Goldberg: Preliminary Survey

After coming up with the basic Goldberg idea, we decided to survey some of our potential users to get some general ideas about what people might want, how they work, and what they thought of the idea. We also drew some observations from discussions on music and composition-related mailing lists and newsgroups concerning computer-assisted composition and musical tools.

This message was posted to two mailing lists and a newsgroup, all frequented by composer-types of various kinds. The two mailing lists were Analogue Heaven and Electronica, both oriented toward electronic musical instruments and composition. The newsgroup was rec.music.compose, which is read by more traditional composers.

Some basic observations from the survey:

• People compose in many different styles. Even a particular individual may work in many different ways.
• Many of the nontraditional composers use a large variety of different tools. A new tool is viewed as an opportunity to experiment and try something out.
• Many of the traditional composers have a dislike for algorithmic composition methods and might tend to view with disdain any tool that hints of computer composition.
• A couple of people suggested that they might use Goldberg in real time for performance.

Some conclusions:

• We can't expect to write a single tool which will appeal to everyone (though it may be possible to write several related tools that will, using different interfaces, appeal to a wider range of people; more on that later).
• Goldberg should be able to integrate with other musical tools.
• Goldberg should support real time use for performance (perhaps by making it possible to audition variations before they play or at least have some idea of what a variation was going to do).
• As originally envisioned and mocked up, Goldberg would be unlikely to appeal to traditional composers.
• If Goldberg were presented as a non-intelligent transcription tool, it might appeal more to traditional composers.