Dr. Frank's What's-it

I thought I'd read a story recently about scientists in Spain who had developed a computer program that could 'write' the music (as in, not the lyrics) to a song that would have a high chance of being a #1 hit. From what I remember, the scientists entered all kinds of info regarding melodies, harmonies, tempo, etc. into a database and then had the computer determine, based on what songs had been #1 hits in the past, what exactly makes a #1 hit. It could then generate a brand-"new" song that would have a high statistical probability (ignoring the face value of who's singing it, etc) of being a winner.

Of course, this was probably in the Onion or something, but... anyway, Dave Bug's idea reminded me of this. [shrug] =)

Speaking of bastardizing your work, I have decided to record my Lovin' Cup version of "Monkey" on a forthcoming CD project. Fits in perfectly....

Jeff Whalen (now of Tsar) sent me a fax in Prague in 1991, saying "write a second verse and a bridge, and fax it right back!" Along with:

C Am
Hey
C Am
You were walking my way
F G
And I happened to say
C Am
Hello, I love you!
-----

I did, and he recorded the song that afternoon ... and it was pretty good, for a stoopid little song.

If you ever find a big Top-40 database, let me know. I've been wishing for an IMDB-type thing for music....

I don't write music, but I write stories. So I've decided to take "Bones on the Road" and "You Scare Me" and come up with short stories to go with the titles.

Thanks for the inspiration!

I was looking at the BMI database and I didn't know you wrote "Coffee Tea or Me" Frank. I was also surprised that there weren't any other songs titled "Even Hitler Had A Girlfriend". I know there was a movie of the same title and then a comic book based on that movie. I think "Even Hitler Had a Girlfriend" is a kind of old joke so I'm amazed no one else has a song with that title.

La musique de MTX est comme un vin fin. Il devient meilleur avec l'âge. C'est une évolution dans le bruit qui est créé par Dr. Frank. Nous attendons ardemment le produit fini démonstration de huit de la petite chansons.

Babelfish did a better job of translating that than any other time I've used it.

For some reason, my first thought was to try Ba Ba Ba Ba Ba, even though it now seems absurd that I'd think there would be another song with that title. But then, there it was (with "Song" attached). The duo offers up quite a few challenges: "Bizarre Squirrel Ritual", "Blazing Entrails", "Filthy's Dance", and two possibilities in "Brotherly Love Hurts" and "I Got People Inside Me."

There are, currently, 5 songs with a minimum of "Ba Ba Ba Ba." Quite oddly, 3 of the 5 are written by Franks. (Not as oddly, one of those is published by Bug Music.)

Sadly, at this time it appears the only hope for the TuneOf program would be manual entry (though that sort of thing is not unheard of (see imdb.com or any of the CD DBs)).

A friend brought up the question of copyright/intellectual property infringement. Does the songwriter "own" the rights to distribution of their rhyme scheme and syllabic count, even if it's one of dozens of songs using the combination?

Maybe someday MP3 ID3 tags (or the future equivalent) will have spots to add in an 8:AABB/CCCD attribute.

Matt, I can't wait to hear your Loving Cup Monkey treatment (hey-- a couple of years ago that would have sounded like a pretty hip band or album name.) That's really the answer to many of my problems: get my friends to record my tunes and let them worry about how to pull off the grandiose ideas. All right!

Michele, now all we need is a story title data base so I can try to write a song with the title of a story written by someone who also wrote a story called Trail of the Slug. I'm ready. This is really a brilliant material-generating scheme. Yes, soon we'll all be hip deep in material.

Merci beaucoup, Maurice. Comme un Vin Fin is a pretty cool title aussi.

Dave, "I Got People Inside Me" reminds me of a double version of "The Moon Inside" by Robyn Hitchcock. If it's about a woman who is pregnant with twins, that is. If not, not.

As for whether you own the rights to your song's structure, we'd better hope not, 'cause then everybody would have to sue everybody else. Like the Rutles. And I'd probably sue myself by mistake, like, uh, Barry was it?

By the way, the Covers Project at coversproject.com is another somewhat intriguing Web site sort of along these lines...

BMI Diving. I like that.

I used to go in there just so I could get myself worked up about all the music that I would never hear. I swear Billy Corgan (bad example?) has only released about half of the songs that he's registered. I'm a record nerd and, unfortunately, a completist for several artists. When I find out that there are songs that an artist felt worthy of recording and registering with BMI, yet not worthy of their albums (or me), it just bugs the crap outta me. I haven't been back in there for three years now.

Has anybody else gone digging around in there just to find that (artist) recorded (song title) during the (album title) sessions only to discard it?

Tim, I'm not sure why someone would register a non-released song. As you can deduce by looking at my roster, I've never registered anything till I knew it was coming out in some form. Even the out-takes from studio sessions haven't been registered till they get "put on" something. Anybody know if I'm blowing it in some way, or rather: *is* there any benefit to registering unreleased songs? Live performances maybe? You never get anything for that in the US as far as I know, but some venues in Europe have you fill out little sheets with the publishing info. for your set list, that theoretically eventually determines some amount that gets paid out in some way. Or maybe those songs were written by Mr. Corgan and recorded by someone else? I've never recorded "She's Another Thing," but the Smugglers did. If you didn't know that, you'd wonder where that song went.

Here's something I've wondered about BMI royalties. As I've mentioned elsewhere, "performances" are credited on the basis of a statistical extrapolation of the appearance of songs on a sample of radio station logs. Your statement doesn't reflect reality. For me, the section for radio "performances" is usually mostly from College Radio. Typically, there will be a dozen or so songs listed, out of which 10 will be credited with, say, 106 performances each, while the other two are credited with, say, 319 performances each. (To put that in perspective, the royalty for 319 theoretical college radio plays of "Sackcloth and Ashes," for instance, on one statement, was $16.82, so it's not like it makes a great deal of difference, though it does add up. It's like five bucks per hundred "plays.") Obviously, it isn't the case that all those songs just happened to get played the exact same number of times. It's a theoretical number.

Sometimes, the tracks listed seem bizarre, a bit arbitrary. There will be several songs from a recent record that you know got a bit of airplay, for instance, along with one really obscure one from an out of print album from 15 years ago that gets credited with twice as many "performances." Perhaps one dude played "Vive la France" one time, and it got fed into the BMI statistical rate generator, which spit out the arbitrary number 383 based on its sampling method. I believe certain stations are weighted differently, too (though I could be wrong about that) so maybe it would have been one of the weightier ones.

Sometimes, though, you really get the impression that it is absolutely random. Like the computer simply arbitrarily assigns a number of random tracks and an amount of plays to credit them for with no criteria at all. (For the smaller fish, at least.) It doesn't seem to bear much relation to your release schedule, or any predictable factor. Yet I have often had the impression (though I could be very, very wrong and I'm sure I am) that as time goes by, and you have more and more titles listed on your roster, you end up with more and more plays getting credited. Surely it can't be that by registering more songs you increase your chances of winning the small but edifying (if puzzling) BMI lottery? Surely not. But it kind of seems like that sometimes.

So I've always wondered what would happen if I just went ahead and registered a couple hundred fake song titles. Would I increase the chances of plays being credited to me? Would any of them ever turn up on my statements, despite the fact that they don't actually exist? Has anyone ever done it? I don't know. I haven't done it and I never would. I'm sure Billy Corgan wouldn't do it either, and would have no reason to do it. But I've often wondered. Now here I am wondering if he ever wondered about it...

(By the way, the TV and film-related royalties appear to be way more logical and believable than the radio ones. Because of the more predictable programming? Because there's more money involved and fewer outlets? I don't have a lot of commercial radio play, but I wonder if the greater uniformity of programming means that statements of those who do would cause them at least to look more accurate?)

Thank you very much for your insight. Before I go further, I feel I should tell you that I'm in way over my head here. If I come off as a little dim, that's because I am.

Most of the Corgan puzzle has been solved. Several songs were released on the Machina II and Judas O comps. An album's worth of them were actually written for Hole. And a ton of them showed up in ten second clips on a 27 minute b-side made up of bits and pieces of outtakes.

About a dozen songs remain a mystery.

I figure we'll see a boxed set in the next year or two.

BMI games: If a radio station played the medley that's made up of fifty other songs, would Billy receive credit for all fifty songs or just the medley as a whole?

BMI games part two: Have you ever thought to look up the song writing credit for one of these Britney songs to see what else that person has written? I'm picturing a middle-aged man that has the formula down writing for her, Christina, Avril and whoever the next one is. Not that there's anything wrong with that. I just doubt that it's teenage girls writing these songs about what it's like to be a teenage girl.

The thrill of the hunt:

Ben posted a great article today about the rewards of seeking out good music. My visits to BMI were solely for that purpose. Once I trust that an artist won't let me down, I must have everything that they've created. Ben is one of them. You're on your way. I picked up Alcatraz before ever hearing an MTX song. Brilliant. I then grabbed The Miracle of Shame right when it came out and now I'm ready to start digging into your back catalog. Judging by recent posts it will have to be Love is Dead.

I'd love to hear your take on what it means to you to "discover" a great album or artist and how you find them. I'd like to see this on the blog itself, if you have the time. I, myself, could go on for weeks.

I understand that you are not my monkey boy and you don't have to do as I ask. I just value your opinion.

Thanks.

I know that most of Avril's songs were written by a 3-person songwriting team called "The Matrix". Search for "The Matrix Songwriting" in google and you can find some write ups about them. Last thing I read said they are currently working with Britney Spears, Rick Martin, David Bowie (?), and Liz Phair.

About the question of the extent to which lyrics determine tunes, here's a near-perfect accidental test case-- a set of lyrics that, due to a misunderstanding, was set to music by two people on antipodal continents (one of whom had written the lyrics) within a couple of hours of one another:

http://interrobang.jwgh.org/songs/pumpkin.html

The tunes are quite dissimilar, but they're both kind of catchy...

good web page but do you know you are a google whack!!!!!