit.

Of course, the best solution is to avoid -Rs, and use only the good technology: +R!

Jinjin_jp, can you confirm that the DVD-R used in the tests are 4,185,792 sectors long?

[jinjin_jp]

Well, there is still something I don't unserstand.

If a DVD-R DL is minimum 4,150,400 sectors long, it means that a single layer is 2,075,200 sectors minimum.
In this case, it is NOT possible to force the layer break at sector 2,092,896, which is higher than the number of sectors per layer!
So, I imagine that the total size of the discs used in the tests is 2 x 2,092,896 = 4,185,792 sectors. As the DVD-R specs are not precise on the size of the DVD, I cannot assume this number is the right one.
I think strange, too.　"4,185,792 sectors" is too big comared with the information.


Also, why is it needed to fill the entire disc with padding sectors? Does it mean that the -R technology is not compatible with the DVD-Video specs? In this case, you should not be able to do a 1:1 copy.
About this, I think it may not be the problem of -R technology, but the defect of drive or burning-software because of the early stage step.
Actually RecordNow4.0 was not be able to burn DVD-Video files appropriately in the case that order of titles isn't order of VTS. (I can't explain well, it means that title1(in VTS_02),title2(in_VTS_01).)
This bug was fixed version 4.61.


Anyway, I will make a special beta version (for the jap market ;)), with a field to input the layer break sector, to be able to force it at this position instead of the optimal position computed by PgcEdit.
Thanks very very much.


Of course, the best solution is to avoid -Rs, and use only the good technology: +R!

Jinjin_jp, can you confirm that the DVD-R used in the tests are 4,185,792 sectors long?
Sorry I don't have the condition of -R/DL tests, so now I requested to a companion who made the sites I've been introdeced.

[Tobii]

Perhaps it helps you along?

My two Verbatims, with the details from DVD Identifier...

Disc & Book Type : DVD-R DL / DVD-R
Manufacturer Name : Mitsubishi Kagaku Media
Manufacturer ID : MKM 01RD30
Blank Disc Capacity : 4,171,712 Sectors = 8,147.9MB = 7.96GB (8.54GB)


Sorry, I don't have more as two. :(

[frank]

@jinjin_jp
Please don't post things from hearing and rumors! :devil:

Things are complicated enough. Why you don't test yourself with PgcEdit?
Trust PgcEdit. Test Dual Layer -R without manipulating and you will see.

I repeat again: There are no new problems.
There are many professions which published reports about dual layer DVD-R technics on the web. I have posted the links above.

[r0lZ]

Blank Disc Capacity : 4,171,712 Sectors = 8,147.9MB = 7.96GB (8.54GB)So, with a Verbatim, this 2,092,896 number doesn't make sense!

BTW, have you already burned these DVDs with PgcEdit? What is the result?