The founder of one of Brisbane’s most successful startups, acquired music platform We Are Hunted, has shared his thoughts on the state of personal music app design.
Released last week, Stephen Phillips’ “Why Do I Like This Song?” is a collection of his insights and research into music software. It includes information on song similarity techniques, the psychology behind music, app design ideas, future projections and a dense collection of background reading.
Phillips helped launch We Are Hunted in 2009, a discovery platform that charted and streamed popular online music. It opened to over 100,000 visitors in its first week and was later acquired by Twitter in 2012, where Phillips continued working on Twitter Music until leaving the company in 2014.
“While a song has objective attributes that can be measured, it is the subjective attributes, unique to each listener, that ultimately determines likability.” – Stephen Phillips
Phillips begins with forms of intelligent music software: “expert” for lists made from descriptive reviews, “social” for charts sourced from listening history, and “acoustic” for analyses made from tempo, timbre, melody and the like.
He also describes how We Are Hunted combined the most valuable techniques by scanning song reviews from music blogs (experts) for the most common descriptive phrases (social).
But in discussing these services’ “quality ceiling of about 80%,” Phillips readily acknowledges the drawback of music tracking software; subjectivity. Here, he significantly pulls focus from the music to the listener, and moves on to the old chestnut of why we like a song.
“When we listen to a song, and we are able to correctly identify and resolve it’s patterns our body rewards us by releasing chemicals that make us feel pleasure. We remember songs that make us feel good.”
While the most lacking in technical advice, the section encompassing the subjectivity and psychology of music is, objectively, the most enjoyable.
Writing as if each of these were a given, Phillips waxes about music as a game, a drug, a mating advantage and a dual stimulant/distraction. One of his most valuable observations is the role of representation in music, in terms of everything from gender, ethnicity, and social status.
Phillips also details the social role of certain genres:
“If you are powerless in your life, powerful music will appeal to you. Teens from broken homes or difficult social environments (bullied) are attracted to the power (bass) of Hip Hop, Metal and Hard Rock.”
All this builds to ideas for personalising music apps, such as on-board initialisation, voice control, and adaptation to age and location. The concept of initialisation, where the service learns from the user, is currently being implemented by headphone developers Audeara.
Before listing his research, Philips has two final thoughts on where the music industry is headed: an increase in holograms (which are already terrifyingly now), and further neuroscience research, as a way of further understanding how the brain responds to music.
The piece is an excellent resource for anyone involved in music software, and you can read it in its entirety here.