I have developed some tools in order to solve some challenging problems I have encountered during my activity as a composer.
The first tool has been developed in co-operation with Jianming HUANG and Rémi GERFAULT over a period while I was working at University of Angers (F), during the academic year 2011-2012. Its purpose is to understand whether small melodic fragments can be played simultaneously so that, by repeating the fragments "k" times ("k" to be defined by the user), no notes will be played (started) at the same moment.
The second tool is used to visualize parts of the musical structures, as discussed by Luigi Verdi in his book "Caleidocicli Musicali". It has been developed by a student I supervised on a final assignment for a computer programming course at the University of Angers, 2012. It has been successfully used to compose my "Some small Kaleydos". (The terminology in any text contained therein has been written in a French.) For an outline of all terminology please refer to the book by Luigi Verdi and to the file "projet.pdf" contained in this archive. (The book is published in Italian, and the file is in plain French, but should not be too difficult to understand).
The third tool implements a basic Tabu Search algorithm to explore the block designs graphs defined in the paper, "Musical Experiences with Block Designs", by Franck Jedrzejewski, Moreno Andreatta, and Tom Johnson. It has been implemented under my supervision by Loic Venerosy during a stage at LERIA, University of Angers (F) 2012, and it has been used to compose something I wrote, the morceaux, "Some Small Taboos".
As a composer, I have always been interested in the computational (and algorithmics) aspects of music composition. This interest comes from my studies in Artificial Intelligence and from my research work over the last few years. Hence, I have investigated the computational models used by composers from the "Rinascimento" (including Mozart, Beethoven, until the explicit utilisations by Xenakis), and I have been attracted by the body of work produced by "Futurismo" in Italy, the first artistic movement which puts explicitly the machines (and, although naively, automatic thought) at the base of their creations.
Amongst the approaches I have been investigating, Block Design have a special place: they represent a branch of "set theory" that, under some specific parameter-settings, lead to a new way of organising the "totale cromatico", different from serialism and, to some extent, complementary to it. The basis of this technique can be found in the paper "Musical Experiences with Block Designs", by Franck Jedrzejewski, Moreno Andreatta, and Tom Johnson (in Mathematics and Computation in Music: Second International Conference: pp154-165).
The approach uses some graphs to represent the musical material. My composition starts from the same graphs reported in this paper, but applies a different algorithm to explore them: Tabu Search. I shall not go in to detail of the algorithm here. (See Fred Glover (1989): "Tabu Search - Part 1". ORSA Journal on Computing 1 (2): 190‚Äì206), but, in brief, it is a non deterministic algorithm, meaning that if we run the same algorithm, starting from the same initial state, we can obtain different results.)
The Composition is based on this principle: Starting from the same graph, we have run the algorithm three times, to obtain three different harmonic series, which have been used in the three parts of the composition.