A Simple way to deal with angles for a DVD9-to-2xDVD5 split - different, English, backing, current

andy

I just got finished backing up Howl's Moving Castle. It is your typical Miyazaki DVD with different angles for English and Japanese versions of cells that have some type of lettering in them -- typically shop signs or movie title. All I did was run a trace of the main movie in PgcEdit to see which cells got played for the English version where there were two adjacent cells with indicated ILVUs. e.g., 2 and 3, 4 and 5, and 29 and 30 in the current title. It turns out that 2, 4, and 29 were played and 3, 5, and 30 were skipped (these get played and the other ones skipped for the Japanese version). Then I blanked the skipped cells in VobBlanker. (I did a DVD9 to 2 x DVD5 split so no complaints from the recoding crowd please.) VobBlanker de-interleaves ILVU cells so they (usually) play normally. Then in PgcEdit I simply added cell commands for cells 2, 4, and 29 to jump to the next unblanked cell. e.g., LinkCN Cell 4 for cells 2 and 3 (in case someone jumps to cell 3). The cell commands get executed after the cell they are associated with get played so the progran jumps to the next playable cell with only a split-second pause. This is barely noticeable even if you know where it occurs and are looking for it. To the average viewer it is not even noticeable. If you don't do this there will be a more noticeable (about one second) pause and a blank cell displayed. This works quite well in my set-top box and is the only way I have found to deal with angles that retains all the menu structures including proper scene numbering. All other methods that have been mentioned in the forums have had problems, including misnumbering of scenes. This is a very simple method if you want to preserve the whole DVD uncompressed.

Interesting.  

Have you tried to delete the non played and blanked cells (so that they are unreferenced), and then remove them completely from the VOBs with FixVTS or VobBlanker?  This technique should be sufficient to avoid the cell commands, and to remove the pauses, at least almost completely.
Note that you will have to blank the unplayed cells before removing them in PgcEdit, to force VobBlanker to de-interleave the cells.

Not sure if it works fine, as I have never tried this method.


Originally Posted by r0lZHave you tried to delete the non played and blanked cells (so that they are unreferenced), and then remove them completely from the VOBs with FixVTS or VobBlanker?  This technique should be sufficient to avoid the cell commands, and to remove the pauses, at least almost completely.
Note that you will have to blank the unplayed cells before removing them in PgcEdit, to force VobBlanker to de-interleave the cells.

Not sure if it works fine, as I have never tried this method.

This sounds interesting. How would you do this exactly? Would you blank the unwanted cells first in VB, then quot;Remove Cellsquot; in PgcEdit, then run FixVTS on this? Can you give explicit instructions. I would like to try this and report back on it.


Originally Posted by foobazWould you blank the unwanted cells first in VB, then quot;Remove Cellsquot; in PgcEdit, then run FixVTS on this?

Exactly!  Do you really need more instructions?

I'm not sure if you must set or clear the seamless joint flags after the remove cells.  PgcEdit will probably clear them when the cells are removed, but since the original cells are played seamlessly, I think it should be OK to set them back.  Try with a RW on your standalone and let me know.

What I usually do is to remove the angle flags and the unwanted angle cells in the PGC (I've a small tool to do this in the IFO, but it can be manually done with pgcedit). Finally run VobBlanker: Unwanted angle cells (now unreferenced) will be removed.

Ah!. Be careful with the program/chapter table

jsoto

Jsoto, with your method, I wonder if the first ILVU end values are automatically cleared by VB.  There is no way to do it manually in PgcEdit.

Yes, they are. It is part of the de-ILV process...
jsoto

Oh, great! So your method is indeed the fastest and easiest one.


Originally Posted by jsotoWhat I usually do is to remove the angle flags and the unwanted angle cells in the PGC (I've a small tool to do this in the IFO, but it can be manually done with pgcedit). Finally run VobBlanker: Unwanted angle cells (now unreferenced) will be removed.

Ah!. Be careful with the program/chapter table

jsoto

Could you explain this in more detail? So far I have tried removing already blanked cells with PgcEdit with disastrous results, even after processing with FixVTS 1.24. The movie crashes the software player after the first blanked cell. How do you adjust the chapter table? How do you reprocess in VobBlanker? I could find nothing about this here or in other forums.

OK, I've improved my little tool to modify the programs table, so the chapters should be OK.

Try this:
del_angles

How to use:
- Select your angle (use VobBlanker or a SW player to detect which angle do you want)
- Run del_angles keeping your angle.
- Open VIDEO_TS.IFO in IFOedit and modify the number of angles of the PGC to 1. (VMG_PTT_SRPT table). I'll do it automatically in a new version, but in the meantime...
- Run VobBlanker in KEEP (default) mode, but be sure the quot;Fix always VOB SYNCI/SRI pointers in Titlesquot; option is active.
- Run pgcEdit, find the SETSTN angle commands and delete them.

jsoto
PS: I didn't test this tool too much.... Even more, I've added just now the IFOs backup and the program table modification... I hope it works fine, but be careful..


Originally Posted by jsotoOK, I've improved my little tool to modify the programs table, so the chapters should be OK.

Hey thanks jsoto for making this available. I followed your instructions exactly and everything worked flawlessly. The unwanted angle was completely gone including cell references. It played fine in the software player, however, the kept angle cells played very choppily in the set-top player, similar to the problem I had before (showthread.php?t=103233 ). Something to do with the de-interleaved cells and my player I suspect. I didn't have this problem when I did it as outlined in my original post on this thread.

Ah! Forgot to mention....
Cell internal pointers have to be fixed, because now they have to be different (the removed angles...). May be this is the problem you have (or may be not)

So please check
quot;Fix always VOB SYNCI/SRI pointers in Titlesquot;
run VobBlanker again and let me know...

I've updated my previous post..

BTW, use 2.1.1.0_beta8. There is an ugly error fixing subs pointers in previous versions (not too important, but can cause VobBlanker to hang)

jsoto


Originally Posted by jsotoAh! Forgot to mention....
Cell internal pointers have to be fixed, because now they have to be different (the removed angles...). May be this is the problem you have (or may be not)

So please check
quot;Fix always VOB SYNCI/SRI pointers in Titlesquot;
run VobBlanker again and let me know...

OK, I tried this with 2.1.0.0 (couldn't find the beta below). This worked like a charm. Plays in the set-top without a glitch. Menus and chapter numbering are preserved. This is definitely my new tool of choice for processing angles. Kudos! jsoto, once again. You are the master.  

BTW, use 2.1.1.0_beta8. There is an ugly error fixing subs pointers in previous versions (not too important, but can cause VobBlanker to hang)
jsoto
Great!.
When I find some free time I'll finish completely the tool and put it in my web space...

Beta 8:
showthread.ph...223#post799223

jsoto

Hey jsoto,

Dunno where you are at with this, but tried it on Clerks X (Tenth Anniversary R1), Disc 2.

Got an error: Ifo file too big.

It's size was 284,672 -- it is kinda big, isn't it?  FYI...

Ooops, try again
del_angles

jsoto

Hi jsoto,

Gave this a whirl, although you had to make me work for that download, didn't ya?

I had to wing this a bit.  This IFO was structured a little differently (and I understand why), so I couldn't follow your instructions verbatim.

There's 2 titles/2 PGCs in the IFO. Title 1/PGC 1/Angle 1 is movie, Title 2/PGC 2/angle 2 is video commentary. Both titles share audio track 2 (commentary), so you can watch the movie with audio commentary, or you can watch the video commentary [with audio commentary, of course].

PGC 1 = PGC 2.

I dropped the IFO in your tool and selected PGC 1/angle 1 and let it do its thing.  Then I dropped the modified IFO in your tool and selected PGC 2/angle 1.  This cleared out all of the unwanted cells so that VobBlanker could see the unreferenced material.  Looks like it worked out -- I didn't test this in a hardware player yet, though.

Also, I didn't delete the SETSTN angle commands, but I don't think it's necessary?  The DVD is designed to play either title 1 or title 2 exclusively based on user selection, so it doesn't really do any type of dynamic switching midstream during viewing.  Does that make sense?

Thanks...


Originally Posted by JFergusonHi jsoto,
PGC 1 = PGC 2.
I dropped the IFO in your tool and selected PGC 1/angle 1 and let it do its thing.  Then I dropped the modified IFO in your tool and selected PGC 2/angle 1.  This cleared out all of the unwanted cells so that VobBlanker could see the unreferenced material.  Looks like it worked out -- I didn't test this in a hardware player yet, though.

That's correct.

Also, I didn't delete the SETSTN angle commands, but I don't think it's necessary?  The DVD is designed to play either title 1 or title 2 exclusively based on user selection, so it doesn't really do any type of dynamic switching midstream during viewing.  Does that make sense?
Well, the settop usually ignores the SETSTN angle commands if there are no angles... I think that deleting these commands is only needed to have a quot;perfect standardquot; DVD

jsoto

The del_angles link is dead...?

BTW, I have just finished coding a new function to remove the angles in PgcEdit.  To strip them from the VOBs, you will have to process the domain with VobBlanker, with the option to fix the pointers in the VOB files.  Works well.
The next release of PgcEdit is almost ready...