[game_preservation] Archivist's Burden

[Devin Monnens]

Thank you all for an interesting and useful discussion! I kind of feel like
I'm late to the party! I will have to look at the links Andrew sent in more
detail. However, there are some guidelines I think you should keep in mind
before you begin this process.
Regarding formats:

NO! NO MP3s!! No OGGs!! You want a LOSSLESS format like FLAC or WAV. FLAC
has not been accepted by libraries and archives as a preservation standard
because it is not ubiquitous (LoC decided against using it because they had
increases in bandwidth and storage, but I don't know other reasons against
it). WAV is ubiquitous - I don't know of any audio player that won't run a
WAV file. It is also closest to CD native format. FLAC is supposed to be
superior: it has a smaller file size, it maintains metadata (including ID3
tags and checksums of the original WAV!), and it has error recovery
algorithms. FLAC is the standard among many audiophiles, particularly in the
game music archival sector. However: FLAC is not supported by a wide range
of players out of the box, and some won't even touch it (like Apple, in
their evil greediness). I would say avoid ALAC because it is a closed
standard. You have 800 games and a 750GB hard drive; file size is not going
to be an issue for you (even with multi-disc games), so you could go the
route of archives and just do it WAV. If you backed up everything in FLAC,
there's no guarantee you wouldn't have to convert everything back to WAV at
a later date...

Honestly, it's this hemming and hawing that has partially prevented me from
starting my lossless audio cd and game backup project.


Andrew and all:

The Library of Congress has a detailed description of each file format and
describes why particular formats were adopted.

http://www.digitalpreservation.gov/formats/fdd/descriptions.shtml


Regarding extraction methods...

I would agree if you are looking at 800cds it would be best to find a system
to automate this. I have 'backed' up 100 Amiga discs using a copy program,
and that takes a LONG time at at least 8 minutes each (I got a lot of
reading done).

You can upgrade your machine so it has multiple disc drives. I have a friend
online who dumps audio CDs and game music for friends, and he has a tower
with something like 8 disc drives, so he can do something like 100 cds in an
hour. With 800 discs, spending another $60-$100 on new drives isn't going to
hurt.

Also, look at your discs before dumping. If it's scratched, set it in a
separate file, so you just do the clean discs first. Scratched discs will be
more time-intensive because you'd have to verify the read/write crc's for
authenticity. If the discs are too badly scratched, you might have to get
them buffed. If you lived in Denver, I might be able to set you up with a
deal at my work.


Format standards:

Another question becomes what standard to use. This depends on why you want
to preserve them, but I wonder if we shouldn't be adopting a format standard
for game backup (say ISO or ISO+WAV). I think ISO or any other virtual
drive-compatible system would be preferable so you can mount the discs, and
I think it is also good to be able to recreate the original format using
your backup file because that might also be useful in the future (say if
someone wants to run a game on original, restored, or rebuilt hardware).

One current 'standard' is that used by redump:

http://redump.org/guide/cddumping/

Redump uses a process-intensive method of extraction. They use MD5 checksum
files to verify, which I think should be a must for your project as well.
However, they never explain in detail why their system is used, unlike say
Slightly Dark's extraction method for CDs:

http://www.slightlydark.com/forums/showthread.php?t=850

To be legitimate, I think they have to explain in detail why the system is
used that way rather than just say 'this is the format we adopted, go use
it.' For doing 800 games, I think it would be useful to explain why you
backed them up the way you did.

Here is a list of PC games they have already dumped:

http://redump.org/discs/system/pc/region/U/


Metadata:

Decide what metadata you wish to include. Do you want to scan the discs and
the covers? This will add time and open up a new bag of formats. A widely
accepted preservation format is PNG.


Before you fully automate the process and go gung-ho backing up discs left
and right, you want to run tests on a few games to see if the process is
feasible. How long does it take to do a single disc? Is the process
determined to back up the disc correctly? Is the resulting file usable in
the way you want it? Once you have successfully evaluated a few test cases,
then it would be time to fully adopt the strategy.


On Fri, Sep 25, 2009 at 9:53 AM, Rachel "Sheepy" Donahue
<donahrm at gmail.com>wrote:


>

>  I guess if the entire intention is to get the files in an easy to listen

>> to format, I'd suggest adding an simple 128 static bitrate MP3 creation

>> line, which would make quick and tiny files in comparison to the FLAC/Apple

>> Lossless files (and easier to provide copies of for people).

>>

>

>

> Noooooooooooooooooooooo! MP3s/Ogg are fine for access, but you want

> lossless for preservation :)

>

> _______________________________________________

> game_preservation mailing list

> game_preservation at igda.org

> http://six.pairlist.net/mailman/listinfo/game_preservation

>




-- 
Devin Monnens
www.deserthat.com

The sleep of Reason produces monsters.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://six.pairlist.net/pipermail/game_preservation/attachments/20090925/41f451cc/attachment.html>