Retro Electro sonifies David Bowie Lyric Emotion up to 2016 using Acid House Bass-line (Cut-off = Arousal, Resonance = Dominance)

David Bowie Lyric Emotion up to End of 1980s Sonified with Hyperspeed Piano using pitch for emotional positiveness and tempo for emotional arousal

David Bowie Historical Album Sales Sonified using Drone Pitch, with Repeated Pulse of Time-compressed Album Sales Data

Wired, UK  “Composer Alexis Kirke and electronic music pioneer Martyn Ware have analysed numerical data relating to David Bowie’s album sales, lyrics and music features and transformed them into musical ‘sonifications’…[Kirke] met Ware — a producer and founding member of Heaven 17 and the Human League — when they were on a panel together at the BBC. “He is an even bigger lover of Bowie than myself. Bowie went to an early Human League gig and said something about it being ‘the future of music’,” Kirke explains.”

“When Kirke was asked by the V&A to work on something for their upcoming Bowie exhibition, and called in Ware to work on the project. You can listen to the compositions at the Bowie-themed Friday Late at the Victoria & Albert Museum on 26 April, which is part of the David Bowie Is exhibition, running until 11 August 2013.”

“Kirke and Ware used statistical analyses of elements such as the changes in emotional content of David Bowie lyrics over the years, or the prominence of major and minor keys in his songs. These are then mapped onto musical features such as tempo, pitch and loudness and transformed into their own musical compositions. The aim was to show how Bowie’s oeuvre developed over time.


“To analyse the emotional content, the pair use the ANEW database, a scientific database of emotionally-annotated words, which are given scores based on their emotional positivity (valence) and emotional physical intensity (arousal). “I created a database of Bowie’s lyrics going from The Laughing Gnome up to the end of the Tin Machine era. These were then searched for key emotional words. This is a simplified emotional analysis in the sense it doesn’t look at the context in which the words were used. But such non-contextual textual emotion analysis is not uncommon. The resulting valence and arousal measures were then low pass filtered in an attempt to find longer term patterns. To our surprise, we found a broader cycle as well as some shorter cycles in between.”