This accessible page is a derivative of https://www.researchcatalogue.net/view/2938321/3620180 which it is meant to support and not replace.
Upics, Sun
Several of the recordings feature dissonant clusters of pitches that slightly modulate and shift in subtle ways. The foundations of the piece ‘Sun’ are these clusters combined with textures that develop according to the analysis data. The piece has a very block-like surface that stays more or less consistent throughout while the more distant sounds seem to be continuously shifting. The focus is local and less based on discrete bits compared to many of the other pieces. The main cluster does drift off towards the end and is replaced with shorter sounds that have been found close in similarity through the data analysis. Again, the queries work to select but also as a surprise since the main dissonant cluster is a combined and granularly extended sonority made through the different interpreting processes.
Audio description: Sound file of Sun. Several of the recordings feature dissonant clusters of pitches that slightly modulate and shift in subtle ways. The foundations of the piece Sun, are these clusters combined with textures that develop according to the analysis data. The piece has a block-like surface that stays more or less consistent throughout while the more distant sounds seem to be continuously shifting. The focus is local and less based on discrete parts, compared to many of the other pieces. The main cluster does drift off towards the end and is replaced with shorter sounds that have been found close in similarity through the data analysis. Again, the queries work to select but also as a surprise since the main dissonant cluster is a combined and granularly extended sonority made through the different interpreting processes.
Click on https://www.researchcatalogue.net/view/2938321/3620180#tool-3622250 to listen to the audio.
Image description: Cover image from the album UPICS.
Click on https://www.researchcatalogue.net/view/2938321/3620180#tool-3622263 to see the image.
Grouping Discoveries
Once an analysis process is complete, the archive is accessed. It is analysed by machine-listening processes and then paths are proposed to further explore. The archived material is thus processed by listening agents that instantiate a flow of events unique to that archive. The dynamic of these listening agents is a central point of exploration; they not only listen to each other and themselves but also form intricate behavioural chains that incorporate manual input and further transformation. The approach highlights specific categories. Groupings of similarity, of difference, of duration, and date. Sound becomes a key factor in guiding autonomous algorithmic behaviour and control mechanisms. The tool facilitates a deeper understanding of the archive, enabling the discovery of material through various processes linked to the data. In this setting, the environment and context for the audio processes and agents are defined by the sound archive itself.
The project is an example of a data-driven approach to composition but also a way to discover sound material through the analysis of data. The compositional approach involves accessing a collection of sounds, regardless of their history. To analyse them and reform them through different reproducing algorithms. There is no connection between how the material was created and how it would be used later. This separation was liberating since it allowed me to realise and understand the UPIC recordings differently from my previous attempts. The recordings made fifteen years ago now become accessible and open through these processes. By actively querying and exploring the material, unique selections are accumulated. Employing diverse algorithms to interpret the material proved effective in altering their context and highlighting their intrinsic qualities. This approach allowed the recordings to stand out, not just as historical artefacts detached from their original context, but as dynamic, contemporary ways of engaging with sound material of the past.
Upics, Fog
The querying methods, interpreting agents and processing methods used for the pieces all became part of the SNDArchive system and repository that was completed and made open source. My approaches to examining sound material have evolved into established procedures accessible to other people and open for further development. While the system itself represents a significant outcome of the research, the most crucial element lies in the interplay between the code and music, and how they have mutually enriched one another.
In early 2023, six of the compositions produced during these analytical sessions were merged into a collection, ‘UPICS’.
Audio description: Sound file of Fog. The querying methods, interpreting agents and processing methods used for the pieces all became part of the SNDArchive system and repository that was completed and made open source. My approaches to examining sound material have evolved into established procedures accessible to other people and open for further development. While the system itself represents a significant outcome of the research, the most crucial element lies in the interplay between the code and music, and how they have mutually enriched one another.
Click on https://www.researchcatalogue.net/view/2938321/3620180#tool-3622264 to listen to the audio.
The primary outcomes here contribute to a broader artistic research discourse by proposing composition as a mode of searching, an exploratory practice shaped by systems, material histories, and shifting contexts. The approach foregrounds a relational process in which form emerges through interaction with archival materials, algorithmic procedures, and the constraints they entail. In this view, the past is not simply recalled but restructured through acts of framing, selection, and translation. The project thus aligns with broader concerns around how knowledge is produced through practice, how tools shape thinking, and how acts of reactivation can generate new configurations of meaning.