I’m about half way through my collection of CD’s at the moment in ripping them to .ogg. Though to tell the truth, that’s my ‘music’ cd collection for the most part. If I consider adding my language CDs, I may have a long way to go. I have more music variety, but some of the language collections are 10 or more CDs.
I haven’t quite decided how to treat the lauguage CDs. Since they are primarily voice, I suspect that I could get away with ripping them to Speex rather than ogg vorbis, but I don’t know if I would end up missing any of the inflection and other characteristics that are good to have in language study. It’s something I have to think about a bit.
I’ve used a variety of CD ripping software over the years. Some do just about everything for you, some take a bit more work. Part of the problem with rippers that ‘do everything for you’ is that they relly upon the correctness of the information they find in the CDDB sources they use. Where this becomes an issue is when your collection contains multiple CDs from the same artists or a CD from an independent with a small distribution.
In the case of the former, the big problem is that how you think something should be entered is not necesarily how I think it should be entered, or someone else for that matter. As an example, I have about 8 CDs from The Steve Miller Band. Some of the CDDB entries for the CDs list the artist as ‘Steve Miller,’ some as ‘Steve Miller Band, The,’ and a few as ‘The Steve Miller Band.’ If you let the ripper created all the various sub-directories for the each CD, you end up with 3 (or more) directories for the same artist.
I suppose I should explain that I do directory structure as ‘MyMusic/ArtistName/AlbumName/’ with each song it’s own file within that structure. Multi-disk collections usually end up with a folder per disk. In theory I could also catagorize things by genra either before or after the album, however I’ve found that letting the music player I want to use handle genera sorting tends to work out best. Some artists have moved across genra’s over the years. ZZ Top for example can be considered Hard Rock, Country, Blues, and a few other possible genera’s depending upon what disk you are looking at. Likewise Jewel has Folk, Folk Rock, and poetry CDs out. If your ripper does not handle ‘folk rock’ you may end up with either folk, rock, or misc as the genra.
The latter problem (independents or small distributions not being indexed) means that I have to enter all the information myself. That’s OK for the rather small collection of these CDs that I have, but since I am not interested in maintaining an account to enter data into the CDDB servers for these small number of disks, I end up with the prospect of running into the other problem later on when the CD does get indexed by someone else. I’m pretty sure that I’ve run into this issue already.
There is one other problem. CD’s do not have an ‘information’ track where the information regarding the content of the disk. As a result the various CDDB servers use information regarding track starting locations, counts, and lenghth of each track to determine what CD you are working with. Over the wide variety of CDs that have been entered this leads to some situations where multiple CDs have the same indexing information. For most CD’s this does not present a significant issue, however I actually have one or two CDs that not only have collisions in index information, they also have collisions within the genera they work within. Worse is that the song titles have ended up being ‘mostly’ right, though there was other information that was wrong.
As I mentioned in an earlier blog post, I’ve already ripped almost everything I have at some point in the past. As a result, for the vast majority of my music, there’s already a folder for each CD. I would prefer to maintain using those folders for a number of reasons, not the least of which is to give myself some idea of where I stand on on the conversion process. This also means that I pretty much need to do this all on one system. If I were ripping enverything for the first time, I could distribute the load across multiple computers. Then I would just drop CDs into open trays, pick out CDs that had been ripped, and when everything was done merge all of the files into one directory structure that then gets copied, shared, etc. to the various systems I want the music on.
If I was willing to wait a couple of weeks to get everything done, and I was taking the fully ‘automated’ route, I would just stack a part of the collection of disks next to the computer, and every time the computer kicked out one disk, I would just drop another in it’s place, and let it grind away. However I am using a multi-core computer, and it can handle ripping more than one disk at a time. I’ve got an Internal cd-drive, and several external cd-drives that I can work from. Actually they are dvd+/-rw drives, but that’s not really all that important. It turns out however that even when ripping through different bus interfaces for USB, that I can effectively rip from only 2 sources at a time. Ripping time goes from about half the play time for a disk on the first two drives to more than the play time for the three disks if I add a third disk to the mix. In other words when I rip from 2 drives at a time, I can rip about 6 one hour disks in 3 hours. Add a third drive, and I can rip 3 disks in 3 hours. (What I was ripping at the time of the testing was disks with about an hour of music on them.)
While I am about half way through my collection, by count I am only about a third of the way through. That’s because the first ‘half’ of my collection consists largely of classical music and a few one or two piece collections, such as radio shows, or ambient sounds. (Hard to call that music in reality.) When the average lenght of a piece of music in the collection that has already been ripped is 10 min, and the average for the remaining part of the collection is 4 min., you end up with some disproportion in the counts…
About the only thing more tedius than riping CDs is doing laundry. Of course it sounds like a good thing to do at the same time. Once or twice.
More snow on the way it sounds like, so now would be a good time to make a quick run to the grocery store before the roads get bad again.
