New Rising Media

View Original

Scientists discover the secrets to a successful pop song

Talentless singers rejoice.  University of Bristol scientists have created software that has the ability to analyse and determine the hit potential of a song.

Called 'Score a Hit,' data has been collected about tunes in the UK top 40 since 1961 for the purpose of using the equation, to see if it's estimations were correct and precise.  The program has been able to predict chart positions with roughly 60% accuracy through it's algorithm of evaluating 23 characteristics of the song such as danceability, harmonic simplicity and volume.  

To keep up with these changing weights of importance on these features of the music, artificial intelligence has been used to actively learn what traits are emphasised within popular music at every point in time.  The analysis has noticed a general trend towards music becoming progressively louder and easier to dance to.  Dance songs were popular in the 80s, and the 60s were rife with harmonically simple compositions.

The software is tricked, however, by songs that make it to number one via means other than the musical qualities (or lack thereof).  Means such as the massive amount of promotion that fuels X-Factor songs to the top of the winter charts, or the social media counter culture that puts select rock songs from the 90s to Christmas number one instead.

The video show how musical characteristics have changed over the decades: the further over to the right it is, the more detrimental to a song's success it is.  Check out the source link for more

 

Source: Score a hit