Filesystem Metadata Doesn't Scale

I do recommend switching to ASCII tags; it defaults to Unicode by default, which most people won’t need, and this doubles the size of the tags.

Interesting advice, Jeff. The tags are, what, a few hundred bytes worth? So doubling the size of tags of everything in your collection costs you literally some megabytes. At current rates ($0.60/GB) that could add up to maybe even a penny with a really big music collection!

Alternatively, you could say damn the expense, spend the extra penny, and let bands like Sigur Rs be represented faithfully in your metadata.

spend the extra penny, and let bands like Sigur Rs be represented faithfully in your metadata

Point taken. It’s a meaningless savings.

But Sigur Ros doesn’t need Unicode, does it?

I escaped from the mp3 tagging hell!
I completely gave up on relying on mp3 tagging. I’m sorting my music library with folders (interpret/album/trackXX.mp3) and that works fine for me. I fiddled around with iTunes for a while but I do not use it anymore.

And I deliberately provided myself with an iRiver MP3 player, which can navigate by folders as well. And as I generally listen to music by entire albums, I do not need Genres, year of release and other metadata.

Works fine for me, YMMV though.

No one mentioned the Godfather?

It’s the best tagger in my opinion.

http://users.otenet.gr/~jtcliper/tgf/

You can search CDDB with it, use regular expressions to lift tags from existing files/folders, batch rename. I’ve yet to find something it can’t do.

I was about to make the same point as Alistair about the true savings of ASCII/Unicode for a few words in a file megabytes in size, but then I noticed I had been beaten to it…

It seems to me that the Genre tag is a problem - with some music you really need to be able to assign it to several genres (or rather, there are several possible genres that might all make sense) and so really you want a collection/set of genre tags for any given recording, rather than just a single, arbitrary one. After all, one person’s rock is another person’s metal, and so on and so forth.

I have a 70Gb+ library on the Mac, and using iTunes I find it’s not too bad really. As has already been mentioned, iTunes has some acceptable tagging features and allows you to create reasonably flexible ‘smart playlists’ too. Whether I encode something myself or acquire it in electronic format already, I always try to get the tags right because I listen to a lot of my music on my old iPod and good tags are vital when you have a large library on something as limited as that.

Perhaps we need a ‘tag authority’, a central database of artists and albums with appropriate, acceptable genre tags. Maybe some neat way of searching CDDB to come up with a ‘general consensus’ of who belongs where? :slight_smile:

I was about to make the same point as Alistair about the true savings of ASCII/Unicode for a few words in a file megabytes in size, but then I noticed I had been beaten to it…

It seems to me that the Genre tag is a problem - with some music you really need to be able to assign it to several genres (or rather, there are several possible genres that might all make sense) and so really you want a collection/set of genre tags for any given recording, rather than just a single, arbitrary one. After all, one person’s rock is another person’s metal, and so on and so forth.

I have a 70Gb+ library on the Mac, and using iTunes I find it’s not too bad really. As has already been mentioned, iTunes has some acceptable tagging features and allows you to create reasonably flexible ‘smart playlists’ too. Whether I encode something myself or acquire it in electronic format already, I always try to get the tags right because I listen to a lot of my music on my old iPod and good tags are vital when you have a large library on something as limited as that.

Perhaps we need a ‘tag authority’, a central database of artists and albums with appropriate, acceptable genre tags. Maybe some neat way of searching CDDB to come up with a ‘general consensus’ of who belongs where? :slight_smile:

I’d echo Nicholas’s comments on Exact Audio Copy (EAC). It really is a superior ripping tool. Jeff is right that “data is data”, but I don’t know how many times I’ll be listening to a song I ripped a long time ago (without EAC), and there will be a skip in the track. I hate that!

EAC let’s you know when there are errors in the read, and even allows you to listen to the exact position of the error, so you can tell if it’s audible. Combine that with being able to use whatever compression algorithm you want (FLAC’s for me), and EAC is the king!

I use iTunes to organize my music as I found doing the filesystem way to be a complete pain. I should preface by saying that I use iTunes on a Mac, so it may not apply to iTunes on Windows; I have no idea.

There is an option in iTunes under “Advanced - General” that says: “Keep iTunes Music folder organized. Places songs into album and artist folders, and names the files based on the disc number, track number and song title.” I’ve used this since the beginning and I can browse my library without iTunes if the need arises (it helps when copying some music out of it sometimes). It’s considerably better than when I tried to organize it myself, but that probably has something to do with the irregularity with which I actually added things. And laziness.

The other trouble with a rental service is that they’re very limited in what they have. AllOfMp3 is nice since at least what they have is available the way I want it (flac) and the tagging is pretty good. But I admit I only bought the plastic junk for about half the music I own. Another 20% or so comes direct from the artist and the rest from allofmp3 (with the interesting legality of that site).

I’ve been pleasantly surprised at how many artists are happy to accept payment for ripped albums, one in particular responded by putting hi-res artwork on his website for me so I don’t have to live with scans (there’s a huge loss from jpeg- print- scan- jpeg).

I would contend that although “Earth, Wind File” isn’t correct, it’s an understandable Freudian slip after your Herculean reclassification effort … :slight_smile:

I have been doing the same thing recently Jeff with over 300 gigs of FLAC files. Albums are the easy part, but what about live stuff. I bet at least half of my audio is live soundboard recordings with no way to lookup information (even though I have it all int he .txt), so I have just been doing the albums first. I have been using Picard, even though I hate it because I haven’t found anything better. I will try out MonkeyMedia thanks for the tip. It really shouldn’t be this hard though IMO I guess thats what happens when your on the edge;)

I prefer id3-tagit (http://www.id3-tagit.de/english/index.htm) to do my tagging. It’s free, feature rich and can be controlled via the keyboard making working with it very fast.

I second this, I use ID3Tagit myself, and it is very good to use for tagging files.

But Sigur Ros doesn’t need Unicode, does it?

Oh dear God, yes! Or Latin-1 at the very least. It might be a lesser sin to misspell non-English words by dropping accents, but “” (thorn), “” (edh) and “” are another matter entirely.

“It’s giant-- currently 10,970 songs and 733 albums in 48.9 gigabytes”

uh-huh, and you have ALL the cds for those right. :wink: (jk). It may be time to look into a music subscription service. I don’t often say that to people since I hate the idea of renting music. But if you’ve got THAT much…it seems like the amount of time you’ve spent ripping them would be offset by a $10/month charge.

Have you tried MusicBrainz? It allows you to pick the tag you want to use from multiple tags in the database (I’m assuming it’s the Gracenote one). I used it to clean up a bunch of dupes and ones with no information other than a partial title in the filename.

Well I got tired of fixing albums after i rip/download them, so I wrote a tool that best fits my needs: http://thecrab.sf.net/ It won’t fix your whole music library, but it does help me out with new albums.

uh-huh, and you have ALL the cds for those right.

Actually, I do have the CDs for 90% of my collection. Like I said, I’m maniacal about VBR and LAME encoding, and the only way to get that is to have the original uncompressed versions…

It seems to me that the Genre tag is a problem

I’m pretty sure the ID3v2 spec allows multiple tags (frames) for Genre and Artist. But none of the applications support displaying it… classic chicken and egg problem.

The only disadvantage is that it rebuids it’s index at every start.

This is also a key difference between TagRename (it lets me set the “disc x of n” TPOS field) and MediaMonkey. TR doesn’t build an internal database of ID3 tag info. So it has to re-scan every time you select the target folder(s). Although both methods have pros and cons, on the whole, I think I prefer the internal database method. It’s much faster for hopping back and forth between folders, particularly if you are only using the one tool to change things.

Jeff, it is often over looked media player but I am a big fan of Musicmatch. It has a feature that will parse the file name (you can choose it parses) and automatically update the ID3 tag. Might help with your problem.

Um if my memory serves me right than WinAmp let’s you edit ID3 tags with ease…

All you do it right click on the file in the play list select “View file info…” and it comes up with all the info that the file has. Simple and free. It even has ID3v2 tags. Now this can’t organize your music, but simple folders do the exact same thing yes? and you can open an entire folder via-WinAmp.

Which of these is correct?

“Earth Wind Fire”, or “Earth, Wind File”, or “Earth, Wind and Fire”?

The correct answer is “all of the above”.

Except for “Earth Wind File”, surely? You’ve been writing about filesystems too long…

Here’s another vote for MusicBrainz… It truly is an amazing app.

Shame it didn’t work the first time you tried it, but I very strongly recommend giving it another shot.