View Full Version : Big music archiving project, need help please

2006-01-29, 03:44 PM
Sorry if this topic is covered elsewhere. I just started archiving my audio CDR collection using the EAC>flac method. I have about 800 CDRs I would like to archive. I wish I knew 8 years ago what I know now. What I have is about 700 CDRs with no lineage information and various, often bad media. I donít intend on ever trading any of the dirty lineage CDRs but I love them all and would like to preserve them.

My goals are as follows:
Make perfect copies of all disks
Compress them all to a loseless codec
Burn DVD archives

I am about 5% of the way through this project and it is going to be extremely time consuming. I would like to do it right. My current method is DRIVE[BENQ DVD DD DW1620]>EAC v0.95 beta 3>FLAC 1.7.1[8] All software settings are configured to the specifications of links off of this forum.

I have concerns about what kind of media to use for the archiving. I have read the http://www.digitalfaq.com/media/dvdmedia.htm page and it appears that TDK and Maxell are good blanks. I am also looking for opinions on Optical drive preferences and lossless codecs. I have limited to no budget for fancy hardware but it seems that you can get Plextor drives for $50 anymore.

Has anyone done this type of project? Your input will be greatly appreciated.

2006-01-29, 03:57 PM
Using high quality media is a must. TY (or anything made in japan, as it is essentially TY) is the standard for quality media.

Use FLAC, SHN is being "filtered" out.

If you can find the original checksums for your shows, then use SHN and compare the original checksums against yours.

And even though your lineage wont be that great, still include a text file with your shows, if anything it whill help you down the road.

Good luck, it is VERY time consuming.

2006-01-29, 04:14 PM
Thanks for the input. What do you mean by:Use FLAC, SHN is being "filtered" out. ? Also 90% of my collection came form snail-mail trades so there are no checksums or possibility to determine lineage, all will be CDR(x). How do you create MD5 files? I understand what they are and that MKWact will verify them but FLAC does not seem to have the option to create one.

2006-01-29, 04:28 PM
What I mean is that people are starting to use FLAC a lot more, and its become the more prefered lossless format.

There are stil some ways to determing the lineage (even though you will never really know) by using databases such as db.etree.org or phishhooks search feature.

There are several different ways to create MD5's, depending wether you are on a mac or pc (there should be links on this page to the programs and guides how to use them).

FLAC doesnt use MD5s, they use Fingerprints. They are essentially the same thing.

So, if you can find the original MD5s for a show you do have on CDR, EAC it and use SHN. Create and MD5 of your files, then compare your MD5 output to the original MD5.

2006-01-30, 05:08 AM
just to further what toto's sayin...get the prog Trader's Little Helper [ http://thor.prohosting.com/roh0205/ ] after extracting w/ EAC, you'll be able to convert yer wav files to flac and create checksums--i would make an .st5 [also known as an md5 fingerprint] instead of a .ffp or traditional .md5

for the shows you are archiving from Audio CDRs that had no info or checksums, yer lineage would be:
CDR(x) > EAC v0.95 beta 3 > wav > Trader's Little Helper v____ > flac

remember to save and include yer EAC Extraction Log [anything less than a 99.9% on any given track should be considered an error and that track should be re-extracted]...also, i usually make a .st5 checksum of the wav files to add as well the checksums for the flac files

as for media--Taiyo Yuden...theres no other way to do for archiving...check supermediastore.com or rima.com for deals

any other ?s, just post back here...good luck!

2006-01-30, 06:51 AM
I think the easiest would be to set up flac as an external encoder in EAC. Your rips should be fine as long as EAC reports "no errors occured", the track quality percentage only has to do with the number of re-reads to get a correct read.