Unstable Audio: Code Bending Text-to-Music Generation

Collins N. Unstable Audio: Code Bending Text-to-Music Generation. In: 2025 AES International Conference on Artificial Intelligence and Machine Learning for Audio. Audio Engineering Society; 2025.pp. 15:1-7. Available from: https://aes.org/publications/elibrary-page/?id=23004

@inproceedings{Collins2025_23004,
  author = {Collins, Nick},
  title = {{Unstable Audio: Code Bending Text-to-Music Generation}},
  booktitle = {2025 AES International Conference on Artificial Intelligence and Machine Learning for Audio},
  pages = {15:1-7},
  year = {2025},
  month = sep,
  publisher = {Audio Engineering Society},
  url = {https://aes.org/publications/elibrary-page/?id=23004}
}

TY  - CPAPER
TI  - Unstable Audio: Code Bending Text-to-Music Generation
AU  - Collins, Nick
T2  - 2025 AES International Conference on Artificial Intelligence and Machine Learning for Audio
SP  - 1
EP  - 7
M1  - Paper 15
PY  - 2025
DA  - 2025/09/02
UR  - https://aes.org/publications/elibrary-page/?id=23004
PB  - Audio Engineering Society
LA  - en
AB  - Text-to-music models provide new audio generational and transformational capability, though their control is not a precise art. Embracing the aesthetics of `code bending', glitch musicians' approach to software manipulation and the sounds of noise music, the open source Stable Audio 1.0 model is manipulated to enable radical new audio outputs. The domain of historical late 1980s Detroit techno is used as a constrained arena for generation; a comparative study is then undertaken of changing model parameters across all 745 layers of the model. Adding random noise to layer parameters is more likely to lead to interesting results than zeroing (pruning) connections or randomly reconfiguring parameters throughout the model, though all manipulations had some chance of creating interesting outputs. Further potential code bending projects for text-to-music engines are discussed.
ER  -