I wanted to write the book because I thought it would be fun. It was. This may sound silly, but in the years I spent designing and running IPP, what I really wanted to do was use IPP.I think image processingapplications are much more interesting than image processing kernels.

In writing the book I was able to play with IPP functions with almost no restriction -except, perhaps, that Iplay very quickly. In order to make the book more interesting, I was actually required to choose and implement examples that were fun.I thinkthe audioprocessing chapter was a particularly good example of this.

But I think thequestion you really wantedme to answer was "why did I think the book was needed?" The answer is that I thought the bookfilled the biggest gap in IPP. IPP has had no "user's guide" to explain how the functions were meant to be applied. The manuals describe every function without much context, and they have some examples, but not for the majority of the functions. Andthe code samples give plenty of context but I think might be too long and/or intimidating as a starting point for a new IPP user. As for the tutorial, it just wasn't long or broad enough.

Do you think that is accurate? I wouldn't be surprised to hear a mix of responses, particularly since I think somechapters of the manualshave more detailed examples than others, and some functional domains have better sample code than others.

Stewart