I was catching up with Computer Music Journal issues and noticed a review by Jeffrey Trevino and Drew Allen of The Audio Programming Book, in which I was a contributing author. The review has some valid points to make regarding the organization of the book and what it covers. However, I did find problematic one section of the review regarding my chapter on Modeling Orchestral Composition:
But why, you might ask, does an introduction to audio programming culminate in a computer model of the orchestra? The chapter should be reframed as an introduction to Csound, to avoid the current sense of aesthetic blindness. Rather than acknowledge the presence of the numerous higher-level electronic music programming systems that invite alternatives to the conservative score-orchestra model built into Csound, the chapter precludes the discussion of other approaches by failing to mention them. Instead of treating the score-orchestra metaphor flexibly, as is the case in most uses of Csound, the author cements its literal interpretation with a table entitled, “Comparing the Steps of Composing for Live Orchestra and Composing with [Csound’s] Orchestra Library.” (Emphasis Yi’s.) Lastly, he deals aesthetic alternatives their death blows in the form of an all-encompassing first-person plural: “By reusing an existing music software system, we can leverage the solutions already available and focus more closely on our compositional interests.”
I found this section of the review rather problematic and takes what I wrote out of context to serve the narrative of the reviewers that “the remainder of the printed text presents an arbitrary collection of special topic essays on possible relationships between Csound and programs written in C.”
There are a couple of points I wanted to discuss. First is that the text is specifically working on modeling orchestral composition, and proposes a library design to do so. The focus of the article was on the composition aspect, or, how the composer writes for an orchestra, and how to model that aspect of the relationship between composer and orchestra. It was not specifically about the orchestra itself, but requires a discussion of the orchestra to understand the composing process for it.
In that regard, I feel that the criticism that I did not discuss a “score-orchestra metaphor flexibly” is in a way, criticism that I did not write a different text altogether. It seems that the reviewers read the text as being about a Music-N style score/orchestra design, or specifically, about Csound’s score/orchestra design. The reviewer’s misunderstanding is notable in their additions of [Csound's] here:
“Comparing the Steps of Composing for Live Orchestra and Composing with [Csound’s] Orchestra Library.” (Emphasis Yi’s.)
Note, the original title of the table did not have “Csound’s” as the article was not about Csound’s score/orchestra at all. The library that is the focus of the chapter uses Csound as a backend, but it in itself is a generic design of orchestral composition that could be developed to target other backends as well (i.e. PD, Max, SuperCollider, Chuck, etc.).
Note: When I wrote the chapter, I did have some reservations that there may be confusion. Csound and Music-N languages do have a long-held design of a score and orchestra. I did worry that it might get confused with the concepts I was trying to model. For me, the Music-N score/orchestra are not the model of what I was doing, but just the tool I used to support the software model of the orchestral composition library. I tried to be clear of this distinction, but it seems to still have lead to some confusion.
This leads to the second issue I had with the reviewers’ comments regarding aesthetics. In the text, I wrote a brief section on wanting to reuse an existing synthesis system. This was so that the article could focus on orchestral composition and less on the synthesis details, which I believe are covered in other chapters of the book. I discussed why MIDI was not sufficient for the library’s design goals, and discussed why I chose Csound. However, I would have thought that readers who are familiar with SuperCollider, PD, Common Lisp Music, or other synthesis systems, would understand the chapter to be a generic design that could easily be implemented in other languages and systems. I did have to discuss Csound’s ORC/SCO model to explain how the results of the orchestral composition library would ultimately map its results to another systems software model to get audible results.
To me, this is not even an aesthetic choice, but an implementation detail. To me, the aesthetic choice here is that I wrote from a perspective of a composer looking at achieving compositional practices commonly found in orchestral composition and reproducing them in a computer music context. In that regard, that is the focus of the chapter, and is even the title of the text. I do not think then that not mentioning other aesthetic goals is a valid criticism. To me, It is as if I criticized an article about Impressionist technique as leaving out mention about Dada or Post-Modernism.
As for other tools and ways of working, I’d like to state that I have supported and continue to support other tools and software synthesis systems. I have often promoted that everyone should use the tools that best suits them, and ultimately the important thing to me is the musical result, and not the tools used. I think it is the reviewers’ reading of the article as being about Csound has lead to misplaced criticism.
To summarize, I appreciate the work of the reviewers to review The Audio Programming Book. However, in regards to the section on my own chapter, I feel they had misread the text. In the end, I still find the chapter and the design of the library interesting and hope it will be as useful to others–using whatever system they use–as it has been for my own compositional work. Hopefully I have explained my own perspective clearly, and would invite any further discussion.