dvdbackup generates huge copy

[Stefan Monnier]

[ Maybe this is the same bug as bug#648920, but I don't understand that one enough to judge. ]

Viewing a DVD in my favorite way is paradoxically inconvenient since it involves copying the DVD to a machine that doesn't have an optical drive. So dvdbackup is a very good friend of mine.

But recently it has let me down a few times, because the VIDEO_TS folders it generates for some DVD are much larger than they should. E.g. it might generate a 60GB folder for a perfectly normal (at worse dual-layer) DVD. The result works (VLC plays it just fine), but the extra space is problematic, both in disk space and in network-transfer time.

In case it's relevant, I'm using Debian's dvdbackup-0.4.1-1 package and I'm usually very pleased with it.

[Gregory Alexander (gwa)]

This appears to be due to aliasing multiple filenames to the same file on the DVD.

This is probably a form of copy protection, though it could be done intentionally for some kind of IFO file behavior.

The "best" solution I can see here is to create links in the target directory.

Thanks!

[Bryce Harrington (bryce)]

I've seen this behavior on a few different DVDs, although usually the resultant VIDEO_TS doesn't play that well (lots of skipping and freezing).

[Fabian Keil (fk-ux)]

For discs like this I slapped together a perl script that takes the output of isoinfo -i and classifies the files as either "master files" or duplicates.

Based on that classification it can deduplicate an existing copy by replacing the duplicates with links to the "master files" (like Gregory already suggested above), or generate a shell script that would do the same but can be adjusted before executing it.

The deduplication logic currently only works for discs where the duplicates are obvious, and I suspect that it may be possible to master DVDs in a way that they aren't, but so far I haven't seen this.

My goal is to eventually be able to only rip the master files. A dvdbackup option to only rip files specified by name (or start sector) would make this possible and I intend to work on this (or another solution) the next time I want to rip such a disc.

The duplicate detection could be integrated into dvdbackup itself, but I haven't looked into how much work that would be.

The skipping and freezing issues Bryce mentioned are probably symptoms of dvdbackup padding blocks that shouldn't be padded and not directly connected to the duplicated files. Using read error mode "b" could help but might take a while.

[Bryce Harrington (bryce)]

The bug may really be more in libdvdread4; it ought to detect this situation and correct for it.

Indeed, with recent libdvdread4 versions, some DVDs that would exhibit this behavior now produce normal sized directory structures. Other DVDs still seem to have the problem, so YMMV.