I’ve been looking for a fun example to push the boundaries of what is possible when doing data-driven journalism in our Online News unit this semester. I used Skrillex in a lecture last year to discuss affect and popular music (Lawrence Grossberg’s work is good on the way affect can be analysed in terms of ‘mattering maps’, but also check out this journal article for a different kind of engagement). Earlier this year someone posted this capture of comments on Skrillex’s Facebook account regarding the quality (or absence) of ‘the drop':
Mark Richardson at Pitchfork (music site, the bastion of indie music etc) had this to say about Skrillex, his fans and these Facebook comments:
The responses were edited down from hundreds of comments, many of which had Skrillex fans mirroring his praise of the tune. But the reason why it’s funny, and why it’s been passed around so much, is clear: These bass fiends have no ear for electronic music genius. They just want that drop.
So what is ‘the drop’?
Typically, the percussion will pause, often reducing the track to silence, and then resume with more intensity, accompanied by a dominant subbass (often passing portamento through an entire octave or more, as in the audio example). It is very common for the bass to drop at or very close to 55 seconds into the song, due to the fact that 55 seconds is just over 32 measures at the common tempo of 140 bpm.
Or urban dictionary:
The part of a dubstep tune where it gets so incomprehensibly filthy that one cannot fathom – therefore, ones mind explodes.
Person 1: “Yo dude, check out the drop in this banger”
Person 2: “Holy shit dude”
The drop is the when the beat kicks after a duration of anticipatory build up (‘intro’). The relative value of the drop or the intro is often debated (sometimes it is the ‘bounce’ that wins out). Dubstep is also known for the ‘wub’, check out this application of the below-mentioned Echo Nest API, the Wub Machine. The results can be truly horrific:
I downloaded the free iPhone app and created a wub machine dubstep version of the They Might Be Giants’ track “The Bells Are Ringing”. I laughed so hard I almost did rofl.
Transversal blocks of musicality
What I find fascinating about the discourse of music enthusiasts and fans around ‘the drop’ is that it is largely congruent with popular music discourses at different points in history referring to the ‘swing’, the ‘riff’, the ‘beat’ (as in ‘house’ or ‘break’) and so on. Each of these elements describes a particular block of musicality that is repeated in different ways within specific genres of popular music and within specific scenes (here I am using Will Straw’s influential definition of a scene).
I don’t know what to call these blocks of musicality in general; I am sure that musicologists have a term for it or someone will invent a term. I am thinking about them following Foucault’s concept of the ‘statement’. A ‘statement’ is a kind of singularity in discourse: the distribution of statements in an archive characterises the field of (onto-epistemological) positivity for articulating ‘truth’ in scientific discourses. Although music scholars have pointed out that discursive repetition is different to ‘musematic repetition’ within an individual song, I am describing something else.
The distribution of these blocks of musicality characterises a field of (onto-affective) positivity as a condition of popular musical appreciation. I am not talking about whether or not a track is ‘good’ or ‘bad’, but that these blocks of musicality will serve as the affective condition (in the philosophical sense of causality) of musical appreciation. Within genre studies, we’d call the drop a trope of the genre dubstep. I am trying to push it a bit further however, because genre studies is largely concerned with complexities of cultural typologies. What I am interested in is the affective dimensions of these blocks of musicality and how they come to organise listening practices.
The different blocks of musicality have different affective qualities. The drop combines anticipation and a pitch of intensity. Anticipation can have negative affective qualities (dread) and positive (‘excitement’), with popular music associated with the latter. The distribution of the drop as a differentially repeated block of musicality is also a distribution of these affective qualities through the communities of practice (online, clubs, etc.). If this seems like an overly convoluted way of saying that beats are dropped in clubs, you’d be right, but I am not (only) saying that. I am suggesting that ‘the drop’ cuts across music, the bodies of listeners and the discourses of music reception (Pitchfork, or any number of other music appreciation sites).
There is a transversality to these blocks of musicality that transcends a purely musical interpretation of them. What if ‘the drop’ became popular not because of the sonorous dimension of its musicality, but because of the shared (ie social) distribution of anticipation and pitch of intensity felt that moves across a community of listeners? You not only ‘hear’ the drop, to paraphrase Adorno, you ‘hear’ the everyone-else-hearing-it. There is a social dimension of the block of musicality present in every ‘drop’. I could imagine a ‘media archeaology’ of such blocks of musicality, as a way to examine the composition of power relations characterising popular music scenes (as well as Straw’s categories such as nostalgia, etc.). The social dimension of ‘the drop’ is accidently captured in the above quoted Urban Dictionary definition. So beyond academic research, what if you could analyse the character of ‘the drop’ not in strictly musical terms, but in terms of its musical capacity for sociality as a predictor of popularity?
The possibility of data-driven music journalism?
There is UPlaya that carries out an algorithmic analysis of music submitted to compare it to previous ‘hits’ to assess whether or not it fits with its predictions of success based on previous popular music. The big player in parsing music and a great deal of associated material is The Echo Nest API. The Echo Nest is described as a ‘music intelligence platform’ and boasts 5 billion datapoints with 30 million songs and 1.5 million artists. Here is a Slideshare presentation where one of the creators of Echo Nest walks through its creation and the “pitfalls and promise of music data”. One of the more amusing uses of the Echo Nest API is this project called The Pitchfork Effect. The project in itself is very cool. I find it amusing that data analysis tools are being used to analyse the qualitative process of judging music and sound aesthetics (as well as whole range of other issues to do with political economy of music, i.e. ‘indie’ used to mean something beyond an aesthetic/marketing category). But I am thinking of something else.
Say, for example, I wanted to analyse Skrillex’s music and reception through the concept of ‘the drop’. Is an algorithmic analysis of his music tracks possible, in terms of when each track ‘drops’ and the quality of the ‘drop’? Certainly. It would be a question of exploring the relation between the anticipatory build up (‘intro’) and then ‘drop’ when the beat kicks. I’m interested in not only an analysis of the music itself but locate the music in patterns of reception. The question here would be, how does ‘the drop’ ‘drop’ (in communities of music listening practice)? Similar to Skrillex’s computational music producing ‘drops’, this would be a computational music journalism analysing meta-drops. (::diabolical cackle::)
Data could be gathered a number of ways including by way of doing a basic sentiment analysis of online commentary about the quality of the drop or number of ‘plays’ of a given track through online sites such as Last.FM. Combining both sets of data we could look for patterns/correlation between the qualitative reception of the socio-musicological ‘drop’ and the algorithmic analysis of the ‘drop’ as a block of musicality. The thesis could be tested against historical examples of ‘riffs’ and so on using different algorithmic measures for a media archaeology of such transversal ‘blocks of musicality’.
As a start here is the ‘fantracker’ data vis of all activity tracked by Musicmetric: