Analyzing Musical Sound

Musicologists have several starting points for their work, of which the two most prominent are text and sound documents (i.e., scores and recordings). One aim of this chapter is to show that there are important properties of sound that cannot be gleaned directly from the score but that may be inferred if the reader can bring to bear knowledge of the acoustic properties of sounds on the one hand, and of the processes by which they are perceptually organized on the other. Another aim is to provide the musicologist interested in the analysis of sound documents (music recorded from oral or improvising traditions, or electroacoustic works) with tools for the systematic analysis of unnotated—and in many cases, unnotatable—musics. In order to get a sense of what this approach can bring to the study of musical objects, let us consider a few examples. Imagine Ravel’s Boléro. This piece is structurally rather simple, alternating between two themes in a repetitive AABB form. However, the melodies are played successively by different instruments at the beginning, and by increasing numbers of instruments playing in parallel on different pitches as the piece progresses, finishing with a dramatic, full orchestral version. There is also a progressive crescendo from beginning to end, giving the piece a single, unified trajectory. It is not evident from the score that, if played in a particular way, the parallel instrumental melodies will fuse together into a single, new, composite timbre; and what might be called the “timbral trajectory” is also difficult to characterize from the score. What other representation might be useful in explaining, or simply describing, what happens perceptually? Figure 8.1 shows spectrographic representations (also called spectrograms) of the first 11 notes of the A melody from Boléro, in three orchestrations from different sections of the piece: (a) section 2, where it is played by one clarinet; (b) section 9, played in parallel intervals by a French horn, two piccolos, and celesta; and (c) section 14, played in parallel by most of the orchestra including the strings. We will come back to more detailed aspects of these representations later, but note that two kinds of structures are immediately visible: a series of horizontal lines that represent the frequencies of the instruments playing the melody, and a series of vertical bars that represent the rhythmic accompaniment. Note too that the density, intensity (represented by the blackness of the lines), and spectral extent (expansion toward the higher frequencies) can be seen to increase from section 2 through section 9 to section 14, reflecting the increasing number, dynamic level, and registral spread of instruments involved.