|Abstract|| || |
Open Sound World, or OSW, is a scalable, extensible programming environment that allows musicians, sound designers and researchers to process sound in response to expressive real-time control . In OSW, components called transforms are combined to form programs called patches. Patches are themselves transforms that can be included in other patches. This gives rise to a hierarchical name space in which every transform has a unique address. This address can be used to send and receive OpenSound Control (OSC) messages to and from a transform. Additionally, the inlets, outlets and state variables (i.e., internal variables and parameters of a transform) are themselves named components with unique addresses. Thus an OSW program forms a complete OSC address space where any object (patch, transform, variable) can be queried inspected or modified using OSC messages. This presentation explores several applications that utilize OpenSound Control in OSW
Discovery of OSW programs by generic clients:
OSW includes a full implementation of OSC queries, which was largely the work of Andrew Schmeder at CNMAT . Client programs can discover a running instance of an OSW by detecting a special “heartbeat” message (i.e., similar to active sensing in MIDI), and then discover the dynamic address space of the OSW programs. The client can then use OSC to send any message understood by any of the objects in the OSW programs.
Custom User Interfaces:
The discovery/query system, as well as standard OSC messages can be used as a mechanism for inter-process communication between OSW and custom user interfaces. Such interfaces can be special front-ends for individual patches, or alternative user interfaces for building patches. We present specialized user interfaces as front-ends for patches that use the advanced synthesis capabilities of OSW.
In addition to communication between clients and servers, OSC can be used for peer communication among multiple instances of OSW. Such communication includes sending OSW time and signal data via OSC, allowing distributed synthesis and signal-processing on networked systems.
Building robust and reliable software is a challenging task, more so when the software must run on multiple platforms. For OSW, we have introduced a system of automated testing in which test cases are Python scripts that send and receive OSC messages that build, run, and inspect dynamically constructed patches. Using OpenSound Control as the “command language” for testing has allowed the rapid creation of test cases and uncovering of bugs that would otherwise be undetected until deployment. The test system as well as the fixes uncovered in the process will be part of the upcoming Open Sound World version 1.2 .
 Chaudhary, A., Freed, A. and Wright, M. An Open Architecture for Real-Time Audio Processing Software, Audio Engineering Society 107th Convention, prepr int #5031, 1999.
 Open Sound World, software by Amar Chaudhary et al. http://osw.sourceforge.net
 Schmeder, Andrew W. Implementation of a Scalable and Dynamic Interface for Open Sound World using Open Sound Control.