DvdShrink file corruption

[Phantom_E]

If you deselect "Split VOB into 1GB size chunks" it generates corrupted navigation blocks in the output.

Is this a known issue?

If it is a known issue, does it corrupt anything else?

Is there a tool that can repair the bad navigation blocks?
Preferably in place repair rather than copy-repair.

[Chetwood]

It shouldn't do this but if it does it's probably due to a copy protection that Shrink can't handle. Best thing is to rip with DVDFab or AnyDVD first and then encode with Shrink.

[Ghitulescu]

If you deselect "Split VOB into 1GB size chunks" it generates corrupted navigation blocks in the output.
Depending on how you noticed this, yes, it may be aIs this a known issue?
How did you notice?

PS: aren't the original VOBs equally defective?

[Phantom_E]

PS: aren't the original VOBs equally defective?
No, the original VOBs are ok. I just had to redo the backup.

I have more than one disk in this particular project and the VOBs I have checked all show the same navigation pack issue. Redoing the backups with the "Split VOB into 1GB size chunks" option checked has cured the problem so far.

[Phantom_E]

It shouldn't do this but if it does it's probably due to a copy protection that Shrink can't handle. Best thing is to rip with DVDFab or AnyDVD first and then encode with Shrink.
One of the disks I tried does not have any DRM at all and I can back it up using the OS copy command. I just used DvdShrink on it because I wanted the end result to be 1 big VOB instead of 7 small ones.

[Groucho2004]

This might be a long shot but you may have a bad memory stick. I had a problem like this a long time ago (data corruption when copying very large files).

[Ghitulescu]

VOB files are binary split, so DRMed or not, you can still use COPY/B with them (provided the file system allows for the joined size :)).
And DVDshrink does not remove any protection.

Again, how did you notice the errors? What tools, procedures etc.?

[Phantom_E]

This might be a long shot but you may have a bad memory stick. I had a problem like this a long time ago (data corruption when copying very large files).
Tried same dvd disk on two different PCs with different OS, DVD reader and HD.

[Phantom_E]

Again, how did you notice the errors? What tools, procedures etc.?
I used IFOedit to get the chapter address and Vobedit to look at the matching address in the VOB.

After a certain point it is duplicating nav pack records. You can tell this because the block internal "this->LBA" address is for a different location.

[hello_hello]

Mmmmmmm.... so it's not MPC-HC's fault I sometimes have problems playing vob files? Seeking correctly.... that sort of thing?

The closest encounter I generally have with a vob file is maybe if I open one to check something while setting up an encode..... but I generally get DVDShrink to re-author the output as a single vob file. Habit really, no other reason. No problem encoding them, but maybe that's why I sometimes have difficulty playing them and why MPC-HC locks up while seeking every now and then.... I'll have to investigate the next time I rip a DVD.

[Phantom_E]

Mmmmmmm.... so it's not MPC-HC's fault I sometimes have problems playing vob files? Seeking correctly.... that sort of thing?
Other than I use WD it sounds like you have the same setup.
I'm in the process of finding out if the navigation issues the WD unit is having are related to this.

I only found the issue because VobBlanker would not reprocess a file I wanted to turn into MP3s.

[hello_hello]

If I keep the original DVD video, which is fairly rare, I generally remux the vob files as MKVs and store it that way. Once they're MKV'd I have no problems playing them..... although I did drag a few out of mothballs for re-encoding a while back and ffms2 wasn't all that happy about indexing them. I though it was just ffms2 finding yet another way to torture me..... I wonder if bad navigation blocks would effect the way the video is muxed and in turn upset ffms2? Something else to investigate maybe....

[Phantom_E]

VOB files are binary split, so DRMed or not, you can still use COPY/B with them (provided the file system allows for the joined size :)).

I thought this worked myself, but it produces corrupted output and fails if /v is used.

OSs tried W2k and Vista.
The output file is way larger than 2g.
The output fails too match the input around the 2g point though.

[Chetwood]

And DVDshrink does not remove any protection.
Except CSS.

I just used DvdShrink on it because I wanted the end result to be 1 big VOB instead of 7 small ones.
Why not one ISO instead?

[Phantom_E]

Why not one ISO instead?
The original impetus was to get rid of racks and stacks of disks.
I just move the big VOBs to a large USB HD and play them with a tiny video server.
My initial attempts at containerization (mpg, mk2, or something) said -> later.
I think later is becoming now :(

[Phantom_E]

I'm revising my guess as to what is going on.

Subject to change, my current thinking is:

Tool Max Usable file size

VobEdit 2G (signed 32 Bit integer)
Copy /b 4G (unsigned 32 Bit integer)
VobBlanker 2G (signed 32 Bit integer)
PowerDVD8 - DVD file mode 2G (signed 32 Bit integer)
PowerDVD8 - media file mode >50G (64 Bit?)
DvdShrink >= DVD-9 max (>32 bit)

Subject to additional testing.

[Phantom_E]

I wrote a bit of code to hack the big 6G VOB back into 1G chunks. File compare shows the recovered chunks match the chunks from DvdShrink when "Split VOB into 1GB size chunks" is selected.

DvdShrink is ok.

Everything else looks like a larger file pointer variable would be useful.

Nevermind

[Ghitulescu]

DvdShrink is ok.
Told you :)
VOB files are binary split, so DRMed or not, you can still use COPY/B with them (provided the file system allows for the joined size :)).



I thought this worked myself, but it produces corrupted output and fails if /v is used.
What? /V does nothing to the copy process, it simply verifies AFTER the writing, whether the data differs from the original. In other words, with /V or without /V, the output data is the same but /V draws your attention to possible errors (in the good old DOS time, a lot of diskettes had errors).

But you are absolutely sure that you used /B?
.... you can still use COPY/B with them ....

[Phantom_E]

Told you :)
Perhaps, it would have been nice to be told the real problem.:)

/V does nothing
I don't exactly think getting a good output file is nothing.:)

My fingers refuse to enter the copy command without /B (sometimes /A) and /V.
In any event it quits with a verify error at the 4G point with /V.

The stream writer classes of Windows OSs do /V by default.
I don't know if this includes the legacy DOS copy command.


PS: Copy should work fine for most DVD-5s.

[setarip_old]

@Phantom_E

Hi! I just used DvdShrink on it because I wanted the end result to be 1 big VOB instead of 7 small ones.I sense that the difficulty you've encountered may be related to the long ago discovery (during the development and beta testing stages of DVDShrink) that DVDShrink did/does not properly handle layer breaks...

[Phantom_E]

@Phantom_E

Hi! I sense that the difficulty you've encountered may be related to the long ago discovery (during the development and beta testing stages of DVDShrink) that DVDShrink did/does not properly handle layer breaks...
I did some searches to try to understand this.

Am I correct in thinking layer breaks would only be a problem if:

A) I chose keep layer breaks in Shrink.
and/or
B) I want to write a DVD-9.

[setarip_old]

If I remember correctly (and that's a big questionable "IF" at my stage of life ;>}), it was established during our beta testing that since DVDShrink operates only in the world of compliant DVD format:

1) Maximum .VOB size (individual or combined) could not exceed the equivalent of a DVD-9

2) Any form of DVD "package" (hard drive or disc) greater than the equivalent of DVD-5 up to a maximum equivalent of a full DVD-9 would require a properly generated "layer break"

I believe this (flawed layer break problem) can now be overcome by using IMGBurn in conjunction with DVDShrink

[Phantom_E]

In reviewing the various attempts I tried to solve the problem I was having...

I looks like two tests did indeed fail due to layer break issues.

It may also explain other issues I have. Particularly with non DRM CRC errors.

Thanks for the heads up.

[setarip_old]

As always, my pleasure ;>}