Log in

View Full Version : Any good tools for n@vigation pack?


Afrinux
13th December 2005, 12:31
Hi!
I was wondering if there is a tool to verify DVD-Video n@vigation p@cks.
Thanks for helping.

jeanl
14th December 2005, 01:30
What do you mean by "verify"?
VobEdit will let you look at the smallest details in the navpack, if that's what you want to do...
jeanl

Afrinux
14th December 2005, 06:44
What do you mean by "verify"?
VobEdit will let you look at the smallest details in the navpack, if that's what you want to do...
jeanl
Sorry, by 'verify', I meant to check the navpack.
Thank you! I will try VobEdit again.

I was thinking about making a tool which function is only to check the navpacks. If there is some problems in the navpacks, the tool will fix them.
Do you think its a good idea? Is there other features I should add to the tool to make it useful?
Thanks for your help.

jeanl
14th December 2005, 06:47
I'm not sure I understand what you mean by check. It's hard to say, by just looking at a navpack, whether there's something wrong with it. The things that can be wrong are usually pointers to other navpacks, or pointers to other VOBUs related to subpics and audio streams, etc etc...
Jeanl

blutach
14th December 2005, 07:21
The Philips Video Verifier. But it costs $$$

Regards

mpucoder
14th December 2005, 07:54
The Philips Video Verifier. But it costs $$$

And is horribly flawed.

jeanl
14th December 2005, 17:41
mpucoder, can you elaborate on that?
jeanl

frank
14th December 2005, 18:16
IfoEdit can correct wrong navpacks! Do we need more?

mpucoder
14th December 2005, 18:23
Let's just say that the author of the Philips DVD verifier had just as hard a time interpretting the specs as the authoring application writers. There are too many misinterpretations to list, and inconsistencies. Examine the output of the verifier carefully before deciding if the mux is as bad as it looks.
Two examples:
1) It is impossible to make a PAL DVD with no menu video (either VMG or VTSM) error free. You will either get an error that the video standard (PAL v NTSC) does not match, or, if the attributes are set to match, an error for having non-zero attributes without video.
2) It is impossible to get pulldown material to pass error free - like many European programs, the verifier does not understand pulldown and complains about VOBUs which are not a multiple of frames duration, even though the quoted spec is "multiple of fields in duration"
And many, many more. SRI pointers and time maps will probably create thousands of incorrectly diagnosed errors, especially with NTSC (where one second can be either 90000 (an actual second) or 90090 (30 frames, agrees with non-drop timecode) ticks, and Philips gets anal about that even though the elapsed time fields pass inspection)

jeanl
14th December 2005, 18:27
Thanks for these details mpucoder... I suppose that makes it almost useless for making sure muxman is doing the right thing... :(
jeanl

mpucoder
14th December 2005, 18:34
Almost - I checked every error it gave me by hand when creating my test muxes. Now every new compile of MuxMan is tested against those (currently 17) regression test results rather than using the verifier.

jeanl
14th December 2005, 18:35
That sounds like a good way to do it...
jeanl

Afrinux
15th December 2005, 09:38
IfoEdit can correct wrong navpacks! Do we need more?
I have made a DVD menu and couldnt figure out why the HL!s werent displayed.
So I needed a Tool to check where its wrong so I can fix it for myself.
IfoEdit and VobEdit dont have that feature; correct me if I am wrong.

blutach
15th December 2005, 09:41
That sounds more like an issue with button sizes, placement, the CLUT or transparency values Afrinux. Check these out in PgcEdit's button editor.

Regards

Afrinux
15th December 2005, 09:55
That sounds more like an issue with button sizes, placement, the CLUT or transparency values Afrinux. Check these out in PgcEdit's button editor.

Regards
Thanks blutach for your reply. I did check it in PgcEdit. Buttons, Spu palettes seem to be fine.
I can see the the Hl!s on PCs using PowerDVD, but if try it on a stand-alone player, the hl!s arent displayed.

Probably, a display time problem.
The HL!_S_PTM is set to the same value of the first video field PTS, tho. :confused:

Afrinux

jeanl
15th December 2005, 17:45
Afrinux, also look at the hightlights themselves, using DVDSubEdit (http://www.videohelp.com/~DVDSubEdit/) (sorry for the plug)! You might find something bizarre in the subpicture commands.
In my experience, this kind of problem is quite hard to debug (we did a lot of that with blutach for menushrink, and it wasn't easy). We might be able to help you with a bit more info, for example you could send us the menu VOBs (if they're not too large)...
jeanl

Afrinux
17th December 2005, 09:36
Afrinux, also look at the hightlights themselves, using DVDSubEdit (http://www.videohelp.com/~DVDSubEdit/) (sorry for the plug)! You might find something bizarre in the subpicture commands.
In my experience, this kind of problem is quite hard to debug (we did a lot of that with blutach for menushrink, and it wasn't easy). We might be able to help you with a bit more info, for example you could send us the menu VOBs (if they're not too large)...
jeanl
jeanl, thanks for your suggestion. I tried DVDSubEdit and didnt find anything wrong with the commands. HL! was also correct. When I was sure that all the conditions to display a hl! on a button were met, I started looking in the ifo file, to find out that the start address of the menu_vobs is one sector bigger.
I fixed it and voila, there were the hl!.
My vob was like this: [navpack][vidpack][vidpack]...[audiopack]...[spupack].
Since the first sector was out, the stand-alone player couldnt get the navpack info.
Thanks for your help.

jeanl
17th December 2005, 18:22
Great! Not much help, but great just the same!
jeanl

Afrinux
18th December 2005, 11:22
Great! Not much help, but great just the same!
jeanl
Well, I think if it wasnt of the tool, I would spend more time trying to figure out what's wrong with the SPUs.
It did help me a lot. Thanks a lot again.