Heavy guitar riffs, heroin addicts wearing flannel shirts singing about suicide. In 2014, they acquired Echo Nest, a company that specialized in these sort of services, to improve playlists and recommendations on their own services. Spotify gets these numbers using signal processing techniques. A more complete explanation of the metrics can be found here. There are eleven metrics provided by Spotify, but I selected three as the most significant for this exercise: danceability, which is a good way to assess the quality of the rhythm of a track, energy, which measures pace, loudness and noisiness, and valence, which is the overall mood of the song. Metricsīut, an important question: what are these metrics, anyway? And how does Spotify get these numbers? Dandy, by the harmless, almost childish Herman’s Hermits is followed by Somebody To Love, Jefferson Airplane’s strange, aggressive, drug fuelled, fast paced rock anthem. For example: the most shocking change in mood comes from 1966 to 1967 – possibly, the most revolutionary year in music history. Some of the results were surprisingly meaningful for an automated playlist. Then, I summed the three scores to reach a zeitgeist index – of course, I did multiply the negative scores by -1, so that one result would not nullify the other. To get to that result, I subtracted the metrics of each song to the average metrics for the corresponding year.
One of the ideas was to see the average score for each year in three of the main metrics (danceability, energy and valence) and find which song would be more similar to that ideal type – which I called the zeitgeist songs. And, the best part, they are available for anyone to use at Spotify’s API.įor this project, I selected over 6,600 songs that made the Billboard Top 100 Year End chart from 1941 to 2016, and, based on Spotify’s metric, I tried to measure how their qualities changed over these 76 years. They are numerical values to assess how danceable, how energetic and how positive they are their duration and tempo their key and mode and, also, whether they’re instrumental, acoustic, or live. The company holds metrics for all the songs it hosts. One of the ways to do so is by analysing how it evolved one year after another using data.
Think of the 1980s, what was the first thing that came to mind? To most of us, is the reverb of the snare, the artificial pitch of the synth, maybe When Doves Cry guitar solo – the music, always the music.īecause of that, to analyse what people were listening to over the decades is a way to tell the history society itself. Music is arguably the strongest mark of an era. An attempt to tell the history of popular music through data
0 kommentar(er)
