How to pick LB position on multi-angle title?

[fordman]

OK, I've ran into a seemingly tough one that PGCEdit choked on (I always run through PGCEdit before building with ImgBurn), but ImgBurn happily offers to allow selection of a layer break.

 

In PGCEdit, the title layout where the split needs to be is listed in the first attached image below. You can see that the current layer break appears to be at cell 39, and the second angle is contained in cell 40, and has a selected, but grayed out LB flag.

 

When I choose to create an ISO in PGCEdit, I receive the warning contained in the second attached image below.

 

So, since I didn't understand that exactly, I simply tried to build the image in ImgBurn and received the dialog presented in the third attached image below.

 

What's especially confusing about the selection of cells 39 or 40 is that it would appear to leave one angle on one side of the disc and the other angle on the other side of the disc.

 

Also, shouldn't the cell times show in ImgBurn be the same for cells 39 and 40, as they are in PGCEdit?

 

So, can I make a valid image with ImgBurn 2.1, or do I need to somehow make further adjustments?

 

Thanks,

Ford Man

Edited September 19, 2006 by fordman

[blutach]

What's especially confusing about the selection of cells 39 or 40 is that it would appear to leave one angle on one side of the disc and the other angle on the other side of the disc.

The LB is placed just before the selected cell that is marked as non-seamless.

 

Not sure why PgcEdit complains. Perhaps r0lZ can help over at doom9 or Digital digest. Perhaps you have done some editing that has thrown things out? But I doubt it.

 

As for the times in ImgBurn, they are obviously times 2. In fact, ImgBurn's time are supposed to show how much is elapsed at the LB, not how much is elapsed after the LB cell is played (so it will differ from PgcEdit by 1 cell normally). However, for angles, things look a bit awry - it appears that we need to divide by the number of angles for angle 1 and this will make angle 2 correct (it would be advanced by 3:20 being the playing time of angle 1).

 

Good find

 

Regards

[fordman]

The LB is placed just before the selected cell that is marked as non-seamless.

 

I realized this morning that this would be the case, so it appears that each angle for a particular cell would indeed be co-located on a single side of the disc.

 

Not sure why PgcEdit complains. Perhaps r0lZ can help over at doom9 or Digital digest. Perhaps you have done some editing that has thrown things out? But I doubt it.

Actually, yes, some content was removed and re-authored with DVD Rebuilder. However, I thought that through the use of padding, as both ImgBurn and PGCEdit do, a valid layer break could be made. Also, the disc is nowhere near the 8.5 GB limit, so there's plenty of room for padding. The entire title is shot in in two angles - first person view and a "making of" view, so each and every cell in angle 1 has a corresponding cell for angle 2 with the exact same time...

 

As for the times in ImgBurn, they are obviously times 2. In fact, ImgBurn's time are supposed to show how much is elapsed at the LB, not how much is elapsed after the LB cell is played (so it will differ from PgcEdit by 1 cell normally). However, for angles, things look a bit awry - it appears that we need to divide by the number of angles for angle 1 and this will make angle 2 correct (it would be advanced by 3:20 being the playing time of angle 1).

 

Good find

 

So, despite the issue with ImgBurn's times and angles, will ImgBurn 2.1.0.0 still make an image with a valid layer break point, or will this time miscalculation affect ImgBurn's ability to write a valid DL image? I had assumed that PGCEdit was the one with the problem, but perhaps ImgBurn's miscalculations are what make it appear that it found a valid layer break point?

Edited September 19, 2006 by fordman

[LIGHTNING UK!]

Nah, the time means nothing. It's purely cosmetic.

 

The sector number is what's important and so you'd just select cell 39 and be done with it.

 

I will work on fixing the time calculation for PGC's with cells for different angles

[fordman]

Nah, the time means nothing. It's purely cosmetic.

 

The sector number is what's important and so you'd just select cell 39 and be done with it.

 

I will work on fixing the time calculation for PGC's with cells for different angles

 

Great - thanks for the confirmation. Perhaps I should point r0lZ to this thread so that he can investigate his code...

[LIGHTNING UK!]

Blu, assuming you come back to read this, should I even be listing anything other than the cell of the 1st angle in a block for a potential LB position?

 

If all cells in an angle block are supposed to be on the same layer / marked in the same way with the SPLIP flag, I would guess not.

[fordman]

Blu, assuming you come back to read this, should I even be listing anything other than the cell of the 1st angle in a block for a potential LB position?

 

If all cells in an angle block are supposed to be on the same layer / marked in the same way with the SPLIP flag, I would guess not.

 

not to hijaak your message to blutach, but I think you are correct. I believe that's why PGCEdit showed cell 40 in my example as having a LB flag, but it was a grayed out check mark. I didn't try to uncheck the LB box for cell 39, but I assume if I had, the grayed out one for cell 40 would have disappeared also.

[blutach]

Hijack away fordman - everyone else here does AND it's your bloomin' thread!

 

@LUK - I'd list only angle 1. I am not at all sure of the spec on this. It is feasible that the LB could be before angle 2 couldn't it? Angle switching would not be at all seamless however and that would be a nasty drawback. So, I'd list only the first angle. Easier too.

 

Regards

[LIGHTNING UK!]

I think it makes sense to always have the LB at the start of an angle block and not in the middle, and duplicate the chosen LB type (seamless etc) to all cells in that block.

 

If you put it in the middle (and didn't duplicate), all cells around it would be marked seamless. Whilst that would be correct for cells before the one you've selected, angles after it (in same block) would be wrong because they WOULD be non seamless. I guess the same then goes for the first cell played by the route taken by angles before the LB cell. That first cell should also probably be marked non seamless.

 

So yeah, that complicates things far beyond what it needs to! Much easier / better to just display the first cell of angle blocks and automatically configure all cells in that block with identical SPLIP flags.

 

(That's what I've now done)

 

Do you know of any good DVD's I could test this on?! I can't remember many (any!) with angles on them throughout the PGC like in fordman's example.

[blutach]

The Matrix 1 has loads of angles and seamless branching.

Jim Taylor's test DVD (it has everything imaginable on it).

 

But neither has angles all the way through a title, I don't think.

 

Regards

[r0lZ]

The PgcEdit method is somewhat limited when there are angles in the PGC. Currently, an angle cell cannot be selected as the layer break position. However, it's perfectly legal. The problem is only due to the limitation in PgcEdit, and the fact that the whole PGC is made of angle cells. Sorry for that.

 

I agree with LUK and blutach: only the angle 1 cell should be selectable, although all angles must be marked as non-seamless. It is absurd to select angle 2, since angle 1 will be cut after its first VOBU.

[fordman]

The PgcEdit method is somewhat limited when there are angles in the PGC. Currently, an angle cell cannot be selected as the layer break position. However, it's perfectly legal. The problem is only due to the limitation in PgcEdit, and the fact that the whole PGC is made of angle cells. Sorry for that.

 

I agree with LUK and blutach: only the angle 1 cell should be selectable, although all angles must be marked as non-seamless. It is absurd to select angle 2, since angle 1 will be cut after its first VOBU.

 

Thanks for the confirmation. I saw you released 7.4 recently and I tried it and received the same message, so you're explanation covers that.

 

I selected cell 39 and ImgBurn 2.1.0.0 made a perfect DL image which switched exactly where it needed to.

 

Thanks again,

 

fordman