Cam and I had an interesting email conversation this week concerning the commercial use of music analysis software to identify potential hits etc. I’d thought I’d post it here as a discussion point. What you YOU think?
— CAMERON WROTE TO HAL
I got signed up to the beta of a service the other day that’s like
Pandora on Steroids ( http://www.uplaya.com/ ). Rather than use
musicologists they’ve got some boffins to devise a set of dynamic
algorhythms which analyse music and put it into all sorts of categories
and find ways of linking stuff together.
Anyhow, as they have some smart artist pages I was wondering about
getting
all our Young Punx tracks into this, and also put them through their “Hit Song
Science” machine (http://www.hitsongscience.com/) to
let us know which songs they think have the highest potential of
being hits, and also analyse which bands we “sound like”.
If you could upload tracks and artwork somewhere, I’ll send him the
link and get them ingested etc.
Obviously the whole idea of a world where computers predict hits is
rather dystopian, but might be interesting to see if there are any
“hidden gems” in our back catalogue / coming up.
— HAL REPLIED TO CAMERON
God, these programs are the work of satan.
They are the one thing that could send the entire music industry to
its grave
A computer can accurately tell you that your song sounds like many
successful songs in the past, i.e. is not original.
When record companies start using them (and they do) all it
guarantees is a stiffling commercial environment where creativity is
actively discouraged, and all songs for promotion must conform to a
formula. No bohemian rhapsody. No higher state of consciousness. No
“Learn to recycle”
They are useful if you are dealing with formula based music, like
packaged pop, but beyond that, the ‘experts’ who go on about how
(assume nerd voice number 2) “well actually you WILL find that
statistically speaking 93% of top 10 hits conform to a blah blah
blah, irrespective of genre and blah blah blah”, totally fail to
understand that art moves forward through breaking rules and
inventing new rules, and will wonder why no-one buys their records
anymore, when they have computer designed them to be the same as the
records people already own!
hal
(actually it is an interesting analytical tool for seeing patterns
in past works. The problem is when it is mis-applied to attempt to be
a measure of whether something is GOOD or not, or to patronise the
public by assuming they will only be interested in something
previously proven. Unfortunately their own marketing material
promotes it in such a way.)
— CAMERON REPLIED TO HAL
Completely agree with the misuse of “HSS” to predict hits – when it
gets interesting is when you get their system to look at your iTunes
collection and get it to look at your music in a more leftfield way,
e.g. house music that sounds like metallica, classical music which
sounds like kraftwerk, or to arrange 100 tracks starting from gabber
and ending in easy listening, etc. Or putting in random variables
like Daft Punk meets Abba. Or for looking at your own music and
seeing what DNA it is made out of.
Might be interesting to see which TYP tracks have hit potential
purely based on previous hits which they sound like. Delivering
innovation and creativity in a way that’s palatable to most people
who have a low tolerance for new music, is all about embedding
surprises within the familiar. Same as with DJ’ing – got to play
stuff people ‘expect’ to get permission to go mental.
Commercial music is a commercial business and this sort of stuff is
inevitable. As I’m sure you’ll know, cultural creativity generally
comes from the fringes of society where people throw out the rules
because they’re doing it for love, not money. This then gets pulled
into the mainstream and then reinvigorates it. The issue is that when
businesses get put under pressure then get more cautious and take
fewer risks, thus precipitating their own demise as doing the same
thing gets the same results…
