Slashdot Mirror
Chords To 1300 Songs Analyzed Statistically For Patterns
First time accepted submitter hooktheory writes "We looked at the statistics gathered from 1300 choruses, verses, etc. of popular songs to discover the answer to a few basic questions about pop music. First we look at the relative popularity of different chords based on the frequency that they appear in the chord progressions of popular music. Then we begin to look at the relationship that different chords have with one another. To make quantitative statements about music you need to have data; lots of it. Guitar tab websites have tons of information about the chord progressions that songs use, but the quality is not very high. Just as important, the information is not in a format suitable for gathering statistics. So, over the past 2 years we've been slowly and painstakingly building up a database of songs taken mainly from the billboard 100 and analyzing them 1 at a time. At the moment the database of songs has over 1300 entries indexed. Knowing these patterns can give one a deeper more fundamental sense for how music works" This reminds me of the work done by two Rutgers grad students last year trying to find a formula for a hit song.
Pachelbel Rant
Just what we need, a database for the RIAA to use to play Whack-A-Mole on upcoming songwriters for 'copyright infringement'. There are only so many chord progressions possible. This will allow the holders of the eternal copyright to sue somebody because the chord progression they wrote mirrors a song their grandparents heard in the womb and thus infringes.
Yet another argument for 7 year copyrights. Too bad we can't convince our Congresscritters of this...
Understanding the scope of the problem is the first step on the path to true panic.
...they were all either:
Right? Or maybe that's just pop songs from the past twenty years....
Check out my sci-fi/humor trilogy at PatriotsBooks.
Don't need no computer analysis for that.
http://www.youtube.com/watch?v=5pidokakU4I
My second theory of the brontosaurus, that which belongs to me, and is mine, and mine alone, can be stated as follows...
This does not "give one a deeper more fundamental sense for how music works", but gives insight into what tends to be popular in the USA. Now, you could say that that has a major effect on the popular culture of the rest of the world because of the way it's exported, but still it's not necessarily the fundamentals of music in general.
Axis of Awesome - 4 Chord Song
http://www.youtube.com/watch?v=5pidokakU4I
Oblig.
with a fast enough computer could you programatically write sheet music for every possible song ever, copy write it, and the sue away? Sure, you'd get dinged for the already copywrited stuff, but you could just cross reference off future product, since you know own all of music.
Hi! I make Firefox Plug-ins. Check 'em out @ https://addons.mozilla.org/en-US/firefox/addon/youtube-mp3-podcaster/
Don't let the RIAA get their hands on this technology. They will just start suing all the off-label artists for stealing their valuable "copyrighted" chord progressions.
Anyone with a rudimentary understanding of music theory could have told you that. It's not just pop songs either, composers have been using that progression for hundreds of years in every conceivable genre. It goes back to the very beginning of what we would consider standard harmony, circa late 16th century, maybe earlier.
Strangely, they found that a vast majority of popular music from the last 60 years seemed to break down into a pattern of 48 beats using three repeated chords (and variants thereof), such as: AAAADDAAEDAE. Odd that.
I'm betting the majority of then were G C D... Oh wait, that's Country music and they were analyzing Pop!
The brits alread did this, years ago. Watch & Enjoy:
http://video.google.com/videoplay?docid=4423562351831425828
Good security is based upon reality and common sense. Common sense is a function of having common knowledge.
So does this mean that if I copywrite these in a 4 cord song called happy birthday 4. I will rule all the cords with my 4 fundamental mind controlling cords for life?
This is B.S. there's plenty of people who don't give a shit about these cords, I guarantee most ambient/electronic beat"less" music and much classical doesn't fundamentally really on these.
I'm talking about artists (or groups) like Jon Major Jenkins, Numina, or Max Corbacho.
Anyway this further illustrates why "music" should not be owned at all. Almost virtually any of the big record labels are producing is non-unique from a scientific or even layman's perspective.
Anyway I would be more then happy if no one could produce music with these 4 cords without a hefty license, because the majority of it is crap anyway (sarcasm).
I think the folks over at The Onion have figured out the pattern to a hit pop song. I present as evidence, K'Ronikka with Booty Wave:
http://www.youtube.com/watch?v=lmCjJ0VBjjU
Watch it on YouTube.
Ant(Dude) @ Quality Foraged Links (AQFL.net) & The Ant Farm (antfarm.ma.cx / antfarm.home.dhs.org).
What kind of software did you use for your analysis?
What kind of chord information do you see when you analyze something like one of the sections of Vivaldi's For Seasons pieces? I'd say borrow a Vivaldi score and then run your analyzer on those blocks of violins that are sawing away in something like the Summer or Winter pieces. The Vivaldi tunes are really "wall of sound" or "wall of musical excitement" pieces. But underneath all the flashy richness of multiple violins playing I wonder would your analysis system find just one simple chord progression, just like a popular tune?
Can your analysis software output detail about what specific musical notes are being played and when one note stops but other notes continue?
Regarding the Vivaldi and other classical tunes I hear on the local radio station, sometimes I just stop and say wow that sounds just like a Jazz riff. Where did I hear that before? I don't understand... I just started hearing things differently lately or else the 8 am DJ on KDFC is an imp.
Suuuure, just ignore the markup-like ChordPro format that many tab/chord sites use. It's only been around for 15+ years.
patent the patterns and then claim you want 5 trillion dollars for using it.
HA no more riaa problems.....
i'm not sure how far you'll get trying to own all music related IP if you're unable to spell Copyright or Chord correctly...
I prefer organic processing devices. The music sounds so much better. I hate when they crash. *sobs*
I'd go on a Vegan diet but the delivery time from Vega is too long. --brownkitty
It really doesn't matter, so long as you have a good marketing department.
Just about every pop song for the past... eternity... has used the same pattern: I, V, vi, IV. It's hilarious how bad it's getting. For instance, look at the preview of Coldplay's "The Scientist" on MusicNotes. The original key is F Major, so we'll work off of that. We start off with a Dm chord, thusly iv (minor sixth). Transition into Bb Major, so we get IV (major fourth). Then down to F, so I, and finally to C, so V (major fifth). They just shifted the pattern two chords over.
to consider that music needs ONLY a succession of notes happening in predefined patterns in order to please people !!!
Couldn't he at least have used that system and then told us the most common chord sequences as well as the most common keys...
Donald 'Duck' Dunn: We had a band powerful enough to turn goat piss into gasoline.
The science fiction author Charles Sheffield wrote a story about a similar idea in the late 1970s called What Song the Sirens Sang. The protagonist is a journalist investigating a politician who has come seemingly out of nowhere and is about to be nominated for president. He discovers that the secret to the politician's success is that he has developed a theory of communications that allows him to combine words and music to evoke optimum emotional responses. Check it out, it's a short read and very good.
cheap labor conservatives - they want to keep you hungry enough to be thankful for minimum wage.
A recent study of the "Freie Universität Berlin" of trends in US charts suggests, that pop songs got slower over the last 50 years and use more minor chords. Doesn't mean that society got sadder, the study explains, it only shows that we listen to more ambivalent stuff and are able to enjoy even sad emotions.
I'm sorry, but the reserch won't tell anything interesting. For a song it is much more important, HOW it is performed, not WHAT is performed. There is no (naive) formula there, that a computer could analyze. The success of a song has everything to do with the charm of the artists and how skilled the musicians are and how it is arranged and so on. The chord progressions are irrelevant. Look, this song had only 1 chord and it was a huge hit. So what now -- we start writing 1-chord songs and every monkey could be a star? Sorry, but no.
The coolest app ever would be one where you hum, whistle or sing a few bars of a song you know, or almost remember, and it identifies the song for you. Of course said app would have to know all the songs in advance in order to find a match. Yeah, that would go over big with the RIAA.
This seems to overlook both regional tastes and language.
For example, myself and 3 other people I have on twitter like to listen to Korean and Japanese songs along. I like to listen to the 'Glee' cover versions of a few songs that I'm not fond of the regular artist's version.
But bravo for trying to figure out what the 'perfect' song americans like, even though it would probably shift tastes were it ever to be produced.
Actually the chord progression of the two are similar.
Table-ized A.I.
So the juicy stuff comes next month?
I did something similar on a smaller scale recently for a still-in-progress hobby auto-composing app using cSound and scripts. I took chord progressions that I personally liked and built a chord pairing chart something like:
A --> C // note: m=minor
Am --> F, Em, Dm
A# --> G, D#, Am
B --> D, F
etc...
The first chord is the "lookup chord" and the second is a list of candidate "good" chords. In a loop it produces a chord sequence. It's almost like a Markov chain, but so far without probability weights.
I also considered a 3-way lookup to give a "good" next chord based on the prior 2, but that's a later experiment.
I notice some gaps. If you transpose to base it on C keys, then the A, B, C#, F#, and G# chords were rare altogether.
(Note that one can transpose the chords, and that's why I didn't use Roman notation. Sample has dummy values only; if you try them, they'll sound like sh8t.)
Table-ized A.I.
"we look at 1300 songs to discover some answers... It was hard because the data wasn't good. We entered the data ourselves." And that is it???? That is your entire summary? What about like maybe telling us what the article is actually about like what were some of the answers they discovered?
Even if chord progressions were copyrightable, one could still "find multiple commercial songs using the same chord progression" among songs at one label, or if if the various major record labels were cross-licensing their chord progressions while forcing indie labels to pay up, much as MPEG LA members cross-license video patents.
I remember meeting a guy [circa 1992] who had a consulting business based on doing just that. He would put a [suspect] music CD into his CDROM drive and [with custom software he wrote] have it analyze the note sequences looking for fragments that matched fragments of his clients' songs/catalog. IIRC, the criterion was either 11 notes or 11 bars [I can't remember which] of music.
It was probably 11 notes, seeing as Bright Tunes v. Harrisongs had George Harrison losing over nine notes. Nine notes is already short enough to produce many coincidental matches among existing songs in the repertories of BMI and ASCAP. So what steps should a singer-songwriter take to avoid such copyright trolls?
Yes, the current length is insane [and unconstitutional IMHO]
Your HO matters not in the real world because you aren't five Supreme Court justices. Eldred v. Ashcroft.
I don't think there'd be enough matter in the universe for [copyrighting everything]. [...] Even for very short songs that's a lot of possibilities.
A judge isn't looking for the whole songs to be identical; he's just looking for the songs to be "substantially similar". This cuts down to comparing the melodic hooks, as shown by the cases listed here. A nine-note hook was deemed an infringement in Bright Tunes Music v. Harrisongs Music, so let's go with that.
Model a "note" as a duration plus a pitch interval to the next note. There are seven distinct intervals in a diatonic (major or minor or modal) scale. Notes can be short or long; the performer's exact timing does not change the fundamental character of a melody. This gives fourteen possibilities for each note. But the last note does not really have a duration, nor does it have an interval to the next pitch because there is none. With eight duration/interval combinations, you end up with 14^8 possibilities, or about 1.48 billion distinct hooks. That's fewer than one for each person on the planet.
"Many tab/chord sites" aren't authorized by the songs' music publishers. Apart from the copyright issues, this means the music publishers have no opportunity to exercise quality control, making the tabs less likely to accurately represent what the musicians performed.
Since I was a young boy in the 80's I've always thought that it would be possible to create a hit using a computer based on analysing all hits.. It's an easy way to create new songs you certainly would like.. Now with current technologies it really should be a breeze... I can't believe I would be the only one who ever thought up such an idea...
For a song it is much more important, HOW it is performed, not WHAT is performed.
Not to a judge in a copyright suit. Judges strip away the performance and look at the sheet music.
They accept user-contributed analyses and etc. Check.
They have no visible license (that I can find) under which user contributions are made. Check.
They do not provide any way (that I can find) to download the database. Check.
Made of fail.
Just analyze "Pumped Up Kicks" - Foster the People. Crack for your ears.
On of my first class teacher told us: You are here because you have some instinct about creating interesting designs, you will now learn the science behind it to make it happens on demand... it really got into me. I still rely more on my instincts though.
Tomorrow is another day...
Anybody who has studied music composition for a short period of time will understand why the results he found are obvious. There have been countless books written on the subject... one that has an interesting take on it (and the author of this 'study' should read) is:
How Music Really Works
http://howmusicreallyworks.com/
Maybe they could have tried on guitarpro files? I'm not sure if they can be read by 3rd party but those a analogous to midi files.
I think charting the path to tonal ambivalence in pop music might be helpful!
V - I - V - I etc
to the
I - vi - IV - V
to the
I - iii - vi - ii - V - I
and the Most Ambivalent
bVI - bVII - I !
etc.
I'd like to see a visual representation of songs over time and their "distance" from the I - V polarity,
a.k.a. how "concealed" the swing of the I - V pendulum is,
a.k.a. if we are trying to eschew that duality in pop music.
Here's the one and only progression:
I, I, IV, IV, I, I, V, IV, (repeat until near the end, then throw in a flat-seven)
https://app.box.com/WitthoftResume Code: https://github.com/cellocgw