The documentation of the spectral envelope project can be roughly divided into four parts.
The first part, fundamentals, consists of chapters 1 and 2, and covers this introduction and the prerequisites of the project. In chapter 2, the basics of digital signal processing will be briefly visited, followed by an introduction to additive sound synthesis, the concept of spectral envelopes and the source-filter model of speech production. Finally, a quick tour of the other programs and systems of IRCAM related to spectral envelopes will set the framework for their application.
In the second part, the methods used and developed in this project will be described formally. After a description of the estimation of spectral envelopes in chapter 3, the internal representation is presented in chapter 4, manipulation of spectral envelopes is covered in chapter 5, and the application of spectral envelopes to sound synthesis can be found in chapter 6.
The third part, implementation, describes the translation of the results of the previous chapters into a working software system. Chapter 7 first presents some general considerations for the implementation, such as a brief introduction to software engineering, the architectural notation used. Then, the architecture of the spectral envelope library is presented, followed by a definition of the file format used. Appendix 1 describes the usage and the structure of the program VIEWENV to experiment with spectral envelope algorithms.
Finally, an overview of existing and possible applications of spectral envelopes is given in chapter 8, and a summary of the project and a conclusion from a scientific and an artistic viewpoint is given in chapter 9.