Subs off by several seconds in a strange movie layout

[StifflerStealth]

Like the title says, the movie That I am trying to back up has a strange layout (explained below). I have been trying to back it up for 2 weeks now. I have read many post here, and tried several options including the InfiniFilm/BOV guide, even though this disc has neither. All three audio tracks are in sync, but the subtitles come around three to four seconds too late. Also, the chapter points are off, and they get off more and more as the one gets futher along in the video. Infact that when you select the chapter for the end credits, the video jumps to well after the credits have ended. This happens when I try to extract the data using the normal extract by PGC method. When imported into Scenarist, the cells under the Track editor tab and then clicking the plus by VTS01P01. The Cells are named VTS01_V3_C2. Is that the correct nameing sceme? If you look below, the reauthored VOBs become out of ordered.

The layout:
Format: NTSC
It is an episode disc.
All three episodes are in the same VTS, VTS 1 to be exact.
Each episode is in it's own PGC. As follows:
Original: (Vob/Cell ID)
PGC 1: (3/02) to (3/25); (4/01)
PGC 2: (1/01) to (1/08); (4/01)
PGC 3: (2/01) to (2/08); (4/01)
Cell 4/01 is a 15 frame cell that tells the player to go back to the menu.
I have pictures of the layout from IfoUpdate if you want to see those, but they are really big. There is no cell with VobID 3 and CellID 1. On top of that PGCs 2 and 3 come before PGC 1 in the encoded VTS 1 VOBs. (I looked at the VOBs).

When I re-encode I get this format in the VTS 1 IFO file:
Reauthored: (Vob/Cell ID)
PGC 1: (1/01) to (1/24); (4/01)
PGC 2: (2/01) to (2/08); (4/01)
PGC 3: (3/01) to (3/08); (4/01)

I think that layout is the cause for delays. I only looked at episode one (PGC 1) to figure out if the disk was correct. I think the chapter points and the subtitles are off because all the tools require a more normal format, i.e. Cell IDs starting at 1 not 2.

I used the new Big 3 (DIF4U --> BatchCCE --> ScenAid). I used ReAuthorist first, but that produced the same results as ScenAid. I used CapterXtracter to overwrite the CCE Chapter frames and then modified the BatchCCE jobs with the new chapters, but I still get warnings. I even tried to demux by CellID, but ScenAid does not support that mode. It only detected the first cell for each PGC, and would not add the other cells to the PGC. I then tried to do it by VobID, but ScenAid gave me a Type mismatch error, and the SCP file would not work with Scenarist 2.7. I would try Scenarist 3, but it keeps crashing during the m(f)uxing phase. I've been too lazy to check for updates.

Interesting notes:
When Demuxed by CellID, some cells required to be pulldowned, but others had the "nopull" lable. Also, the audio files had varied delays ranging from 0ms to + and - 166ms.
When Demuxed by VobID, none of the three video files required pulldown, and none of the audio files had an offset.

All Wranings and errors I get:
Warning Sub-Picture (stream 1 in Track "VTS01P01") is across scene boundary at 00:11:37:18
Warning Sub-Picture (stream 2 in Track "VTS01P01") is across scene boundary at 00:11:37:18
Warning Sub-Picture (stream 1 in Track "VTS01P01") is across scene boundary at 00:38:20:24
Warning Sub-Picture (stream 2 in Track "VTS01P01") is across scene boundary at 00:38:20:24
Warning Sub-Picture (stream 1 in Track "VTS01P01") is across scene boundary at 00:11:36:28
Warning Sub-Picture (stream 2 in Track "VTS01P01") is across scene boundary at 00:11:36:28
Warning Sub-Picture (stream 1 in Track "VTS01P01") is across scene boundary at 00:38:18:14
Warning Sub-Picture (stream 2 in Track "VTS01P01") is across scene boundary at 00:38:18:14
Warning Sub-Picture (stream 1 in Track "VTS01P01") is across scene boundary at 00:38:18:12
Warning Sub-Picture (stream 2 in Track "VTS01P01") is across scene boundary at 00:38:18:12
Warning Sub-Picture (stream 1 in Track "VTS01P01") is across scene boundary at 00:11:36:25
Warning Sub-Picture (stream 2 in Track "VTS01P01") is across scene boundary at 00:11:36:25
Error Class(Sub-Picture Play) Tag(Duration, 7203) failed to paste
Error Class(Sub-Picture Play) Tag(Duration, 7203) failed to paste

Any ideas on how to back this DVD up? I know this was a long post, but I though it would help to make things clear.

@Eyes: This is the DVD I told you about in the email when I registered for your proggy.

Thanks
Stiff

[influenza]

Well maybe and I say maybe there does exist a cell 1 which doesn't show up in the ifo (I've seen this once on spirited away). to find this out maybe you can do a cell demux on it? (no subs if I may advice you, cause that will take long. If this cell exists and is before the normal video this maybe a cause.

DOn't worry about the numbering on the vobids they won't matter.

[StifflerStealth]

When I did a CellID demux on this disc, a dummy Cell called VobID 3 Cell ID 1 was created. I thought it was created just to keep the nameing structure of PGC 1. On regular PGC and VobID demux, there is no mention of a Cell called vobID 3 cellid 1. After doing a demux by cellid, how can I write a scp script to import this into scenarist, or do I need to do the importing manually?

I know, it's a tough one to back up. That's why I thought I would post here to see if anyone can help me.

Stiff

[influenza]

Try to import this cell manually and add it to your existing layout. Maybe it works You'll need to edit the ifo afterwards I think, to keep the correct lay-out

[StifflerStealth]

I started to look around in the IFO file for VTS1 at the sector numbers to try to see if there were infact other cells that were not listed.

These are the sector ranges for each PGC minus VobID 4: (Vob/Cell ID)
PGC 1: 1,873,683 (3/02) - 3,931,077 (3/25)
PGC 2: 0 (1/01) - 936,612 (1/08)
PGC 3: 936,613 (2/01) - 1,873,301 (2/08)
VobID 4: 3,931,396 - 3,931,399

As you can see, Cell (2/01) is just the next sector over from Cell (1/08), so one can surmis that the next sequential cell is in the next sector. However that is not the case going from Cell (2/08) to Cell (3/02) or from Cell (3/25) to Cell (4/01).

Cells (2/08) and (3/02) are supposed to be sequential of one another, but there are 382 sectors that are unaccounted for. Could this be where Cell (3/01) is, but hidden??? Is this why the sub are running late.? Does the video and audio get corrected down 382 sectors, but the subs don't, so they start 382 sectors late???

But how do you explain the difference between Cells (3/25) and (4/01)??? The difference is 319 sectors. Is that normal for that type of gap for a control cell that is only 15 frames long?

Do I need to create a dummy cell called (3/01), and if so, how do I make it 382 sectors big?

Thanks,
Stiff

[StifflerStealth]

I had to delete all the demuxed files to concentrate on the DTS problem of the other DVD I posted about. Now that I have that cleared out, I re-demuxed the DVD using CellID, again. Fun! Some questions that I have in the above post can be answered. I noticed a video file for cells (3/01) and (3/26). Thinking that these were just fillers for that sector discrepancy I mentioned. However, I noticed that the audio file for cell (3/26) was delay corrected, So I open the video file, and it played. I opened up (3/01), and it played. There is actually valid video in those two cells. They're just a logo, but valid. The length of the video of (3/01) is the exact time that the Subs are out of sync. Strange thing is that these cells are not mentioned in any of the IFOs from the DVD. It's like they included the video but forgot all about it. Now all I need to do is import these cells, and tell Scanarist to forget about them. Which I have no idea how to do. I need the $100,000 dollar software that the professionals use. :P

Stiff

[influenza]

Well exactly what I thought. STrange thing for sure. I think you can compile and edit the ifo afterwards, set the cell value to 0 or so. Just play a bit with it.

[StifflerStealth]

I reproduced the exact layout of the disc in DVD Maestro. I created three Playlists for the 3 different PGCs. However, DVD Maestro creates a default PGC that list every cell including Cells (3/01) and (3/26), So I can't use ATM mode because there are two more cells than the original. The other PGCs are spot on. There is one minor problem, Maestro puts all the cells into one Vob ID, so when I use IFOUpdate in DVD Maesto Mode I get a new IFO file that has all the cells in one Vob ID. The DVD refuses to Play like that. For the Audio tracks, I created a Sync audio track selecting the correct audio file for each cell. Is there any script that I can use to either a) delete the Mastro created PGC from the IFO so I can use Adjusted Cell Mode, or b) tell Maestro to create Vob IDs for the cells instead of putting them in one Vob ID so I can use IFOUpdate in DVD Maestro Mode? I seaarched the entire forum for help, but no satisfactory results for a DVD in my condition. There must be a way because Pros use this software to make DVDs for sale and the DVDs have multiple Vob IDs.

Thanks,
Stiff

PS. You can see how long I've been working on this by the Delay time of my post.

[influenza]

I can't help you much with maestro. I always use scenarist myself

[StifflerStealth]

I could just cry! OK! Can anybody give me a few pointers on how to do this in either DVD Maestro or Sonic Scenarist or any other software program where there is a good functional trial version I can get (some trail versions limit what you can do)?

Is there anybody that whats to write a custom DVD Muxer with me?

Stiff

[D3s7]

Stiffler.

Can you email me that ifo... I'd love to take a look at it

- D3s7

[StifflerStealth]

@D3s7: I sent the IFO files to the address on the ScenAid website. The header is "The IFO files you wanted from StifflerStealth".

I thought I had this thing backed up yesterday, but there was a minor problem with the audio. I demuxed by VobID, imported the files manually into Scenarist, and then added the files so that Episode 1 was PGC1 vobid 1, Episode 2 was PGC2 vobid2... and so on. Rember that PGC1 had vobid 3, so I reordered things a bit. I compiled the DVD and everything from audio to subs were in sync, but the audio for episode 2 refused to play until the chapter 2. I played the audio files for episode 2 and it was all there. I guess the audio is used to being first in the set of VOB files and now it's in second place.

These episode DVDs are expensive. That's why I want to create high quality backups to play instead of the original DVD. None of the transcoders gives me the quality of CCE, and I can say that I tried every single one including the Sneak Peak at CloneDVD 2 (it's free so I figured what the hey). However, I have not tried ReJig yet, and since DIF4U supports it I might try it.

Stiff

[influenza]

ok: comapre the ifos with each other. The original and the newly authored one (see if the stream values for the pgcs are the same under pg_city). If not adjust them

[D3s7]

I may know why your scenes are off slightly with the ScenAid script. - I'll check the code tonight

And I'm thinking maybe RA has the same problem. I thought I took this kind of scenario in mind when building the layouts however, it's possible both apps are having trouble finding the "start" position for cell 2 because cell 1 is missing.

But by looking at the layout, I see no reason a PGC backup of this wouldn't work - no vobid's are reused except 4 which is layerbreak and doesn't matter anyway.

Might be easier to go back to that method and if it still holds true, just adjust the .sst files 3-4 sec back :)

[StifflerStealth]

@ D3s7: There are nearly a thousand Subs for PGC 1 times 2 for two Languages.

There is another issue I noticed, and I don't know which program is at fault. The re-encoded MPEG2 files end up being a little over two seconds too long in any demux I do with this DVD. I found this out when I did a CellID demux. I loaded the original Cell (1/01) in DVD Maestro (I've been using Scenarist and Maestro to figure this out) and then I loaded the reauthored Cell (1/01) into Maestro. Then I noticed the time difference. I used a program I wrote to calculate the correct number of frames, and loaded that new number into BatchCCE. The new, new Cell (1/01) was the correct length, but 2 seconds of the video was missing compared to the next cell.

Stiff

[StifflerStealth]

OK. I finally got this thing muxed together. What I did:

1) Used CellID demux to give me the hidden cells.
2) BatchCCE to re-encode.
3) Used DVD-RB to put it back together again.

That was the only tool I could get this thing together with. I had all the cells, and Ifigured what the hey and tried it. For the life of me, I could not figure out how to get this thing together with Maestro or Scenarist and keep everything in sync. The only way to do that would be to manually edit the subs or audio files to get them in sync like D3s7 suggested. I think I will try his new version of ScenAid on this DVD to see if it work when it comes out. I sent the IFO files to him like he wanted, so I am hoping that helped.

Thanks for all your help! This is a wonderful, freindly community.
Stiff

[influenza]

Out of curiosity: did you have to rename all cells to get dvd-rb to rebuild them? That must have been some work :D

[Trahald]

Could you PM me with the DVD name.. i'm always up for a challenge.

Been dissapointed so far. :(

[StifflerStealth]

Originally posted by influenza
Out of curiosity: did you have to rename all cells to get dvd-rb to rebuild them? That must have been some work :D

Yes, I did, and that was a lot of cells, but it worked.


@Trahald: I sent you that PM.

Stiff