DVD-RB v1.26.2 - Missing VTS

[r3m1gus]

'lo all,

I recently set up dvd-rb to re-encode an NTSC disc of X-Men and ran into a problem. DVD-RB is not seeing some of the VTS's on the disc.

At first I thought it was a mistake I made re-authoring the disc before running it through dvd-rb. I cut a lot of extras, including all the BOV's and moved the deleted scenes from vtsm's to vtst's. I think that in order to support BOV, RSM, and multiple audio tracks for those deleted scenes the original authors set up two VTS's with the same deleted scenes but one a normal audio and on the other a director's commentary. Since I didn't really need them to be interactive, it saved me ~300MB's to simple create a title section with two audio tracks.

Anyway, I thought that the rather drastic pre-auth was messing things up, but when I tried to load the original ISO into dvd-rb, I ran into the same problem. I tried lowering vts_min_size to 1, and though that allowed me to see more vts's on the original ISO it didn't help on my problem. Incidentally, doesn't dvd-rb load vts's with small title sections and large menus now (1.26.2)?

I did come across this post on the extact same issue with an older version on dvd-rb:

http://forum.doom9.org/showthread.php?t=106021


There is some reference in this post to dvd-rb not encoding or loading menu VOB's, but jdobbs has long added that function to the util, and the vts that's not loading on my pre-authed version has a timy menu and 300+MB title VOB. Is it possible there's a reference to file size in the IFO that's preventing dvd-rb from loading the vts? I created the deleted scenes vts with a combination of pgcdemux, dvd-lab pro, and pgcedit.

It's not that big a deal as I can certainly encode the main vts with dvd-rb, calculate how much space to leave free and then import the titles and menu of the deleted scene vts w/o encoding them in pgcedit for a bit of post-authoring. But it'd be nice to steal a lil' more space for the main vts.

The file listing of my VIDEO_TS pre_auth dir:

20,480 VIDEO_TS.BUP
20,480 VIDEO_TS.IFO
686,080 VIDEO_TS.VOB
83,968 VTS_01_0.BUP
83,968 VTS_01_0.IFO
299,141,120 VTS_01_0.VOB
1,073,049,600 VTS_01_1.VOB
1,073,489,920 VTS_01_2.VOB
1,073,358,848 VTS_01_3.VOB
1,073,608,704 VTS_01_4.VOB
374,116,352 VTS_01_5.VOB
26,624 VTS_02_0.BUP
26,624 VTS_02_0.IFO
141,312 VTS_02_0.VOB
406,921,216 VTS_02_1.VOB


Thanks in advance to everybody for putting up with my blathering and really rather minor issue.

[blutach]

First, the latest version 1.26.5; you should upgrade.

Secondly, the sizes of those VTSs will be seen by DVDRB in its standard INI configuartion (except for the small VTSM 2). To see all menus, set a hidden setting VTSM_MIN_SIZE=1

Regards

[r3m1gus]

Thanks for the reply.

I set VTSM_MIN_SIZE to 1 in the INI and dvd-rb is still not loading certain VTS's from either my pre-auth'd version or the original ISO.

I'll get the latest version and see if that helps.

[r3m1gus]

OK. upgraded to 1.26.5 - I should have done that first - and I"m still having the same problem.

I did run through an encode at a lower bitrate so I could re-add the missing vts later, so I guess that's my next step. There is one oddity in the encoded vts set. VIDEO_TS.IFO contains the title play maps for all eight titles in my pre-auth'd version. Titles 2-8, however, which should be in the deleted scene vts that was not encoded, are all duplicates of title 1 (the main title). Also VIDEO_TS.IFO does have the proper info for the deleted scenes title set, even if it's not there.

If anybody has any thoughts on this I'd appreciate a response, though I think I'll be able to muddle through regardless.

Thanks!

[jdobbs]

That means the VTS contained unreferenced data (it is there just to take up space), sometimes it is a repeat of video found elsewhere on the DVD... if that's the case -- you should leave it out.

A suggestion: Let DVD-RB do its job, and don't second-guess it. It works.

[blutach]

IIRC, one of the series of Little Britain was authored this way with what were apparently deleted scenes authored as unreferenced material (shame that, as they were quite funny).

Check that the titles are actually called. Use PgcEdit's Info - Find Uncalled PGCs function. Note this is not 100% accurate as a PGC may have a Jump to title 2 but that PGC may be uncalled except by the title to which it is referring. But, more than likely, it will confirm jdobbs' diagnosis.

Regards

[r3m1gus]

Thanks to both of you for your replies.

@jdobbs
I wouldn't presume to second guess dvd-rb. I always used the big3 for my backups and continually lamented their inability to handle ILVU segmnents, not to mention the occasional crashes of Scenarist. When I went hunting and found that dvd-rb supported seemless branching I was interested and I've been impressed with it ever since.

@blutach
I believe that your suspician is indeed accurate.

I went back and opened the studio authored ISO w/ pgcedit and scanned it. There were 69uncalled pgc's before a bov scan and 9 after it. Among those 9 were the RootM and spaceholder titles in the VTS's that weren't appearing in dvd-rb. However, the BOV's called pgc's in the VMGM, which set a gprm and then called the angle menu's of the offending vts's. They in turn used LinkPGCN to call each of the Menu pgc's that contained the deleted and behind-the-scenes content. So, I'm still not entirely sure whether dvd-rb is ignoring those vts's because they're initially called by BOV's, their Root Menu's and Titles are unreferenced, the angle menu's utilize LinkPGCN or something else.

In any case, I went back to my pre-authored version, removed the gprm sets and Jump_VTS_TT if statements from my offending VTS, had the menu post cmd's call a new vmgm dummy pgc I created, and used my set gprm statements with Jump_TT if/then's instead. 'lo and behold, dvd-rb now sees my recalcitrant VTS.

Of course, I'd already encoded the main VTS, imported the deleted scene VTS and fixed the title map table, but it's always nice to get a solution.

Thanks so much for your help.

[blutach]

Hmmm...

After the BOV scan, it sounds like the VTSs were not uncalled. You might try a trace in PgcEdit (after scanning for BOVs) to see if you can actually get to the offending VTSs (or just play the DVD and see).

Regards

[r3m1gus]

Both the original ISO and my pre-auth'd version (even before I altered the TTN jumps to title jumps) played everything (WinDVD 4) and traced fine...

Some folks seem to frown on pre-authing, prefering to use vob-blanker or some other tool to get rid of extra data w/o really messing with dvd cmd structure. But I hate having blank vts's, titles, and audio/sub tracks tagging along so I try to strip everything I don't want completely out. As a result, I have to pay very close attention to pgcedit traces as it's far easier to wreck a cmd structure when stripping as opposed to blanking.

I suppose it's a moot point anyway as I eliminated the problem, but I certainly am curious about it.

Thanks again and please let me know if you think of any other way I can investigate this, it'd be nice to know what the deal is.

[blutach]

Well, if that's the case that it traced to the supposedly unreferenced PGCs, jdobbs will wanna look into this further, I expect. I can say that except for small VTSs, in my experience, DVDRB has never omitted a titleset.

Regards

[r3m1gus]

I can say that except for small VTSs, in my experience, DVDRB has never omitted a titleset.

Yup I've certainly never seen it either. I do seem to detect a hint of doubt regarding my claim. It's perfectly natural, after all I don't post here often and you have no reason to think I'm anything more than an idiot. In fact, it's entirely possible that I am an idiot. Still I can't see any other road than to try and prove that something here is behaving oddly. So I guess I'll walk through what I've found and try to provide as much proof as possible. Incidentally this will only be data from the original studio ISO, not my pre-authored version. All files refered to are in pgcedit_trace.zip.

First a directory listing of the VIDEO_TS folder in dir_listing.txt.
As you can see by the file sizes, VTS's 2,3,5,9 and 10 should all be loaded into dvd-rb assuming default vts(m)min_size settings and that they are not unreferenced material.

And what dvd-rb comes up with in dvd-rb.jpg

Now I've added my Rebuilder.ini to show that I'm using default size settings in this test, proving that those vts's are referenced takes a bit more doing.

So I guess The first thing would be a pgc listing from the offending dvd by pgcedit. Note that this pgc_uncalled.txt also labels those uncalled pgc's that were found after a BOV search.

So next I've got three pics of pgcedit after tracing through the entire dvd clicking every BOV in Title 1 (vts5). As most of you already know, the pgc's highlighted in yellow are those that the trace has covered. As you'll see in vmgm.jpg, vts2_3.jpg and vts9_10.jpg, nearly all of the "unreferenced" material is, in fact, referenced and covered in a trace. The vtsm3 pgc's went untraced because they are called only when the commentary track is active, and I didn't feel like running the trace twice. Here's what the vmgm pgc's 2-7 look like:

1 Set gprm(12) =(mov) sprm(1:Audio stream number)
2 Set gprm(6) =(mov) 10
3 Set gprm(0) =(mov) 4
4 if ( gprm(12) == gprm(0) ) then { (JumpSS) Jump to VTSM 3, Angle menu (TTN 1) }
5 (JumpSS) Jump to VTSM 2, Angle menu (TTN 1)

The only thing that changes between 2-7 is the gprm(6) setting. That number controls what vtsm pgc the angle menu sends the trace to.

During the trace, pgcedit complained twice about it's inability to handle 1 cell that had two BOV's in it. I'm pretty sure that's why vmgm pgc's 2 & 7 are listed as uncalled and why vtsm2 pgc's 3 and 8 are not covered by the trace; their source was the second bov in the cell.

Finally, if you don't believe the jpg's, I've included the trace log (pgcedit_tracelog.txt). It records each of the BOV jumps to the VMGM, from there to the vtsm angle menu's and then to the vtsm pgc's containing content. As trace log's through an entire movie get rather long here are the highlights.

Line 196: Set BOV subtrack ON
Line 229: Select Play
Line 322: Main title actually starts. All the BOV selects and jumps to the relevant vtsm's start after this line.

I hope this is useful. If there's anything else I can provide that would help please let me know. Or if I"m just crazy and barking at the wind, please tell me that as well.

Thanks again.

[blutach]

I do seem to detect a hint of doubt regarding my claim. Not at all. By posting here about a potential situation, you are helping jdobbs and all users of DVDRB and it's obvious you know your stuff.

I'll have a close look at your attached files, but, on first blush, they seem verrrry unusual - the directory listing itself, with all those huge menus in several VTSs, some of which seem identical, would perhaps point to an advanced protection scheme and maybe a rip that wasn't quite 100%.

In any event, even if the PGCs are called, doesn't mean the cells are played (the PGCs might only be entered at their precommands and jump to other PGCs) - I'll take a closer look at your trace.

In the meantime, ensure enable menu encoding is ticked on your settings.

Regards

[r3m1gus]

Not at all. By posting here about a potential situation, you are helping jdobbs and all users of DVDRB.
Well, thanks. I just didn't want to waste anybody's time I guess.
...the directory listing itself, with all those huge menus in several VTSs, some of which seem identical...
No doubt. vtsm's 2 and 3 are both the same video (Deleted scenes) with a 2channel DD track, the have identical pgc's. The audio in vts2 is film audio. Actors acting, doors opening, etc, etc. The audio in vts3, however, is dvd commentary. I think they wanted the deleted scenes to to have both tracks, but I don't think a menu can have dual audio tracks.

As for vts's 9 and 10 and their huge menu's, well, they both contain different vtsm video content and pgc's. My bet here is that they ran out of space -couldn't go over a GB- for behind-the-scenes stuff in VTS_09_0.VOB, but as they couldn't create another one, or increment it to VTS_09_1.VOB, they created VTS_10_0.VOB for the rest.

I also think they wanted the BOV accessible scenes in vtsm pgc's because RSM won't work from a VTST.

the PGCs might only be entered at their precommands and jump to other PGCs
That's true. However, in vts 2, 3, 9, and 10 all the vtsm pgc's with content have only one command, RSM, and it's post. Anyway I think my trace log will show that these pgc's did actually play.
In the meantime, ensure enable menu encoding is ticked on your settings.
Yup, it is :)

I have had a few x-files discs with deleted scene's encoded into the vtsm pgc's but nothing on this great a scale. Please lemme know if you find anything interesting in my files.

[blutach]

First look, it seems like non-seamless branching (via the BOVs), where the video from the branching is in VTSM 9 and 10.

And you're right on all fronts; the menu can't have more than 1Gb nor a single audio track, so that would explain why VTS 2 and 3 were the same (stupid idea really, they could have authored it all in various PGCs of a single VTS since VTS 5 is hardly full).

As for the RSM not being able to be in the titles domain, this could also have been avoided (eg making a chapter point after the BOV, playing the branched content, and then issuing a Jump VTS_PTT).

Either way, it seems this is referenced content so long as the BOVs are in your project.

Regards

[blutach]

Oh yeah - the other thing here (referring to post 1) is that DVDRB does not encode VIDEO_TS.VOB at this stage of its development.

As well, in the options section of the INI, I'd have a VTS_MIN_SIZE as well as a VTSM_MIN_SIZE option. Seems the VTSM_MIN_SIZE by itself is insufficient to recognise the titleset, as you have observed.

Regards

[r3m1gus]

OK, So if 2, 3, 9 and 10 are all referenced content, why isn't dvd-rb loading them? I assume that the BOV's are a part of the the project as I haven't touched that original ISO, and, as you noted, all the weird vts's are traced.

I have set:
VTS_MIN_SIZE=1
VTSM_MIN_SIZE=1
in rebuilder.ini, but dvd-rb still doesn't load certain vts's. See attached pic.

Thoughts?

[blutach]

The titles in those titlesets are not referenced - the titlesets exist to play the menu content! Something for jdobbs to fix, I expect, as it appears the whole titleset is being excluded if there is no played title, even though content is being played.

Try this as an experiment. Using PgcEdit, author in a new uncalled PGC in VMG. Add 4 JumpTT lines: to titles 10, 11, 12 and 13 (these are the titles in VTS 2, 3, 9 and 10). Save and do a prepare and see if DVDRB brings up these VTSs.

Regards

[r3m1gus]

Worked like a charm.

The recalcitrant vts's aren't displayed in the input settings, but they are processed by the "Prepare" phase after adding the title calls.

Thanks, blutach, for spending so much time on this. It's hardly something that can come up very often as it's a ridiculous way to author a dvd.

[blutach]

No problems. jdobbs now has a couple of things highlighted by you if a VTS is to be displayed:

1. The need for VTS_MIN_SIZE to be there as well as VTSM_MIN_SIZE

2. The need to make a VTS unreferenced only if both the menus and title PGCs are uncalled.

I'm sure he'll tend to this ASAP.

Regards

[jdobbs]

Which disc is this, the original X-Men, X-Men United, or X-Men the Final Stand? I'd like to do some testing...

Definitely an unusual situation. But if a menu is referenced -- the titleset should definitely be kept. I'll see what's up.

[jdobbs]

One other option you have if you ever run into a disc like this again, you can disable the removal of unreferenced matrial with this hidden command (in the "[Options]" area of REBUILDER.INI):

DISABLE_UNREFERENCED_REMOVAL=1

[r3m1gus]

Hey jdobbs,

It was an R1 NTSC copy of X-Men (The first one).

And thanks for that option!