Phil and I got into a conversation this morning (which he’s already mentioned) about the iTunes Music Store and the metadata (ID3 tags such as Artist, Year, Track #, Composer, etc. that are included with each song in iTunes) that they provide.
While I’ve played with it a bit, I don’t see myself becoming a big user of the iTunes Music Store for one very simple reason — their metadata doesn’t meet my standards. Specifically, the “Year” field is often wrong (for instance, Meat Loaf‘s ‘Bat Out of Hell’ is tagged as 2003, when it was re-released, rather than 1977, when it was originally released), and for the majority of the tracks on the store, the “Composer” field is empty — the Classical genre is the only time the Composer field seems to be used consistently.
Now, I fully recognize that for 95% (at least) of the population, this isn’t going to be a major thing at all. As long as the Artist, Album, and Track Name are there and correct, we should be happy, right? Well, sure, for most people. I’m just in that 5% who are picky (ahem…anal) about this (and it’s certainly not limited to my music, as I tend to be quite meticulous about keeping my books and movies alphabetized, and sometimes broken down by genre).
Part of why I like having all that information available is just the amount of different searches that can be done when it’s all in and entered correctly — and when you’re dealing with a music library that is upwards of 80Gb, emcompassing over 10,000 different tracks from around 1,200 CDs, that can be important!
To use one of the examples I gave Phil, Al Jourgensen has been active in a ton of different industrial groups over the years, including Ministry, the Revolting Cocks, Lard, and many, many others. As long as I have the Composer field entered correctly, then I can do a quick search through my library for “Jourgensen” and instantly I’ve got a list of every track in my collection that he’s worked on.
Another example: Bob Dylan has written an incredible amount of music, much of which has been covered by many different artists over the years. Suppose I felt like listening to all the covers of Bob Dylan tracks I had in my collection. Without good metadata, it’s not happening — but with the metadata, I can set up a smart playlist using the terms “Arist does not include Dylan, Composer includes Dylan”, and I’ve got a list of Dylan songs performed by anyone but the man himself.
As far as the Year field goes, I like to keep smart playlists for each decade — I touched on this briefly earlier this month — or just be able to sort a listing of songs chronologically. Having the correct year in the metadata is necessary for this, and years that are off can be pretty jarring (for instance, listening to a modern music playlist and suddenly having 25-year old rock could be a little odd).
So that’s why I have some of the habits I do (well, ignoring the deep-seated childhood psychological trauma), and why I’m not likely to use the iTMS for much more than occasionally grabbing a track to replace a scratched section of one of my CDs. I may use it for expanding my classical collection — I just bought a great collection of Bach‘s Brandenburg Concertos last week — but that’s probably going to be the extent of it for now.
What to do to fix that (and send more of my money Apple’s way)? Well, Phil and I batted this one about for a couple minutes. I’ve looked into this a bit in the past, and it seems that Apple doesn’t really have a lot of control over what metadata is and isn’t present for the songs in their store. If I’m remembering correctly, all encoding and tagging is done by the studios themselves, then provided to Apple for inclusion in the iTMS. So sniping at Apple isn’t likely to do much good, and it would likely be a bit beyond my ken to start harassing each of the various studios to pay attention to these little details (especially when, as I stated before, most people couldn’t give a flying fig about things like this).
The best case scenario, I think, is one that Phil mentioned: if Apple treated the iTMS library’s metadata in a similar fashion to the Gracenote CDDB (this is the service that iTunes connects to in order to automatically discover album information when you put a CD into your computer). With the CDDB, all information in the database can be updated by the users — if you put a CD into your computer and notice that some of the information is incorrect, you can make the edits and then re-submit the updated information back to the CDDB, essentially creating a self-updating and self-correcting public service.
If Apple could implement something like this for the iTMS, I’d be thrilled. Logistically, it’s a bit of a quandry, though, as it would likely need to include some sort of moderation to prevent someone submitting information for an album with every track titled “tHIz ALBum SuXX0Rzz!!!!!” or something equally intelligent. That extra added overhead creates more work for Apple, and as the iTMS is currently running in the red, Apple probably isn’t going to be anxiously looking for ways to put more money in for a feature that only 5% or less of its users are going to care about.
My suggestion, then, would be to create something akin to an “iTMS Pro” service. For, oh, \$5 a month (billed directly to the credit card already registered with the iTMS), an “iTMS Pro” subscriber could edit and submit information on tracks in the iTMS library, correcting information that might have been entered badly the first time (I’ve already noticed the occasional dropped space or inconsistent naming conventions), or adding information that wasn’t included, such as the Composer field. The information would still probably need to go through some sort of moderation process at Apple, but limiting the editing ability to people who were willing to pay the slight extra bit of money would weed out casual pranksters, and provide a few extra dollars to Apple to pay for that moderation process.
Is this going to happen? Oh, I seriously doubt it. But if it did, I’d be sure to sign up.
I like my metadata.
a new project at work is to create an online database of museum objects – it will tie in to a soon-to-be-published online library catalog for three libraries in alaska.
now that i’ve explained, i’ll get to my real comment:
metadata is a bitch. metadata is a ravenous bloody murderous bitch, and i hate it.
i’m sure it’s very useful in the end, but fuck, it’s annoying.
thanks for the rant space. ;)
Er, recording companies are called “Labels”, not “studios”. Studios make movies.
Studios also record music, but they don’t distribute the finished product.
Why would you pay them for the privledge of correcting their metadata, they should allow you to volunteer to do that and maybe give you a free song here and there as thanks.
On a related note, check out musicbrainz.org, which is a site that does a) metadata editing b) recognition of mp3 and ogg files based on a calculated fingerprint and c) using (b) to tag your mp3/ogg with data from (a).
So basically pop an mp3 into their tagger, it’ll recognize it and you can hit save to save it with the right metadata. Of course it’s not perfect. The data is not always correct, but anyone with an account can edit it (though it has to be voted ok before it’s “live” to prevent people from doing mass changes), but for the most part it’s pretty good.
Musicbrainz and itunes is a powerful combination :)
i don’t like itunes breakdown of genres as well. i wish they included sub-genre information, which would be a vast improvement. i end up redoing the genre info with sub-genre data that i manually grab from allmusic.com. it’s a beotch to maintain. :|
I think that the metadata article is almost right. I think that all the information about a song should be in the metadata tags, and it all should be correct. The author of the article is wrong about his being part of only 5% of computer users who might care. I think that at least 50% of people who use computers and store tunes on their hard-drives want complete and accurate information about the tunes. I think that at least half of all computer users who go through the trouble of downloading music files and saving them to their hard-drives also categorize and organize them; I bet they want all the information to be complete and accurate.
For those who use Windows NT, 2KPro, or XP, and who use the NTFS file-system, I’d wager that most are learning about he “Summary” section that is accessed by right-clicking on files and selecting Properties. I’d wager that they are using the Windows Media Player and are (or are becoming) very picky about the accuracy of the data that is stored within the music files themselves.
Any company that serves up tunes for downloading and storing but does not have all the correct information in the song-file’s metadata is just scrrewing its customers and assumes that all its customers are just witless dullards who should be raked over the coals.
Well, a quick fix that you may already know about is this: select the track (or group of tracks), press Cmd-i and change the year. (or other info) This doesn’t help other people who will get the wrong metadata, and may be a pain if you have a lot of different songs, which can’t be conveniently changed as a group. I just do this when my metadata is infrequently wrong.
well every Win9X proggie has the same screw ups except with Itunes the quallity is better than mp3’s.
mac can do no wrong. THE END.
I’ve been going through all the things that Google turns up about me. Man, was I ever opinionated. The things that have turned up that I have written are so stupid, including the item #6 in “Responses to “Metadata is a good thing””
If I had known then what I know now! Oy, vey!
2007-11-10
Heh. Lots can change in four years, Dean — I know that I’ve certainly mouthed off from time to time in the past, much to my dismay later on down the line. All part of getting older, I think.