Found (IMHO) source of wrong time and chapter displayed while rewind through chapters

[Sir Didymus]

I know it is a minor problem, but anyway, it is good at least to know what is causing this behaviour...

I collected some evidence it should be related to the improper filling of the Search Information
Table (VOBU_SRI) in the navpacks...

For the navpacks at the beginning and at the end of each cell, the entries of this table should be
filled with 0x3fffffff for all of the search positions that are undefined or that are referring to
time positions of video packs belonging to another cell.

For example the very first navpack of a vob should have 0x3fffffff filled in all of the "previous
search positions" (because there are no previous video contents...), while it is easy to see that
the filled values are 0x800000.

All of the first navpacks in each cell of a VTS have the entries of this table referring to
"previous search positions" (address 0x545 ... 0x595 in the navpacks) uninitialized:
sometimes there is the value 0x8000002b, other times there are other values, but still
improper values.

Manually clearing these parts of the VOBU_SRI tables across of the cell boundaries, makes the
display of my standalone player showing properly the time and the chapter changes when rewinding
through chapters...

Cheers
SD

[jdobbs]

You are correct that 0x3FFFFFFF should indicate the fact that there is not a VOBU within the time span specified within an entry in the table. I know I'd addressed that issue, but if you're finding incorrect entries I've obviously done something wrong.

Outstanding investigating! Thanks!

[HanSolo00]

Excellent work, and perfect timing for myself--I'll leave my latest project encoded and ready to complete with the next version of Rebuilder when it's released. I think this is the only remaining 'stream flaw' that I've seen in my backups and minor as it is, good news to have it found.

[jdobbs]

@Sir Didymus

Thanks for the info. I just found this error and corrected it, I will post the update in a few minutes. I had populated the table correctly (with the 0x3FFFFFFF entries) but had a comparison that was checking against the wrong value when deciding whether a legitimate value should be entered -- as a result the 0x3FFFFFFF was being overwritten with some variation of 0x800000nn.

With beta testers like you -- who needs programmers!

Thanks again.

[Sir Didymus]

Congratulation, boss...

I confirm this little problem is fixed with version 0.48.

Just tested and verifyed, both using Vobetit and my standalone player.

Thanks to you for your wanderful application.

SD

[jptheripper]

while not rb related, i am having problems I think this addresses. I authored dvd in ifoedit, windvd playback is fine and everything works. however, on standalone:

straight playback works
if i skip chapters beyond chapter 6 it plays for like 5 secconds and then freezes
If i FF or RW it freezes.

I looked for the VOBU_SRI tables in IFOEDIT 0.96 but couldnt find them..

any help would be appreciated.

[jdobbs]

Originally posted by jptheripper
while not rb related, i am having problems I think this addresses. I authored dvd in ifoedit, windvd playback is fine and everything works. however, on standalone:

straight playback works
if i skip chapters beyond chapter 6 it plays for like 5 secconds and then freezes
If i FF or RW it freezes.

I looked for the VOBU_SRI tables in IFOEDIT 0.96 but couldnt find them..

any help would be appreciated. Those tables are not in the IFO, they are in the VOB. Even with VOBEDIT they aren't shown easily, you have to find the values with the HEX editor.

MPUCODER's site outlines them:

http://www.mpucoder.com/DVD/dsi_pkt.html

[jptheripper]

thanx so much