|
|
Two types of material revealed by frequency analysis
- Phonetic (determined) material based on vowels and consonants; was worked into two acts with 4 scenes each; operatic gestures; quotations, pastiche; double coding
- Statistic (derived from answering-machine noise; indetermined) material was worked into radio plays (Prolog, Intermezzo, Epilog)
 |
 |
| Spectrogram of the eight strongest components. Notice the nearly perfect alignment of the partials in segments with harmonic content (vowels). Click inside the graph to bring up a typical spectrum for the selected region. |
Summing up the ampltiudes for the partials shows the boundaries between the sonic events. These events were turned into parts, scenes and sections. The flat areas correspond to the answering machine noise. |
 |
- Waveform of the sentence (top) and the 90-minute structure derived from it (bottom).
- Two of the five parts are divided into scenes which are further subdivided into sections.
|
|
|
|