The value of the sonogram as an analytical tool has been successfully demonstrated for approaches to electroacoustic and acoustic music (Waters and Ungvary 1990). In particular, it can be used to show the overall form of a work at a single glance. Both Schreffler (1983) and Cogan (1985) have revealed valuable analytic information using the sonogram, which could not have been shown using conventional analytic tools: Schreffler, the connection between instrument and musical structure in baroque flute music; and Cogan, the unique spectral and formal characteristics of a range of musics, from plainchant to Stravinsky and Risset.
Despite its apparent objective scientific nature and ability to reveal previously unseen aspects of musical works, the sonogram is far from a complete notational or analytic solution (particularly with regard to mixed electroacoustic music). More recent approaches, such as those taken by Helmuth (1996) in analysing her own acousmatic music, have acknowledged these limitations and attempted to supplement purely sonographical information, for instance charting overall dynamic levels and phrasing. Even Helmuth fails to broaden this to include all the most pertinent aspects of electroacoustic works however, omitting for example spatial information. The approach outlined here can been seen as an extension of this process of supplementing sonographical information, and in particular the idea that a single analytic viewpoint, including the sonogram, is necessarily partial.
One of the basic strengths of the sonogram has been proven to be its use in giving an overview of a work, allowing its shape to be seen at a single glance, as will be shown. In the context of analysing mixed electroacoustic music it can also help us to reach an assessment of the roles of the instrumental/orchestral part and the tape. Mixed electroacoustic music often plays with blurring the boundaries between these two elements and so it may not be possible to separate them through hearing (a recording of) a performance. If we have access to the electroacoustic tape part as a separate recording (which should be possible), then we can produce two parallel sonograms, one of the 'whole' performance, both acoustic and electroacoustic elements combined, and one of the tape part only. This allows us to see the place the tape has in the final piece and coupled with our listening of both these recordings, we are able to begin to uncover the relationship between the two parts. This process is further supported by a close reading of the orchestral score, (which would be our main source of information were we to be without separate recordings).
Supplementing the sonographic image itself is the waveform display. This allows us to see the general dynamic levels as the piece unfolds in time. It also allows us to access spatial or panning information which the tape part may contain, through examining the separate volume traces for each channel of sound of the tape part and comparing the height of the trace in each channel. This can supplement our listening experience and help us reach a highly detailed picture of this (potentially important) aspect of the work being analysed.
Figure 3 shows a sonogram of the whole of Verblendungen. As something of an interesting aside, mention should be made of a connection which can be clearly made to another image concerned with this piece. Saariaho, in the paper previously mentioned (Saariaho 1987), gives one of the initial ideas for this piece as coming from the idea of a brush-stroke dragged across the page, and its resulting image (Saariaho 1987). This poetic image is surprisingly close to the visual representation of the sound of the start of the work created by computer sound analysis. Another striking aspect of the sonographic overview given in figure 3 is the clear stress placed upon the horizontal and vertical axes - in musical terms: harmony and rhythm.
At the start of the work it is possible to see a clear contrast between the strong vertical banding in the lower trace, showing the tape part, and the horizontal lines created by the orchestra's sound in the upper. By the very end of the piece, the tape's image has changed from a vertical emphasis to the horizontal, as if the orchestra and tape had swapped their initial roles. Leading on from this, we can begin to trace an overall tendency for the very tightly clustered horizontal lines at the opening of the piece (created by the orchestra's close harmony) to open up, moving further apart. This process reaches its conclusion in the widely spaced lines at the very end, the tape solo (c.7' 30") showing an intermediary step in this. We could further explore this by identifying the pitch areas within these bands, checking them against the orchestral pitch material occurring simultaneously, moving from large-scale observation to small detail to identify any relationships that might exist. An example of this comparison is given in the first sectional analysis in part two.
The overview sonogram also seems to support the large-scale segmentation begun, particularly identifying the large change in dynamic levels and graphic density near to the middle of the piece identified as belonging to the termination phase of the piece. It also gives us some sense of the visual weight of these two sections, showing how much darker (louder) the opening section really is in comparison. Compared to the overview graphic score which shows a number of different features of the work which we can further explore, the overview sonogram gives us no real details of the piece. It is necessary to work with more detailed sonograms which show us only a portion of the piece at a time but allow us to see the detail of the sound of the piece. Once again this supports an open, multiple view of the work.