LMJ13_02body_005-096 11/25/03 2:57 PM Page 53
Displaced Soundscapes: A Survey of Network Systems for Music and Sonic Art Creation
Álvaro Barbosa ABSTRACT
The introduction of various collaborative tools, made possible by the expansion of computer network systems and ollaboration has long been a key element in perimeter: 40,009 km). Even with communications technology, has C led to new methods of musical music; therefore, the use of collaborative systems based on data transfer at the speed of light composition and improvisation. computer networks to achieve musical results were a natural (approximately 300,000 km per The author describes a number development. The advent of computer network music dates sec) and unlimited bandwidth, of recent music and sound art back to the late 1970s with early experimental performances bidirectional latency would reach projects involving the use of network systems that enable in California by the League of Automatic Music Composers approximately 133.4 msec, which is geographically displaced [1]. Up until the early 1990s, systems for musical collabora- much higher than the tolerable creators to collaboratively tion using computers were based on local networks. In the last threshold. With such numbers in generate shared soundscapes. decade, the massive worldwide growth of the Internet has in- mind, it is not surprising that net- Various system designs, ideas creased the possibilities for composers, performers and audi- work delay is an assumed part of and concepts associated with this interaction paradigm are ences. However, it is well known that the phenomenon of creating music on-line. presented and classified by the network delay is a major drawback for real-time collaboration The thought that network delay author. over global networks. At the global level, physical limitations is particular to the medium and in current network technology introduce higher latency than that composers should embrace is acceptable for real-time acoustic collaboration. this fact when creating music is For the human ear to perceive two sounds as simultaneous, clearly expressed by experimental artist Atau Tanaka in an in- the sounds should not be displaced in time more than 20 msec terview with Golo Föllmer: [2], which means that for mutual awareness to be supported I find Internet time delay rather interesting and I think of it as in a bilateral performance, the maximum threshold would be a kind of unique acoustic of this media. . . . [R]ather than to around 40 msec (the time it would take a performer to per- play existing music on this new time basis, what is interesting to ceive a second performer’s reaction to his or her action). It me is trying to find a musical language that works on this time should be noted that the perception of two different sounds axis. ...[3] performed simultaneously is strongly dependent on sound This same idea motivated the SoundWIRE group at the Cen- characteristics (timbre, pitch and loudness), musical style and ter for Research in Music and Acoustics (CCRMA) at Stanford other types of feedback such as visual or physical stimuli. Nev- University to conduct significant research over the last few ertheless, a 20-msec threshold should be adequate. years to address the influence of network conditions in acoustic If we consider the smallest possible peer-to-peer connection communication [4,5]. between two opposite points on the planet, we have an ap- Latency has a highly variable and unpredictable nature, cre- proximate distance of 20,004.5 km (half the distance of earth’s ating time-based errors, de-sequencing and even partial loss of content, resulting in difficult conditions for performance Fig. 1. Screen shot of the FMOL software bamboo interface. With control. There are major efforts under way to improve the sit- this interface the user can manipulate and generate individual or uation by increasing bandwidth and data compression options. collaborative pieces. (© Sergi Jordà) Such improvements, along with restricting projects to Large Area Networks (LANs) or to Wide Area Networks (WANs) in geographically constrained territories (e.g. a country or even a continent), may reduce network latency to levels that will not impede real-time acoustic communication over the Inter- net. Network latency tends to be central to the discussion of network-based activities, yet there are other important issues, such as who will perform with these systems and whether time- based musical events/performances are always appropriate.
REDEFINING THE ACOUSTIC COMMUNITY Designing and implementing a network music system supposes that new, meaningful sonic results can be achieved by collab-
Álvaro Barbosa (researcher), Music Technology Group, Pompeu Fabra University, Passeig de Circumval•lació 8, 08003 Barcelona, Spain. E-mail: [email protected]