Problems of the analysis of performance. A trial of computerized comparison of recordings of the first Prelude of Bach’s Well tempered Clavier

Abstract

Tabulation, annotation, analytic rewriting of the score are essential graphic and
textual operations of any analysis of performance. They allow the analyst to visualize
some of the phenomena that form the object of the analysis and make them readable.
So doing, they pertain both to the product of a singular listening experience — that
of the author with his tools of analysis — and to the matrix of a reproduction of the
experience — the understanding of the analytical text by the reader.
These characteristics of the process of analysis are questioned here starting from
computerized comparisons of performance, realized at the Ircam in the frame of the
conception of a software allowing representation and manupulation of the
measurements of note durations in multiple performances of the first Prelude of
WTC I. Such a device allows varying the modalities of representation of a set of
information resulting from the measurements, directing the auditive attention, with
the help of these representations, towards phenomena that would otherwise remain
unheard, and keeping track of the path that led to the production of a fact of analysis.
If such properties are privileged characteristics of electronic data processing and of
computerized assistance to the analysis of performance, it seems possible and
desirable to put them in evidence in other analytical processes as well.