Fluid Corpus Manipulation (FluCoMa): creative and programmatic approaches to sound banks
A one-day workshop on music making with collections of sounds using machine listening and machine learning
November 24, 2021, 10:00-17:00
Free and open to the general public (limited to 25 participants): Please register here
Facilitator: Pierre Alexandre Tremblay (University of Huddersfield, UK)
A great deal of music making with computers involves processing, combining and playing with recorded sound in various forms, and as storage gets cheaper and archives larger, our collections become harder to manage and explore. Meanwhile, advances in signal processing and machine learning show promise for working fluently with these collections, but their musical potential is hard to explore because these technologies have not always been readily available in music software.
These musical questions are not, and can not be, solely technical. As repeated controversies to do with the applications of machine learning make clear, the increasingly urgent questions of how questions are formulated, and by / for whom will be of lasting cultural consequence. As well as equipping participants with a understanding of what these technologies can do, we aim also to help develop a critical perspective on their possible role in creative practice.
This intense one-day workshop will give participants a hand-on introduction to these technologies and their possible usefulness using the Fluid Corpus Manipulation toolkits, available for Max, Supercollider, and Pure Data. We will explore topics in signal decomposition and analysis; dataset exploration; and machine listening / learning in relation to musical tasks of curating and developing a corpus in the context of one's piece / instrument / system.
### The Fluid Corpus Manipulation Project
The Fluid Corpus Manipulation (FluCoMa) project is about trying to enable and animate development of the music potential of signal analysis, machine listening and machine learning technologies. We aim to do this by providing open source implementations in a range of environments (Max, SuperCollider, PureData and the command line), supported by workshops and learning materials, in the hope of seeding a sustainable community for exchange and exploration.
The project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No 725899).
## Aims
Participants will learn:
- non trivial aspects of decomposing, segmenting, and describing sounds
- how to explore this new functionality as a creative device
- basic machine listening and data analysis approaches to make sense of the newly segmented and described sounds
- how (or whether) these tools might play a role in their future practice
This workshop is hands on: each participant will be experimenting with the toolsets throughout. It will be followed by a few days of semi-supervised, optional creative coding in a room at CIRMMT TBC.
## Requirements
Participants need to come with a portable 64-bit computer (Mac, Windows, Linux) with Max, Supercollider or PureData installed, and headphones. Experience in one of these environments will be assumed.
About 2 hours of preparatory work will also be expected, sent at least 2 weeks before the tutorial: installation, and a few basic tests, alongside a video tutorial (below). Remote support will be available.
>System requirements:
Mac OS 10.8 or above; Windows 10 or above; Ubuntu 16.04 or later (or equivalent)
Max 7 or above; SuperCollider 3.10.3 or above; PureData 0.51 or above
### Facilitator
Prof. Pierre Alexandre Tremblay is a composer and an improviser on bass guitar and sound processing devices, in solo and within various ensembles. He is a member of the London-based collective Loop, and his music is also released on Empreintes DIGITALes and Ora. He formally studied composition with Michel Tétreault, Marcelle Deschênes, and Jonty Harrison, bass guitar with Jean-Guy Larin, Sylvain Bolduc, and Michel Donato, analysis with Michel Longtin and Stéphane Roy, studio technique with Francis Dhomont, Robert Normandeau, and Jean Piché. Pierre Alexandre is Professor in Composition and Improvisation at the University of Huddersfield (UK). He previously worked in popular music as producer and bassist, and has a keen interest for creative coding. He enjoys spending time with his family, drinking oolong tea, gazing at dictionaries, reading prose, and taking long walks. As a founding member of the no-tv collective, he does not own a working television set.
www.pierrealexandretremblay.com
============
# Teaser Videos
# Tutorial videos that I hope students will do before so we can jump right in more advanced concepts
# Other Links / Resources
https://youtu.be/MTWklm1oXWQ (Onset detection with FluCoMa used to drive synthesis processors)
# Searching a modular synth parameter space:
# Audio Classifier:
# Using FluCoMa tool inside REAPER with tutorials:
# FluCoMa’s Commissioned artists talking about their work: