Polar Express Angle Problem

[foobaz]

I haven't been able to find anything here or elsewhere on dealing with the angles present in Polar Express R1. Usually the angles are in separate cells and I just blank the cells containing the unwanted angles in VobBlanker and then insert a cell command to jump to the next good cell to minimize the delay in the player. I have tried all recommended methods dealing with angles and this is the only one I've come up with that works reliably. All other methods have failed.

The problem is, in this title the angles are contained in the same cell so my method won't work. The angles are not selectable via the angle button on the remote but are selected automatically based on the language you choose. The angle used determines the language of the newspaper displayed for instance in the first chapter. Worse, IfoEdit doesn't allow you to choose an angle. 'Remove Angles' is greyed out in the VOB Extras dialog.

I tried just burning it as is but had the usual problems in the set-top player - all angles were played.

I have no idea how to deal with this. Can someone with experience with this please help out. BTW I want to make a two-disk split so DVDShrink, etc., are out.

[jeanl]

foobaz, what you describe is extremely common, at least in cartoons. I've seen lots of disney titles that have this kinds of angles (for titles, newspapers etc). Why do you want to remove the angles anyway? If you're doing a two-disc split, you don't need to remove the angles, right? What's this about your settop player playing all angles?! Was that using your original, or a backup? How did you do this backup?
jeanl

[foobaz]

The problem is that the angle cells are interleaved (ILVUs) and all angles play in succession. You can see this in the PgcEdit previewer as well. Try for example cell 3.

As I stated, usually the angles are in separate cells, including Disney titles, so I can just blank the unwanted cells. VobBlanker de-interleaves the remaining cells so they play normally. This is the first time I have encountered them all in the same cell.

I have no idea why a rip would behave differently from the original but it always does. It works just fine in a software player like WinDVD. Why can't some tool, like DVDD, PgcEdit, VobBlanker, etc. just fix this?

[jeanl]

foobaz, I'm not sure what you mean. True angles are always in ILVUs, otherwise, they would be played following one another. PgcEditPreview always shows the angles consecutively, because it's not smart enough to jump from one ILVU to the next. But maybe you mean, in some movies you have the english/french/spanish versions in different PGC that reference different cells. Note that in such cases, playback is usually not seamless (that's the reason why angles are used, really).

In any case, you're saying that when you rip, then create your two-disc backup, everything works fine on your soft player, but no in your settop box? This is very strange and abnormal...
Anyway, I'm afraid I don't have much advice to give you... Sorry....
jeanl

[setarip_old]

But maybe you mean, in some movies you have the english/french/spanish versions in different PGC that reference different cells.
That is exactly the situation for "Polar Express"

I tried just burning it as is but had the usual problems in the set-top player - all angles were played.
I have no idea how to deal with this.
This is what I did: If you simply rip this DVD, using DVD Decrypter set to its default settings, the rip will not exhibit the behavior you've described.

If you then want to split it over 2 discs (and retain the active menu), one simple method would be to use the OLD(freeware) "DVDFab". You can find it at:

mrbass.org

BTW - If you're talking about converting to .AVI, that's a whole different story (Just rip one PGC)

[foobaz]

True angles are always in ILVUs, otherwise, they would be played following one another. PgcEditPreview always shows the angles consecutively, because it's not smart enough to jump from one ILVU to the next. But maybe you mean, in some movies you have the english/french/spanish versions in different PGC that reference different cells.


VTST 1 , 1 TTN 1 (1:39:45) Title 1 - Chapters: 25, Programs: 25, Cells: 54

Playback time: 01:39:45.21 (at 30 fps)
PG Playback mode: sequential
PUOs: 0 (0x00000000)
NextPGCN: 1
PrevPGCN: 1
GoUpPGCN: 0
PGC Still Time: 0
Audio stream 1 status: 0x00008000 (stream=0)
Subpic stream 1 status: 0x80000100 (streams for 4:3=0, wide=0, letterbox=1, pan&scan=0)
Subpic stream 2 status: 0x80020300 (streams for 4:3=0, wide=2, letterbox=3, pan&scan=0)
Subpic stream 3 status: 0x80040500 (streams for 4:3=0, wide=4, letterbox=5, pan&scan=0)

Chap. Prog. Cell Type Layer Ang VOBU Cell Cell Playback End Entry First Last Last VOB Cell
(PTT) Flags Break Still Still Cmd. Time Time VOBU ILVU VOBU VOBU ID ID
Time # sector End Start End

1 1 1 2 yes - no 0 0 00:03:28.10 00:03:28.10 19726 0 107103 107419 5 1
2 10 no - no 0 0 00:00:04.00 00:03:32.10 107420 0 108575 108803 6 1
3 14 no - no 0 0 00:00:40.06 00:04:12.16 108804 109290 147598 147791 7 1
4 12 no - no 0 0 00:00:04.00 00:04:16.16 147792 147876 150213 150359 7 2
5 10 no - no 0 0 00:00:45.14 00:05:02.00 151538 0 164925 165073 10 1


VTST 1 , 2 TTN 2 (1:39:45) Title 2 - Chapters: 25, Programs: 25, Cells: 54

Playback time: 01:39:45.21 (at 30 fps)
PG Playback mode: sequential
PUOs: 0 (0x00000000)
NextPGCN: 0
PrevPGCN: 0
GoUpPGCN: 0
PGC Still Time: 0
Audio stream 2 status: 0x00008100 (stream=1)
Subpic stream 1 status: 0x80000100 (streams for 4:3=0, wide=0, letterbox=1, pan&scan=0)
Subpic stream 2 status: 0x80020300 (streams for 4:3=0, wide=2, letterbox=3, pan&scan=0)
Subpic stream 3 status: 0x80040500 (streams for 4:3=0, wide=4, letterbox=5, pan&scan=0)

Chap. Prog. Cell Type Layer Ang VOBU Cell Cell Playback End Entry First Last Last VOB Cell
(PTT) Flags Break Still Still Cmd. Time Time VOBU ILVU VOBU VOBU ID ID
Time # sector End Start End

1 1 1 2 yes - no 0 0 00:03:28.10 00:03:28.10 19726 0 107103 107419 5 1
2 10 no - no 0 0 00:00:04.00 00:03:32.10 107420 0 108575 108803 6 1
3 14 no - no 0 0 00:00:40.06 00:04:12.16 109291 109768 148039 148265 8 1
4 12 no - no 0 0 00:00:04.00 00:04:16.16 148266 148350 150802 150948 8 2
5 10 no - no 0 0 00:00:45.14 00:05:02.00 151538 0 164925 165073 10 1


This title has 3 separate PGCs for the three languages. In each PGC there are cells with ILVUs. For example, cell 3 above. The PGCs all reference the same content and are therefore parallel. The only difference is that the Entry VOBU sector, etc. and the VOB IDs for corresponding cells in each PGC are different for the ILVU cells.


Chap. Prog. Cell Type Layer Ang VOBU Cell Cell Playback End Entry First Last Last VOB Cell
(PTT) Flags Break Still Still Cmd. Time Time VOBU ILVU VOBU VOBU ID ID
Time # sector End Start End

PGC 1
3 14 no - no 0 0 00:00:40.06 00:04:12.16 108804 109290 147598 147791 7 1
4 12 no - no 0 0 00:00:04.00 00:04:16.16 147792 147876 150213 150359 7 2
PGC 2
3 14 no - no 0 0 00:00:40.06 00:04:12.16 109291 109768 148039 148265 8 1
4 12 no - no 0 0 00:00:04.00 00:04:16.16 148266 148350 150802 150948 8 2


From this can be inferred that each PGC uses a different section of the same cell for its particular language.

I have commonly seen the case where there is only one PGC but multiple ILVU cells for the different languages. A good example is Myazaki films. This allows simply blanking unwanted cells. But this case is different since the same cell is used for the different languages. If there is a tool that will blank just the unwanted VOB IDs and de-interlace the cells, that would probably work.

In any case, you're saying that when you rip, then create your two-disc backup, everything works fine on your soft player, but no in your settop box? This is very strange and abnormal...

I don't think this is that unusual. I recall many posts where people had the same problem with Star Wars and The Matrix.

[foobaz]

This is what I did: If you simply rip this DVD, using DVD Decrypter set to its default settings, the rip will not exhibit the behavior you've described.
It plays fine in a software player but did you try it in your set-top player? This is not an uncommon problem.

[setarip_old]

It plays fine in a software player but did you try it in your set-top player?

Yes. I compressed the DVD Decrypter v.3.5.4 DVD "package" with DVD2One. The resultant burned DVD plays perfectly on all of my standalone DVD players.

BTW - Mine is also Region 1...

[jeanl]

Same here. I can't speak about polar express, but all the star wars, and all the matrix DVDs have been no problems. I think it's rather unusual for a backup to exhibit problems with angles. I think you should try to figure out what the problem is withe the settop box, rather than try to remove angles (which isn't that easy).

From looking at your last example, there is a problem here. The angle flag should be set! I can see that 7/1 and 8/1 are interleaved, but if that's the case, then there needs to be an angle flag, right? can you check in your original if the original IFOs show an angle flag?
Jeanl

[foobaz]

I think it's rather unusual for a backup to exhibit problems with angles. I think you should try to figure out what the problem is withe the settop box, rather than try to remove angles (which isn't that easy). I don't know but I see many posts about removing angles here. Why would you do that unless you were having a problem playing the angle? As far as the set-top, the problem is it is cheap and apparently not smart enough to deal with the rip anyway. It has no such problem with the original though. Still, I don't understand why the rip would not behave identically to the original. BTW I have tried DVDFab in the past and it copied all the angles and they behaved just like the original on the set-top. So why don't DVDD rips behave the same way?

From looking at your last example, there is a problem here. The angle flag should be set! I can see that 7/1 and 8/1 are interleaved, but if that's the case, then there needs to be an angle flag, right? can you check in your original if the original IFOs show an angle flag?
Chap. Prog. Cell Type Layer Ang VOBU Cell Cell Playback End Entry First Last Last VOB Cell
(PTT) Flags Break Still Still Cmd. Time Time VOBU ILVU VOBU VOBU ID ID
Time # sector End Start End

1 1 1 2 yes - no 0 0 00:03:28.10 00:03:28.10 19726 0 107103 107419 5 1
2 10 no - no 0 0 00:00:04.00 00:03:32.10 107420 0 108575 108803 6 1
3 14 no - no 0 0 00:00:40.06 00:04:12.16 108804 109290 147598 147791 7 1
4 12 no - no 0 0 00:00:04.00 00:04:16.16 147792 147876 150213 150359 7 2
5 10 no - no 0 0 00:00:45.14 00:05:02.00 151538 0 164925 165073 10 1
It's that way in the original too but it does work OK in the set-top. I really don't know enough about ILVUs to understand the nature of the problem and I have not been able to find detailed information on the web. And I don't want to pay $5,000 for Book A and Book B of the official DVD spec. (Anyone have a link to the pdf of this? :sly: )

[setarip_old]

BTW I have tried DVDFab in the past and it copied all the angles and they behaved just like the original on the set-top. So why don't DVDD rips behave the same way?
I mentioned this previously as a suggestion, so now I'll phrase it as a question:

Have you altered ANY of the original default settings for DVD Decrypter? (You shouldn't have)- or used any additional software to make any changes (You shouldn't have)?

[blutach]

@foobaz - as you have noted, each of the different title PGCs seems to have different VIDs - it's just the way the titles have been constructed. Why don't you strip the VIDs you don't want with IfoEdit?

Regards

[foobaz]

Have you altered ANY of the original default settings for DVD Decrypter? Yes I changed them years ago and I changed the number of read error retries due to the recent spate of discs with the dread ARcoSS corruption. But why would this affect angle play?

BTW are you ripping the whole disc using File mode?

[setarip_old]

BTW are you ripping the whole disc using File mode?

Yes...

[foobaz]

Have you altered ANY of the original default settings for DVD Decrypter?I reset DVDD 3.5.4.0 to the default settings via the 'Defaults' button and did a rip in 'File' mode. The results were exactly the same, as I expected. I don't see why this would make any difference. So maybe jeanl is right. Maybe it is my player. But my player has no problem with the original so why the difference? DVDD must be altering something . . . ? ? ? Can anyone here explain this?

[jeanl]

foobaz, maybe more info might be needed: I assume polar express is a DVD-9, what did you use to do your two-disc split? Maybe the problem is here?
I'm still very surprised by the lack of angle flags, but I'm not an expert with angles, and if you say the original does not have them and plays just fine everywhere, then that should be fine. There is something different, however, between your original, and the modified version obviously... So, if I were you, I would try to compare the original unaltered IFOs with the modified one (after splitting), and try to see if there's any difference. I use a program that I wrote to do that, if you send me the IFOs at jeanldvd at free dot fr, I can try to see if I see any difference. Of course, there will be difference because of the splitting...

Can you describe the entire procedure you followed to get the your final burn?
jeanl
P.S. Also, if you have a friend with another DVD player, try to play it on his/hers, that would be informative...

[jeanl]

foobaz, I took a look at your IFOs, and I can see something wrong. You still haven't told us how you did the disc split, so I'm not sure where the error comes from but here's what I see that seems wrong.
Open VTS1 - PGC1: This is the original

Chap. Prog. Cell Type Layer Res- Still Cell Playback End Entry First Last Last VOB Cell
(PTT) Flags Break tric- Time Cmd. Time Time VOBU ILVU VOBU VOBU ID ID
ted. sector End Start End

1 1 1 2 yes no 0 0 00:03:28.10 00:03:28.10 19726 0 107103 107419 5 1
2 10 no no 0 0 00:00:04.00 00:03:32.10 107420 0 108575 108803 6 1
3 14 no no 0 0 00:00:40.06 00:04:12.16 108804 109290 147598 147791 7 1
4 12 no no 0 0 00:00:04.00 00:04:16.16 147792 147876 150213 150359 7 2


And this is your modified, disc 1, version:

Chap. Prog. Cell Type Layer Res- Still Cell Playback End Entry First Last Last VOB Cell
(PTT) Flags Break tric- Time Cmd. Time Time VOBU ILVU VOBU VOBU ID ID
ted. sector End Start End

1 1 1 2 yes no 0 0 00:03:28.10 00:03:28.10 20 0 87397 87713 5 1
2 10 no no 0 0 00:00:04.00 00:03:32.10 87714 0 88869 89097 6 1
3 10 no no 0 0 00:00:40.06 00:04:12.16 89098 0 102177 102370 7 1
4 8 no no 0 0 00:00:04.00 00:04:16.16 102371 0 103232 103378 7 2


As you can see, in your modified version, the fields "First ILVU end" are all 0. This is crucial information that's missing. First ILVU end indicates the end of the first interleaved unit in the cell (i.e. in this case, where cell 8/1 starts).
Furthermore, your flags are missing the "interleaved" tick, which means the player isn't aware that the cells are interleaved...
I think this explains what you see. You mention in your email that "What
happens is, the correct angle does play back in the set-top but it is
choppy and pixellated in transition spots" . This would be consistent with the player not being "ready" to jump over the interleaved cells and showing some problems at boundaries. This would also explain why soft player can deal with that (they have much longer look-ahead).

In any case, there's a problem with the way you've done your split. Can you give us more info?
Thanks,
Jeanl

[foobaz]

I assume polar express is a DVD-9, what did you use to do your two-disc split? Maybe the problem is here?Wow, that was fast. OK, I ripped it with DVDD 3.5.4.0 using 'File' mode. I modified the entire title with PgcEdit 0.6.1 to eliminate warnings, etc. I used AddChapter 1.2.0.2 to add a blank cell to the movie PGC and enabled all operations so the forward button can be used to skip the closing chapter. I then split it using VobBlanker 2.0.0.2. I only blanked chapters 19-24 in PGC 1, the English version and didn't bother with the other two language PGCs since they reuse the video content. I then added a LinkTailPGC cell command with PgcEdit to skip over blanked cells after kept cells are played. I added a 'Continued on Disc 2' VOB with PgcEdit and removed irrelevant chapter buttons. Then I created an ISO and burned it with ImgBurn using the PgcEdit interface. I have done this procedure many times and the only time I ever have a problem is when there are ILVUs.

BTW I get the following warning in PgcEdit when I open the disc 1 split:

Warning: There are different commands
in the various PGCs of the menu LUs
of the VTST 1 domain.

This is somewhat unusual, and may cause
problems with the 'Set Menu Type' and
'Copy LU Commands' functions.
I'm still very surprised by the lack of angle flags, but I'm not an expert with angles, and if you say the original does not have them and plays just fine everywhere, then that should be fine.Yes, I posted the original above. You can see it is identical.

I use a program that I wrote to do that, if you send me the IFOs at jeanldvd at free dot fr, I can try to see if I see any difference.Done . . .

Thanks so much for your kind consideration in helping me with this frustrating problem.

[foobaz]

Why don't you strip the VIDs you don't want with IfoEdit?I tried this but the chapter points were all messed up. I tried to fix them with IFOUpdate but it didn't work. In any case, I'd rather not have to remove angles since it is always such a hassle and it would be nice to have them available. Thanks for the suggestion though.

[jsoto]

. I then split it using VobBlanker 2.0.0.2. I only blanked chapters 19-24 in PGC 1,
VobBlanker de-interleaves the cells. So it clears all the First ILV pointers in the PGCs. (And many others in the VOBs)

This is NOT a multiangle DVD, but a ILV multistory (may be it is not the fully right name, but...) In this kind of DVDs the length of the different branches can be different, and IIRC the joints can (must) be non-seamless. Alien Quadilogy is another example of this.
In a multistory made using a true multiangle (like Star Wars or the incredibles) the length of the different branches must be the same.

In any case I do not understand why your player is playing all the cells. The player must follow the PG chain, and the "stories" are different cells (7/1 & 7/2 versus 8/1 & 8/2)



jsoto

[jsoto]

In any case I do not understand why your player is playing all the cells. The player must follow the PG chain, and the "stories" are different cells (7/1 & 7/2 versus 8/1 & 8/2)
jsoto
Could you check (in the processed DVD) if 7/1 and 7/2 contents is different from 8/1 and 8/2?. you can use pgcEditpreview or VobBlanker preview for this.

[foobaz]

foobaz, I took a look at your IFOs, and I can see something wrong. You still haven't told us how you did the disc split, so I'm not sure where the error comes fromYou responded so fast, I hadn't yet finished my reply to that. ;)

here's what I see that seems wrong.
As you can see, in your modified version, the fields "First ILVU end" are all 0. This is crucial information that's missing. First ILVU end indicates the end of the first interleaved unit in the cell (i.e. in this case, where cell 8/1 starts).
Furthermore, your flags are missing the "interleaved" tick, which means the player isn't aware that the cells are interleaved...Holy $h!t! You're absolutely right. And I was assuming all along that this all would be identical in the split . . .
This is what VobBlanker is outputting. And, come to think of it, I get a warning when I open the rip in VobBlanker:

There are at least one ILV Cell.
The VTSs containing these cells have been marked as SKIP.
You can process them, but playback problems can appear
in all PGCs originally containing ILV cells except the first one.Actually the playback problems are in the first PGC. So VobBlanker is not handling the angles properly. One thing I don't understand is why the start/stop positions are all different in the split.

So is there an alternative to using VobBlanker to split this? One that will deal with the angles? BTW I have tried using DVDD and IfoEdit to do this with poor results.

[jsoto]

Well, the warning about "except in PGC1" is quite obsolete.... It was true in VobBlanker 1.X...

In any case, VB is de-interleaving, so it must modify these values, because , after `processing, the cells are NOT interleaved.

BTW, why are you using 2.0.0.2 instead 2.0.1.0? (Although it should be the same)
jsoto

[jsoto]

So is there an alternative to using VobBlanker to split this? One that will deal with the angles? BTW I have tried using DVDD and IfoEdit to do this with poor results. There are NO angles in your original DVD.

[foobaz]

VobBlanker de-interleaves the cells. So it clears all the First ILV pointers in the PGCs. (And many others in the VOBs)I see. In fact when I play the ILVU cells in PgcEdit, they are now de-interleaved.

This is NOT a multiangle DVD, but a ILV multistory (may be it is not the fully right name, but...) In this kind of DVDs the length of the different branches can be different, and IIRC the joints can (must) be non-seamless.You are right. It is not multi-angle. That would explain why IfoEdit doesn't allow angle processing and why the flags indicate interleaving but not multiple angles. I believe the term is 'Partial-Interleave Blocks':
http://www.uemforums.com/2pop/ubbthreads/printthread.php?Board=DVDAuthoringCreation&main=86352&type=post

In a multistory made using a true multiangle (like Star Wars or the incredibles) the length of the different branches must be the same.Yes and in that case each angle is in a separate cell in one PGC. They are of couse interleaved.

In any case I do not understand why your player is playing all the cells. The player must follow the PG chain, and the "stories" are different cells (7/1 & 7/2 versus 8/1 & 8/2)Yes it must. I mentioned to jeanl in my email that I had slightly misstated the problem. The set-top does only play the desired cell but it is choppy and pixellated at transition points. But then, this is the de-interleaved version that VobBlanker has output. In the original, all three angles are present in the one cell and you see all three in the previewer for cell 3 VOBID 7, 8, or 9. Presumably that's because all VOBIDs are in one cell which has not been de-interleaved.

[foobaz]

Could you check (in the processed DVD) if 7/1 and 7/2 contents is different from 8/1 and 8/2?. you can use pgcEditpreview or VobBlanker preview for this.Yes they are different. The same scene but different languages in the newspapaer for example.

[foobaz]

In any case, VB is de-interleaving, so it must modify these values, because , after `processing, the cells are NOT interleaved.I see. Is it possible to still retain the "First ILVU end" values jeanl is talking about? This may eliminate the choppy transitions.

BTW, why are you using 2.0.0.2 instead 2.0.1.0? (Although it should be the same)
jsotoJust resistance to change I guess. ;) I will have to change all the parameters.

[jsoto]

I mentioned to jeanl in my email that I had slightly misstated the problem. The set-top does only play the desired cell but it is choppy and pixellated at transition points. But then, this is the de-interleaved version that VobBlanker has output.
Well, this is the problem we have to track. But I have no clue.... The cells have a new VID and are marked as non-seamless, so the initial GOP must be self-contained and the video should not be pixelated.


In the original, all three angles are present in the one cell and you see all three in the previewer for cell 3 VOBID 7, 8, or 9. Presumably that's because all VOBIDs are in one cell which has not been de-interleaved.Dont trust the previewer. VobBlanker's one advices you that there are ILV cells, and the playback will not be OK. A player will not play all the cells.

[jsoto]

I see. Is it possible to still retain the "First ILVU end" values jeanl is talking about? This may eliminate the choppy transitions.

No, after processing, the cells are "normal" and non-interleaved so these pointers must be zero.

[foobaz]

Well, this is the problem we have to track. But I have no clue.... The cells have a new VID and are marked as non-seamless, so the initial GOP must be self-contained and the video should not be pixelated.The cells seem to have the same VID:

Original
Chap. Prog. Cell Type Layer Ang VOBU Cell Cell Playback End Entry First Last Last VOB Cell
(PTT) Flags Break Still Still Cmd. Time Time VOBU ILVU VOBU VOBU ID ID
Time # sector End Start End

1 1 1 2 yes - no 0 0 00:03:28.10 00:03:28.10 19726 0 107103 107419 5 1
2 10 no - no 0 0 00:00:04.00 00:03:32.10 107420 0 108575 108803 6 1
3 14 no - no 0 0 00:00:40.06 00:04:12.16 108804 109290 147598 147791 7 1
4 12 no - no 0 0 00:00:04.00 00:04:16.16 147792 147876 150213 150359 7 2
5 10 no - no 0 0 00:00:45.14 00:05:02.00 151538 0 164925 165073 10 1

Disc 1
Chap. Prog. Cell Type Layer Ang VOBU Cell Cell Playback End Entry First Last Last VOB Cell
(PTT) Flags Break Still Still Cmd. Time Time VOBU ILVU VOBU VOBU ID ID
Time # sector End Start End

1 1 1 2 yes - no 0 0 00:03:28.10 00:03:28.10 20 0 87397 87713 5 1
2 10 no - no 0 0 00:00:04.00 00:03:32.10 87714 0 88869 89097 6 1
3 10 no - no 0 0 00:00:40.06 00:04:12.16 89098 0 102177 102370 7 1
4 8 no - no 0 0 00:00:04.00 00:04:16.16 102371 0 103232 103378 7 2
5 10 no - no 0 0 00:00:45.14 00:05:02.00 131832 0 145219 145367 10 1

But the sectors are all different.

Dont trust the previewer. VobBlanker's one advices you that there are ILV cells, and the playback will not be OK. A player will not play all the cells.No I know. I was just pointing out how you can see that the VobBlanker version is de-interleaved where the original is not.

[jeanl]

jsoto, thanks for the clarification between multi-angle and simply interleaved. I don't think I had encountered multi-story, single angle PGCs before!

Well, this is the problem we have to track. But I have no clue.... The cells have a new VID and are marked as non-seamless, so the initial GOP must be self-contained and the video should not be pixelated.

jsoto, the original cells were all marked seamless, so we have a problem here. The original VOBUS might not have had closed GOP, right? After processing with vobblanker, the cells are marked non-seamless. The player won't behave nicely in this case. Especially if the first VOBU of the next cell has an open GOP, I'm not sure what the player will do in that case, when it's not told it's a seamless joint.
I'm guessing vobblanker does not remux anything within the VOBU when it de-interleaves the ILVU cells, so foobaz could possibly re-mark the cells as seamless right? The audio multiplex is identical in the output as it is in the input, right? So the player shouldn't have any problem then...

foobaz, this might be something you might want to try: check the seamless joints in your cells.
Jeanl

[jeanl]

This is NOT a multiangle DVD, but a ILV multistory (may be it is not the fully right name, but...) In this kind of DVDs the length of the different branches can be different, and IIRC the joints can (must) be non-seamless. Alien Quadilogy is another example of this.
jsoto
All joints are seamless in the original...
jeanl

[jsoto]

Yes, the seamless mark was the first thing I thought in....

But I think this is not the case.

The joints we can see in the posts are:
between 6/1 and 7/1
between 7/2 and 10/1

Yes, you're right, both are marked as seamless (I was wrong here...) But I'm still confused... VB does not change anything ....

AAh!
I know.
There is a restriction (there is a mpucoder's post about) in the difference (in secs, IIRC) between two different cells/angles seamlessly joined. I'm sure VB is breaking this rule deinterleaving the cells... (Now, the laser needs to jump the whole VID 8 and 9 to go from VID 7 to VID 10)

So, try making the transitions non-seamless and see...

jsoto

EDIT: Found mpucoder's post: 20.000 sectors
http://forum.doom9.org/showthread.php?p=456639#post456639

[jeanl]

So, try making the transitions non-seamless and see...

jsoto
you mean, seamless, right? ;)
jeanl

[jeanl]

Jsoto, there's no way one can coerce vobblanker into NOT de-interleaving the cells, is there?
jeanl

[jsoto]

No , I mean non-seamless, let's say 2 or 0.

This will not fix the open GOP problem. for sure, but may be the player behavior will be better...

jsoto

[jsoto]

Jsoto, there's no way one can coerce vobblanker into NOT de-interleaving the cells, is there?
jeanl There is no way to process a title domain w/o de-interleaving the cells.

[foobaz]

jsoto, the original cells were all marked seamless, so we have a problem here.foobaz, this might be something you might want to try: check the seamless joints in your cells.
JeanlJeanl, they are flagged as 'Seamless joint' (8) in both cases. The main difference is they are also flagged as 'Interleaved' (4) in the original which is appropriate and does not apply to the new cells.

[foobaz]

So, try making the transitions non-seamless and see...So do I make 7/1 and 7/2 non-seamless? And how about 8/1, 8/2, 9/1, and 9/2?

[jeanl]

I would think 7/2 (seamless flag applies to the transition between the previous cell and the current one - the one with the flag).
jeanl

[jsoto]

I would think 7/2 (seamless flag applies to the transition between the previous cell and the current one - the one with the flag).
jeanl
In the PGC
5/1 6/1 7/1 7/2 10/1
the jump will happen in 7/2 to 10/1 transition so I think you have to made non-seamless 10/1.

In other PGC (Not sure if it is exact... )
5/1 6/1 8/1 8/2 10/1
The jumps will happen in 6/1 to 8/1 transition and in 8/2 to 10/1 so I think you have to made non-seamless 8/1 and 10/1.

jsoto

[foobaz]

In the PGC 5/1 6/1 7/1 7/2 10/1
the jump will happen in 7/2 to 10/1 transition so I think you have to made non-seamless 10/1.So the next cell wherever there's a jump to a non-consecutive cell should be made non-seamless. I can try it but I have little hope at this point. I watched the offending cell frame-by-frame in the set-top and I saw that pixellation occurs at every scene change (cross-fade) within the one cell. There are two or three frames that are severely and noticeably pixellated in each case. And there is a two or three frame pause as well. Really horrible. It happens 8 times in cell 3 VID 7/1, a pretty small cell . . . :(

It looks like some sub-genius has inadvertantly stumbled upon a very effective copy-protection scheme. At least as far as making 1:1 splits is concerned. (Doom9: Maybe you ought to censor this last paragraph. It might give ideas to the DRM gestapo. ;))

[jeanl]

foobaz, there might be other ways of doing your split that don't require to de interleave the ILVUs. For example, I think DVDRemake (maybe pro) can do that (but it's not free, but not very expensive, given the amazing features). You might also be able to do this one with IFOEdit...
jeanl

[foobaz]

foobaz, there might be other ways of doing your split that don't require to de interleave the ILVUs. For example, I think DVDRemake (maybe pro) can do that (but it's not free, but not very expensive, given the amazing features). You might also be able to do this one with IFOEdit...
jeanlThanks for the suggestion. Maybe I'll look into DVDRemake though I'd prefer sticking to freeware. I already tried removing VIDs with IfoEdit with disastrous results. Do you know any way to recover correct chapter points after processing with IfoEdit?

[jeanl]

foobaz, the point is you would not have to remove any of the interleaved cells... Simply cut the vob in half and create two discs...
jeanl

[foobaz]

foobaz, the point is you would not have to remove any of the interleaved cells... Simply cut the vob in half and create two discs...
jeanlWith IfoEdit you mean. Well I have had many problems with doing that in the past. Do you have much experience with doing splits in IfoEdit? Do you think DvdReMake (not PRO) can do a split and retain the interleaving and all functionality of the original. If it would it might be worth buying though I hate to set a precedent here. ;) Are there no other freeware alternatives?

[jeanl]

foobaz, the best way to know would be to ask dimad (you can send him a pm, he's on this forum). I can assure you you won't be wasting your money with DVDRemake Pro. It's an amazing piece of software, if you look at the list of things it can do.
I've never done splits with IFOEdit myself (I don't believe in splits! ;) )
Another alternative is DVDStripper. It's an old piece of software that's no longer developed and can be hard to find on the net, but I have a version if you can't find one, somewhere on my PC... Other than that, you might have to do a search, for example at VideoHelp...
jeanl

[jsoto]

I do not understand the pixellation problem.... May be VB is not deinterleaving correctly...

Could you extract the small cell VID 7/1 (in the processed DVD) and send it to me? (email jesus_soto_viso at terra dot es)

Anyway, I'll try to get this particular DVD....

jsoto

[foobaz]

I do not understand the pixellation problem.... May be VB is not deinterleaving correctly...I was able to make a 1:1 split that worked in the set-top with the freeware version of DVDFab. I don't like to use this but I have found it does do a good job on titles with ILVUs. It does not de-inetrleave anything and all angles/stories are available. I still had lots of things to fix up like adding menus back, correcting sectors and the like. But at least the ILVUs play normally. I checked every one of them closely. One thing though; I could not remove all PUOs for some reason. I removed them in DVDD, PgcEdit, and IfoEdit but they were still there when I tested my modifications. Any idea why? I'm talking about being able to forward past transition clips between menus. I like to leave them but it is convenient to be able to bypass them when testing modifications.

Another weird thing that happened was with the second disc. DVDFab inserts a short transition cell for all the blanked out cells in the PGC before the valid cells. I simply insert a cell command in the first few cells to jump to the first valid cell or program. This worked fine in both the PgcEdit previewer and the software player. In the set-top however, when the play button is hit it starts at some weird position, not even at the beginnng of a cell, although it says cell 1. Forwarding to the next cell plays the same segment as does forwarding to each of the blanked cells. Finally when you reach the first valid cell all is normal after that. What the heck is going on here? I did copy over the VTS_01_0.VOB of the original rip to preserve menus and added back pre and post commands to the movie PGC. I did a 'Get VTS Sectors' in IfoEdit. To fix this odd behavior, I finally added a branch to the pre-commands to play the first valid program if program 1 was 'selected'. This effectively bypassed all the blanked cells and it worked fine in the set-top.

I tell you though, I much prefer using VobBlanker to do a split. Is there any way of preserving angles/stories in ILVUs without de-interleaving, as jeanl suggests? That would be the optimal situation. It seems that de-interleaving would actually be more trouble than just leaving the cells interleaved. Is that not true?

In any case, thanks much for looking into this for me. I appreciate the help. VobBlanker is a really useful tool.

[jsoto]

I tell you though, I much prefer using VobBlanker to do a split. Is there any way of preserving angles/stories in ILVUs without de-interleaving, as jeanl suggests?
No way. May be in a future.... (a far future...)

But, sorry, I still do not understand your problem de-interleaving (see your mail).... It should work (may be with minor problems in the transitions)
One thing to do to save the transitions is to set the broken bit in the GOP header. This should be enough to advice the settop to skip the very first frames (the ones with uncomplete data).

jsoto