AI, “fat” sound and streaming
Today, we usually listen to music via streaming. Algorithms analyse our listening habits and suggest further music tracks we might like. And yet they don’t always find the right music to suit listeners’ tastes, because streaming services rely on what other users have streamed. A map-based AI can improve this system because it is able not only to classify genres by rhythm, timbre or melody but also to find other music pieces by identifying transitions to other genres. The bass line is an important parameter here. The louder and more prominent the basses, the “fatter” the music is. AI is therefore able to identify different currents in electronic dance music by the “fatness” of the music.
Mapping Dance Music with AI
The Kohonen map shows how AI sorts electronic dance music. The dots on the graph indicate individual songs. The different colours represent different genres.
Touch the dots and listen to the music!
Operate the buttons above the Kohonen map. You can see how the individual musical parameters are distributed (where yellow refers to a strong influence).
The graph to the right displays evaluations by experts on the ‘fatness’ of songs. It is noteworthy that over the past few decades, music has become increasingly ‘fat’.
References
Schmedecke, Lars (2021). Bass Drop! – Der Bass in der modernen, populären Tanzmusik (Bass Drop! – Bass in modern popular dance music). PhD Thesis. Hamburg: UHH.