This is an analysis of early House and Techno Music from Germany and the united states of AMErica (HOTGAME). Based on recording studio features, a Self-Organizing Map (SOM) is trained to explore similarities and differences between US-American and German house and techno music and their evolution. Some results are available on the techno analysis project site, particularly some interactive SOMs.
The project contains the SOM algorithm, the trained SOM, and the data normalization. To use it, simply download the HOTGAME corpus (zippped to about 10 GB), adjust the file paths in the jupyter notebook, and run it.
Feel free to rearrange the songs using a different folder structure. This way, you can explore, e.g., the sound of particular artists, regions, (sub-)genres, record labels, or alike. You can even include your own music in the SOM. Simply use the recording studio feature extractor from the HOTGAME feature extraction project, and analyze your own audio files (wav, mp3, m4a, opus).
Detail on the recording studio features can be found in:
-
Tim Ziemer, Nikita Kudakov and Christoph Reuter, "Producer vs. Rapper: Who Dominates the Hip Hop Sound?", Journal of the Audio Engineering Society 73(1/2), 2025, pp. 54-62, https://doi.org/10.17743/jaes.2022.0183.
-
Tim Ziemer, "Goniometers are a Powerful Acoustic Feature for Music Information Retrieval Tasks", DAGA 2023 - 49. Jahrestagung für Akustik, Hamburg, Germany, pp. 934-937, https://pub.dega-akustik.de/DAGA_2023/data/articles/000600.pdf.
-
Tim Ziemer, Pattararat Kiattipadungkul and Tanyarin Karuchit, "Acoustic features from the recording studio for Music Information Retrieval Tasks", Proceedings of Meetings on Acoustics 42(1), 2020, paper number 035004, https://doi.org/10.1121/2.0001363.