Composition by Conversation

Composition by conversation (CbC) is a scenario in the MUSICA project where a human and machine collaboratively compose music primarily through the use of natural language commands. This task involves modeling natural language for musical concepts, developing computational models of music that facilitate integration with natural language processing algorithms, querying a score for musical features based on partial specifications of concepts, and generating new music based on user descriptions. CbC’s implementations use MusECI as part of the language-handling framework, and we are conducting user studies to inform the development of parsers for musical language.

Online Versions

New version: (browser interface developed by Chris Kim)
The most recent interface, which supports both chat and point-and-click operations. This version will have periodic feature updates over time. We recommend Chrome on OSX or Windows. Android may work on some devices. The Edge browser and iOS devices have known compatibility problems.

Old version: (browser interface developed by Chris Burrows)
This chat-only version is very close to the original offline version. Start your session by giving a name, then try the command “generate 4” (without the quotes) to get started. You can use “help” to get information on what kinds of commands work. We recommend using Chrome to view the web interface. Firefox and Safari should also work, but there are known compatibility issues with Edge.

Offline Version

From late 2017 through early 2019, the MUSICA group used a Unity and Python-based version of CbC. This offline version gave a chat-style interface and used Lilypond for typesetting the score, and it was used for periodic user testing. The offline version contains certain features that are not yet present in the current web versions (such as multiple staves). There are currently no plans to make this version of the software publicly available, since the web version currently under development will eventually contain all of the same features and more.