Slashdot Mirror
New Computer Program Determines "Hitability"
illuminatedwax writes "It looks like the process of homogenizing the mediocrity of Top 40 radio is going to be aided by a computer, according to an
article from the Music Industry News Network. Polyphonic HMI has developed a new program called Hit Song Science (HSS) and compares "underlying mathematical patterns" in current hit songs and compares them to a new song to determine if it will become a hit or not. Looks like we can expect even more of the same old junk being recycled for us on the radio, although the article claims that it 'will allow new sounds and styles to flourish.'"
Isn't that a valid assumption? I can't think of any top 40 "hits" of any genre that are different and groundbreaking. That's why I use an MP3 player in my car instead of a radio.
1) Bohemian Rhapsody
2) Smells Like Teen Spirit
3) London Calling
Music and Math are closely related, as anyone who's read Godel Escher Bach knows. Musical scores have themes that appear in many different variations such as canons (when a melody is offset in time) and fugues (more complicated than a fugue, read the book if you want to know).
I'm not acoustically talented, and I'm sure I couldn't recognize a fugue or a canon if I heard one, but I know that there is some music that I really like, and that sounds better made and more complete than others. I wouldn't find it hard to believe those songs have properties that a computer could pick out.
For example, have you ever listened to a song for the first time, and been able to anticipate what the next notes would be? I think on some level our brain recognizes patterns that we can't see conciously. With statistical analysis, a program could determine if more hit songs always follow a pattern or a specific pattern (easy to hum songs that get stuck in your head), or if more hit songs would break the melody and hit a note you weren't expecting (like those really mind-blowing high notes).
As a music lover, I would be thrilled if this application worked. It would really enhance websites that try to suggest other songs that you might like based on your favorite songs. In a lot of the music I like, the singer's voice gets deep and gravelly in parts. There could be bands that I hadn't considered listening to who match that profile, and a program like HSS coudl find them.
hey, how long till i can have it automaticly D/L the songs that i will [Mathematically] like?
Can I run the program on music I might consider listening to and rule out anything it approves?
Actually, this is useful on a person by person basis. I can tell it which songs I like, and it can pre-scan new music and decide what I'm more likely to enjoy.
Jason
ProfQuotes
Today, Kenwood announced a new model in their line of automotive head units that incorporates the fruits of their recent alliance with the Chinese software group SinOn.
SinOn is providing the AI side of the new MoDI car stereo that can be trained to recognize the owner's favorite style of music, and subsequently anticipate which streams, with permission, will be selected for play. The user simply puts the unit in training mode for approximately 10 hours, after which it is then set for autoplay. When set for autoplay, the software will prescreen all incoming audio streams and compare "underlying mathematical patterns" to determine if they match the listener's preference in music.
We tested the unit against the North Atlantic music satellite weave, giving it the suggested 10 hours of training. Once switched to autoplay, we travelled along the coast for two days, allowing MoDI to select music for us. We were happily surprised with the serendipity of track selection, and pleased with the seamless performance of the unit at all times.
We can report a positive experience with Kenwood's latest, and a recommendation for anyone looking for the newest in mobile audio while avoiding the pap of modern programmed listening.
Rivers uses a mathematical formula when writing his songs based on songs by several bands including Nirvana. As a huge Weezer fan, I'd have to say he's on to something. He's talked candidly about it in interviews. I'm at least fairly interested in what comes of this.
As far as the media telling you what you'll be listening to...
You've got a point, but it's slowly eroding away. Payolla (sp) is now illegal. With all the attention companies like Clear Channel have gotten for owning such a high percentage of the nation's radio stations could soon result in regulation. Then we've got those nasty little P2P file sharing networks lurking around with mp3z to download. *wink*
You've got to face the fact that these record companies and radio stations only care about the money. If they can run a program that will reliably tell them if song A is more likely to be a hit than song B... maybe they can spend less money on promoting song A and get the same results as if they had released song B with extra money for promotion. That's just common sense, man.
thundercatzlair
They cold call people and have them listen to 5 seconds of the song. This tortured person is then asked to rate the song 1-5. The music industry then takes all the songs that get 1s and 5s and discards them. It turns out that often when one group rates a song a 5 another will really hate the song and rate it a 1. So what the industry is really looking for is songs that score 3s.
I don't know about country music, because, thankfully, I've never worked in that format. Most other music stations do something like this, but in different forms. Sometimes it's calling people and asking what they think of the songs currently in rotation, i.e. "Will you vomit if you hear this Nickleback record again?" Other times they pick a panel of listeners, and have them listen to snippets of about 100 songs (normally 20-30sec of each), and rate them. The ones that rate badly among everyone are thrown out. When you're focusing on your listeners, you can be less concerned about the positive extreme.
The reasoning behind all this is that if you hear a song that you'd rate a 1 (hate) you're likely to turn the radio dial. But if you hear a 3 you're not likely to have any particular response at all -- thus you'll stay tuned in for more comercials.
Well, my friend, if you listen to stations that don't beg for money every five minutes (in addition to the millions of dollars they get in tax money every year), that's kind of the name of the game: hold the audience long enough so that they'll listen to some commercials. You do it by having good programming and good talent.
I guess that's why when you listen to "Classic hits of the [6-9]0s" you hear the same tripe over and again.
Ummm....not quite. Classics stations are safe. There is a certain segment of the population that has been under the influence of illicit substances since 1968. They'll dig Iron Butterfly until they die in about 30 years.
I'm surprised I didn't see this mentioned anywhere. I remember one of the particularly depressing things from 1984 was the music generating machine used to create music for the proles.
A machine that checks to see if a song is going to be a hit with the masses based on mathematics is not far behind a machine that will be able to generate a hit for the masses.
Creepy.
you can take the road that takes you to the stars...
The result was the coolest station I had ever or since heard. Dont know exactly what killed them, but i yearn for something half that cool among all the clearchannel stations i have to fight with.
Somewhere, something incredible is waiting to be known. -- Carl Sagan
... and as a hobbyist musician I love articles like this.
When the general public get sick of all the pop and 'reality' stars made for them, they turn to the underground, and this is where you'll find people who truly allow new styles to flourish.
All this Hollywood stuff is for chumps. If you want real music, and real musicians, just look for the underground.
It's out there.
; -- the corruption of government starts with its secrets. a truly free people keep no secrets. --
This is perhaps the most undeserved 5 score I've ever seen. People, ignore this definition. Quality of music is ABSOLUTELY NOT how well an artist follows some supposed rules of beat, harmony, melody, rhythm, and tone. If this were the case, then classical music would always be considered the best, since these measures of music originate from classical study. So this poster is contradicting himself by saying "I can even enjoy classical" as if it's some mediocre form of music.
The two most (only?) important things in music are whether an artist puts to sound exactly what is in his or her mind, and whether the listener derives value or enjoyment from the music. Quality is subjective. Is this poster throwing away the entire rythmic tradition of many African peoples simply because they don't have harmony and melody? Indian classical music is considered one of the most advanced forms of music on the planet, and they don't really have a concept of harmony.
Anyone slightly educated in music knows this poster is full of shit. He's talking about discriminating between different top 40 songs as if they are caviar. as if
LS
If you think of this in a positive way: Maybe the record companies will be more willing to give new artists that try something new a chance if this tells them it might be a hit. Ofcourse this is all dependant on how well this works, and the music alone is not always enough to make hit.
Maybe there's another application: I have been playing the same old cds for years, and prolly will do so for years to come, maybe they can make a version that can be trained to predict whether I will like it and recommend new songs/artists to me.
beauty is only a light switch away
This reminds me of the Negativland album "Escape From Noise." The first track on the record is a parody of a radio announcement telling you that the next track you will hear has been scientifically tested and engineered to be a big hit. I guess this shows how parody often becomes reality. Spooky. Great album, by the way, if you haven't experienced Negativland yet.
IAAL
Launch.com currently allows you to vote on a scale how much you like/dislike a song - then based on that it recommends other songs to you.
But it doesn't analyze anything in the acutal music.
For that, I would recommend FFT and backprop Neural Nets being added to the existing ranking methods that they have - but in the end, your own brain is likely better at it.
There are some odd things afoot now, in the Villa Straylight.
This is just more evidence of the growing irrelevance of record companies. As technology moves forward, the record companies seem determined to find ways to decrease creativity and thwart musicians, not promote artistry. This will prove a fatal approach, in my opinion.
Musicians can now create and engineer music in their own homes with a relatively modest investment. They can advertise and distribute on the web. By charging a modest sum to download the music, they could quickly out-earn the average 35 cents a cd they now make. When someone (Napster?) comes up with the appropriate delivery vehicle for this scheme, the music-as-big-business era will have come to an end.
Record companies ought to recognize this now and stop treating their talent as noisome middlemen. It seems like they start with packaging and marketing, and add in the music as an afterthought.
But all is not lost --- great musicians want to create great music, and people will want to hear it. You can't keep the two apart.
http://ob-la-blog.blogspot.com/
The Polyphonic site says:
Well, much of what attracts us to a particular song is found in the basic structure of the music. Particular rhythms, changes in key and certain
HSS visualization of an album superimposed into the recent "hit universe" melodic patterns define the psychological and very human response we all have to music.
This seems to imply that the reaction to hit songs is universal. I'm not ashamed to say that I absolutely abhor most of the hit songs on the charts nowadays. Um, "we all have" this "very human response"? My response is to change the station (which rarely actually happens, since I almost always just listen to my MP3 CDs anyway).
My point is, how can some software like this transcend people's individual tastes? I mean, sure, it wouldn't be hard to recycle some music to sell it wholesale to the masses, but it's pretentious at best to assume that some sort of program can accurately reflect my tastes at the same time as the average 12-year-old boy band fan.
--------------------
"Time is an illusion.
Lunchtime doubly so."
-Douglas Adams
David Borowitz
Yes, so long as "what I will like" = "what I like now".
Personally, while there may be some relationship between the two, I'll happily use my own brain, listen to stuff and DECIDE if I like it. It's actually pretty effortless.
-Styopa
Ok.. The conecept is a great CompSci project, in its own idea. Determining and matching patterns in music. GREAT! The bad part is they applied EVIL statistics to the game. Statistacs can be manipulated in sooo many ways. My problem is ok for this to work.. they'd have to survey all the top hits and make a Master pattern. Not only top hits from now.. but top hits from as far back as human archives have recorded, and stratified as popular. Then look and see what common ground you find. Remember.. taste changes over time.
Now my concern personally about this is.. I don't like mainstream music *most* of the time. There is a lot of crap out now, and has been out before that I am completely boggled as to why its popular. I'd say 90%. Mostly I many of the bands I like haven't seen much pop top 40 play. I don't try and be snobbish about it.. its just what I like. If its popular I don't mark it automatically off the list. I mean I'm ashamed to admit it, but I really like that Pink song, Party started.. or whatever it was called. That was a great pop song, and dance hall song really. On the other hand I really like listening to Mike Doughty solo and from his days with Soul Coughing.
My point is everybody is different.. I hate it when everything is playing to the lowest common denominator. I guess thats a cruel fact of life though.
Oh well chances are everbody will get bored of what the program determines as pop.. and they'll have to reprogram it.. thus the industry will still be behind the trends.. as always..
Who makes you Sig?
The way to write a hit song is to imitate what you here on the radio. So said Frank Zappa, in an interview shortly before his death. Listen to the current hits, and imitate them: nothing more, nothing less.
-kgj
You can bet, though, that there will be some Norah Jones soundalikes being pushed through the great Music Industry machine in the next few months, if they aren't already. And they probably will be teen sex idols to boot.
Considering the proportion of garbage on the air, it would be a fair trade. Still, a better use would be as a "Junk Filter", to decide ahead of time what I *wouldn't* like. (And remember, you need to check the junk filter every once in awhile to make sure the settings haven't gotten bollixed up.)
I think we've pushed this "anyone can grow up to be president" thing too far.
No.
Though I find it really spooky that they'd be anyone's yardstick for a music's goodness.
It's just a meta-effect of herd mentality, means nothing.
Tune out. Be yourself.
Keep your packets off my GNU/Girlfriend!
1. Theoretical system. This, I sense, could actually be a great technology. If there is a solid basis behind what we like and why we like it, I feel it would provide some great insights into culture. The record industry would issue music based on what people really do want to hear.
2. Empirical system. This would use data from previous hits and misses to predict how a current song will do. What I don't like about this is the prior database: it will be made up of the generic music the industry has been pushing for years. So it will not be making a decision based on how the masses really would like a tune, rather it would make a decision based on how the masses liked the other stuff they sold. In this way, innovative new styles will still take a long time to pry their way into the mainstream.
No I'm not trolling.
Sounds to me like Eminem is just Elvis all over again.
-- HG Pennypacker, wealthy industrialist and philanthropist
Food for thought:
(The songs I list here are my examples, you may disagree, just substitute appropriate songs for you...)
When I think back to the first time I heard particular songs, even without knowing who the artist was, I recall certain times I loved the song right away, for example:
Enter Sandman- Metallica
Still the One- Shania Twain
(I still enjoy those songs today.)
Then there was some songs that I thought were interesting, for example:
Informer- Snow
I'm Blue (ah ba dee aba dah)- still don't know by whom
These songs were interesting, not great but the third or fourth time I heard them I kind of liked listening to them.
(Now I hate listening to them.)
Then there's songs like Abercrombie and Fitch girls, which I always hated, and still do. I think there's an obvious marketing trend. The Abercrombie and Fitch song was hyped so much, that they "MADE" you like it, or at least they "MADE" the people like it who would call up and request it to be played, thus making it a hit.
The first group represents good songs that stand the test of time. The middle group represesnts something somewhere in the middle. Now all of these songs were top hits. How do you suppose a computer program will differentiate them? (Or does it matter? A hit is a hit.)
The daughter of a Sitar Player, Little Miss Shankar is merely a different kind of pre-packaged. Pre-packaged pseudo-intellectual pseudo-meaningful but still overpoweringly commercial music. A step above Britney Spears... but only one step.