Previewer AR Info Bug w 16:9 P&S still frames?

[Taelon]

One of the last stages when I'm reauthoring a dvd is to modify the menu backgrounds to eliminate the text in areas where buttons have been removed. The process I use is similar to the one outlined in goonix's sticky thread (http://forum.doom9.org/showthread.php?t=99306), the procedure seems daunting at first but I can now customize 4:3 AR menus quickly and easily, it's only when 16:9 menus are encountered that things become a little dicey. I've been using PgcEdits Menu Editor to examine the 'groups' to determine which substream is for WS or LB, for setting Muxman's subpic assignments. But this latest compilation had me momentarily stumped because of conflicting reports as to the nature of the still-frame cells Aspect Ratios.

The dvd in question is "Crash" Region 1, the Menu's appear to be 16:9 when viewed in a software player. However the Previewer in PgcEdit reports the Static Cells as being 4:3 AR, but checking the Cell's Domain Attributes reports them as 16:9 Automatic Pan & Scan.

To further complicate things, when I checked the Cells in question in VobBlanker, some still-frame cells would show as 4:3 at all times, another one said it was 4:3 unless it was played, then it would report as being 16:9. Feeling somewhat confused, I previewed the FMV Cells, and they were reported as 16:9 in both programs. Knowing that all the VTSM1 Cells needed to use the same AR, I concluded that PgcEdit's Domain Attributes listing had to be correct, so I set up Muxman as 16:9 P&S with subpic 1 as Wide and 2 as PS.

The final result appears to be correct the button alignment of both groups looks good and the newly muxed Cells are listed in 16:9 in both programs. But felt I should report the inconsistencies in the Info reported by the Previewers in PgcEdit and VobBlanker. I also noticed that often when previewing Menu Cells in VobBlanker, I'd have to hit the Info button twice in order to get the Information window.

[r0lZ]

Thanks for submitting this interesting problem.

As you know, the AR of the domain is defined in the VOBs and in the IFOs. The definition in the IFOs takes precedence.

For the menu groups, the value at offset 0x9B in the nav pack describes the AR supported by each group. Strangely enough, it is possible to combine several aspect ratios for the same group. For example, I've seen some groups with both the letterbox AND pan&scan bits sets!

In PgcEdit's menu buttons viewer, I use the AR defined in the nav pack for the group. It should be compatible with the domain streams attributes.
In the 'special' case of a combined display mode, I look at the domain attributes, and I try to choose the best display mode. For example, in the case of the letterbox + pan&scan definition, I use the mode defined in the domain streams attributes. Of course, if both letterbox and pan&scan are defined in the domain streams attributes and in the menu group, there is a problem. But I have never encountered this case.

In the case you reported, I suppose there is an authoring error somewhere. Maybe a group is defined for 4:3, while the streams attributes are 16:9. Can you verify that with VobEdit?

[Taelon]

r0lZ, always happy to find strange and confusing exceptions to the rules ;) Since I have a hard time finding things in VobEdit, I extracted the static cells vob 9 & 10 with VobBlanker along with the cells preceding and following, vob 8 & 11 as seen here.

http://img400.imageshack.us/img400/2764/pgc15164qk.th.gif (http://img400.imageshack.us/my.php?image=pgc15164qk.gif)

Here is (hopefully) the information you requested from VobEdit:PGC 14 VOB ID 8, Duration 21.00 sec

[0000] MPEG Sequence Header start code 435 [000001b3]
[0004] Video attributes 755097652 [2d01e034]
Video width: 720
Video height: 480
aspect ratio code: 3
16:9
frame rate code: 4
30000/1001 (29.97)
bit rate: 360000
VBV buffer size: 112
constrained parameters flag: 0
load intra quantiser matrix: 1
load non-intra quantiser matrix: 1

PGC 15 VOB ID 9, Duration 00:12 sec

[0000] MPEG Sequence Header start code 435 [000001b3]
[0004] Video attributes 755097636 [2d01e024]
Video width: 720
Video height: 480
aspect ratio code: 2
4:3
frame rate code: 4
30000/1001 (29.97)
bit rate: 320000
VBV buffer size: 112
constrained parameters flag: 0
load intra quantiser matrix: 1
load non-intra quantiser matrix: 1

PGC 16 VOB ID 10, Duration 00:12 sec

[0000] MPEG Sequence Header start code 435 [000001b3]
[0004] Video attributes 755097636 [2d01e024]
Video width: 720
Video height: 480
aspect ratio code: 2
4:3
frame rate code: 4
30000/1001 (29.97)
bit rate: 320000
VBV buffer size: 112
constrained parameters flag: 0
load intra quantiser matrix: 1
load non-intra quantiser matrix: 1

PGC 17 VOB ID 11, Duration 03:00 sec

[0000] MPEG Sequence Header start code 435 [000001b3]
[0004] Video attributes 755097652 [2d01e034]
Video width: 720
Video height: 480
aspect ratio code: 3
16:9
frame rate code: 4
30000/1001 (29.97)
bit rate: 360000
VBV buffer size: 112
constrained parameters flag: 0
load intra quantiser matrix: 1
load non-intra quantiser matrix: 1 As you can see, the still frame cells are listed as 4:3, the motion cells as 16:9 in VobEdit. Here's a composite screenshot from PgcEdit, PGC 15

http://img400.imageshack.us/img400/3527/pgc157yk.th.gif (http://img400.imageshack.us/my.php?image=pgc157yk.gif)

[r0lZ]

Obviously, this VOB is funny!

In the second image, I see the preview info window. The preview doesn't check the IFOs, and retrieve the AR in the VOB. But, as I said, the information in the IFO takes precedence.

But I would like to know which is the value known as "nibbles which describe grouping of buttons" in VobEdit, af offset [009b], in the nav pack. To see it, you must enable the "button" checkbox. Check a menu with buttons, or this value will probably be 0.

[Taelon]

Ok here's the info from VOB 9
[Pack Header]
[0000] Pack identifier (start code) 442 [000001ba]
[0004] SCR (System clock reference) 68 0 4 0 4 1 [44 00 04 00 04 01 ]
SCR 00000000.000
[000a] Program Mux Rate: 25200 (1260000 BPS) (10080000 bps) 1 137 195 [01 89 c3 ]
[000d] Pack stuffing length: 0 248 [f8]

[Private Stream 2]
[0026] Private Stream 2 start code 447 [000001bf]
[002a] Length 980 [03d4]
[002c] Sub-stream ID 0 [00]

[PCI Stream]
PCI General Information
[002d] this->lba 104408 [000197d8]
[0031] Macrovision (APS) 0 [0000]
[0033] Reserved 0 [0000]
[0035] Prohibited U-OP 0 [00000000]
[0039] Vobu Start Presentation Time (90KHz clk) 21021 [0000521d]
[003d] Vobu End Presentation Time (PTM) 57057 [0000dee1]
[0041] End PTM of VOBU if Sequence_End_Code 24024 [00005dd8]
[0045] Cell elapsed time (BCD) 192 [000000c0]
00:00:00.00 / 30 fps
[0049] International Standard Recording Code 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 [00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ]

Button Highlight Information
[008d] Highlight status 1 [0001]
[008f] Highlight start time (PTM) 21021 [0000521d]
[0093] Highlight end time (PTM) -1 [ffffffff]
[0097] Button selection end time -1 [ffffffff]
[009b] nibbles which describe grouping of buttons 8512 [2140]
number of button groups: 2
button group 1 type (normal,lb,p/s): 1
button group 2 type (normal,lb,p/s): 4
button group 3 type (normal,lb,p/s): 0
[009d] starting button number 0 [00]
[009e] number of buttons 6 [06]
[009f] number of numericaly selected buttons 6 [06]
[00a0] Reserved 0 [00]
[00a1] force select button number 0 [00]
[00a2] force action button number 0 [00]

[r0lZ]

It confirms that your menu is in 16:9, with the Pan&Scan option.
So, everything is fine, except that strange 4:3 aspect ratio of some VOB cells.
I suppose that the cells have been authored separately, with a wrong AR parameter for some of them, and then joined together with a software that didn't take care of the AR. No big deal.

BTW, this problem can be fixed on the fly by DVDPatcher.

[jsoto]

I've found more than once AR mismatches between IFO and VOBs, but it is the first time I see the mismatch in only one cell.

For the record (sorry if it is a little bit OT) VobBlanker checks the AR of the VOB (not the one in the replacing IFO) in the replacing PGCs and warns if it is different to the one in the original IFO. But it only checks the first cell!
Up to now, it does not check the AR replacing cells...

jsoto