This is the best book I have seen so far on theories for sound design. I used it for my graduate design research thesis 4 years ago at Carnegie Mellon, and I am about to require it for my students at UT next week, for a course in sound design. Chion opened up so many new ideas for me on different ways that sound affects us. Like the book itself, the forward by Walter Murch (considered by many to be a hero in film sound design) is inspiring.

After using Shigley & Mischkes' "Mechanical Engineering Design" for years, a colleague of mine loaned me this book. In comparison to "Mechanical Engineering Design," Deutschman, Michels, and Wilsons' "Machine Design, Theory and Practice" is a much easier (and quicker) read. I only wish my professors in school had chosen this textbook instead of the 'Shigley' title. Information in "Machine Design" goes into great detail for given, specific applications. An example of this would be shown in the text's explanation of Stub Acme & Buttress Threads (used extensively in my industry). Good luck finding finding information of this detail in Shigley's book.

This text is a collection of tests for multivariate time series models. While by no means comprehensive, this is a concise and well written text that extends to the multivariate case the necessary tests that are performed for the more commonly generated univariate time series models. I was pleasantly surprised at the job that the authors performed in developing some of the important ideas, as opposed to simply presenting equations.

In the text, methods covered include testing for joint stationarity, normality, independence, cointegration, and causality. There is also a section on the specification of multivariate linear models. Unfortunately, discussion of multivariate nonlinear models amounts primarily to literature references. Finally, there is a discussion of testing for model order and forecast accuracy. Overall, I found this to be a highly useful text.

If you are looking for a text to teach you all of the how's and why's of model building, this is probably not a text for you.

However, if you have a decent background in statistics and have had some model-building experience, this book provides a concise guide to the tests one typically uses when building time series models. The authors cover testing for stationarity, normality, independence, linear & nonlinear dependence, model order and testing the residual process, in addition to sections on the specification of linear & nonlinear models.

The authors do a very good job of including adequate text to explain the concepts that they are introducing, in the event something might be new to the reader. This is not a comprehensive text by any means, but it nicely covers the most common tests [certainly the ones I need for my work] and includes a list of references for those wishing to expand their area of knowledge.