isoterik: a region-code scrubbing, non-remastering DVD cloning tool

Introduction

isoterik extracts a "purified" .iso (UDF) image of a video DVD, devoid of region code. (And CSS encryption.)

When this image is burned to a writable medium, the resulting disc can be played in any DVD player, regardless of region setting -- so you no longer need an extra DVD player just to play imported discs...

Although not the primary motivation, this can also be used to just easily copy any video DVD on the command line, even if you don't care about region code removal.

While other Free Software programs that can copy DVDs (such as k3b or Brasero) creade a new ISO/UDF filesystem from the decrypted video files (using genisoimage/mkisofs), isoterik keeps the original filesystem structure, replacing only the actual purified data. The approach used by the other programs yields slight differences from the orignal at the filesystem level (such as different file creation dates) -- unless there are some buggy corner cases in genisoimage and/or some DVD players, this shouldn't really make any difference in practice though.

Note: isoterik itself doesn't do any CSS decryption. Rather, it invokes dvdbackup to extract decrypted files; which depends on libdvdread to do the actual work; which in turn needs yet another library to deal with CSS, that I might not even be allowed to tell you about without risking legal repercussions... Yay for copyright legislation.

Installation

isoterik consists of two components: the main isoterik program, which is a shell script; and the udffindfile helper binary, which wraps a particular functionality of libdvdread that is not otherwise available in any way that could be used from a shell script.

To get isoterik working, first you need to build the binary. In addition to the basic build tools (on Debian and derivatives, install the build-essential package), you also need to install the libdvdread development package. (libdvdread-dev on Debian.)

Once the dependencies are in place, on typical GNU/Linux systems just running make in the source directory should be sufficient on yield the udffindfile binary. (On other systems, possibly some adjustments might be necessary to Makefile.)

For isoterik to be able to use the binary, it needs to be made available in the executable search path -- typically by copying it to /usr/local/bin or ~/bin. While the actual isoterik program can be invoked from the source directory, I suggest installing it to one of the bin directories as well, for consistency and convenience.

At runtime, isoterik also relies on the dvdbackup program, which needs to be installed separately. (It's available in the dvdbackup package in Debian.) Furthermore -- as mentioned in the introduction -- to actually obtain "clean" copies of CSS-encrypted video files, the libdvdread library underlying dvdbackup needs to have access to a particular extra component...

If your system can already play the DVD you want to purify (using Free Software video players such as VLC, Totem etc.), then the extra library must already be present, and isoterik should work just fine. If you can't play the DVD, then isoterik won't work either, and you need to install the extra library first.

(Or maybe there is a problem with your DVD drive -- in which case there is nothing you can do, except using a different machine... My notebook for example fails with at least one DVD having a "foreign" region code, although it plays other DVDs just fine; while my two desktop drives work with any DVD. I have no idea how common this is.)

Another runtime dependency is /sbin/blkid, which should always be present on GNU/Linux and similar systems, but might be missing elsewhere.

Usage

The main program currently doesn't accept any command line arguments -- so all you can do is simply invoking isoterik. This will attempt to access a video DVD through /dev/dvd -- on most systems, this is a link pointing to the primary DVD drive device. If you need to use a different device, you have to edit the actual isoterik script, changing the value of the DVD_DEVICE variable defined near the top of the script; as well as dvdfindfile.c -- which requires a rebuild/reinstall to become effective...

(Some additional setting variables are available in the isoterik script -- most notably to change the verbosity of status output.)

If a video DVD is present, isoterik first reads its "volume label", which is a short string uniquely identifying the DVD -- usually a shortened but recognisable variant of the official title. It then creates a "purified" image in the current directory, named after the volume label with .iso appended to it.

During operation, isoterik first creates a temporary directory (also named after the volume label, but without .iso appended) within the current directory, where it stores decrypted copies of all video files from the DVD. When assembling the final image file, these temporary files are successively removed. Still, the partition where the final image and the temporary directory are stored does need some temporary extra space during processing, on top of the size of the final image -- in the worst case, up to the maximal size of one VOB file, which is 1 GiB.

ToDo

isoterik is still pretty rough around the edges. Possible future improvements:

• A "proper" build system for the binary, using Autoconf.

• Command line options -- most notably for selecting the source device; and for changing verbosity of status messages.

• Operation could be optimised by storing the decrypted video contents directly to the final image, instead of extracting them to temporary files first. This would avoid the need for the temporary directory (and the temporary extra space usage); and avoiding the extra copy would also speed up the operation considerably.

  This different mode of operation shouldn't be too hard to implement in the isoterik script itself -- however, it would require extra functionality from the underlying programs: specifically, the ability to extract individual video files.

  (And probably also a more explicit way to get a listing of all video files on the DVD. While dvdbackup can print this information, the output doesn't appear suitable for reliable machine parsing...)

  There are several possible approaches for tackling this problem:

  • Use vobcopy instead of dvdbackup. This is another program doing mostly the same, but with some different functions -- including extraction of individual files. However, I'm reluctant to introduce a dependency on vobcopy, as it seems rather unmaintained, and the code quality is horrible...

  • Implement additional helper binaries. However, unlike the udffindfile functionality, extracting files is a rather complex process, which would require duplicating quite a lot of functionality from dvdbackup.

  • Implement the whole thing in C. If we have to implement large parts of the core functionality anyway (rather than relying on existing programs), there is really not much point in using a wrapper script at all...

  • Add the required extra functionality to dvdbackup, and use that in our script. This shouldn't be too hard to do -- but it would complicate deployment of isoterik, as it would now rely on a new version of dvdbackup to work. (Unless we keep the old mode of operation in place as a fallback, complicating the script...)

  • Implement the whole thing within dvdbackup: introducing an optional new mode of operation doing the same as isoterik...