Download of the Month: JSyn 13.4

Phil Burke, a developer of interactive music software, has filled an interesting niche in the field of software synthesizers. His newest creation, JSyn

Phil Burke, a developer of interactive music software, has filled an interesting niche in the field of software synthesizers. His newest creation, JSyn 13.4 (, is a cross-platform API (application program interface) for adding real-time, interactive audio-synthesis capabilities to a Web page that incorporates Java. The host-based shareware application gives users the opportunity to explore creative and interactive music-making over the Web in a bandwidth-efficient manner without having to stream or download audio content.

For example, you can use JSyn to enhance your Web site with simple sound effects or with complex sound environments that never repeat. The application offers a multitude of useful synthesis components-such as oscillators, filters, and envelope generators-as well as effects, 16-bit sample playback, and a programmable scheduler.

To hear what this powerful program can do, download the JSyn plug-ins and place them in your browser's plug-in folder. The JSyn Examples page includes applets for generating bird and rain sounds, an interactive synthesizer/sequencer called Performance Rack #1 (see Fig. 1), and compositions featuring JSyn by composers Barbara Benary, Nick Didkovsky (aka Dr. Nerve), and Matthew Yee-King.

On Mixman's Radio Mixman page (, users of Mixman Studio, Studio Mac, or Studio Pro can post their work and listen to and critique the music of others. Visitors view audio selections by artist, title, or upload date, and play them using onscreen tape player-style controls. Minimum system requirements are a RealNetworks G2 player and either Microsoft's Internet Explorer 4.0 or Netscape's Communicator 4.5. To upload your work, use Studio Mac or Studio Pro to create and export the necessary Real Audio file . . . Emagic ( and Rocket Network ( have teamed up with Harmony Central ( to promote Logic Rocket. This freeware version of Emagic's Logic Audio includes RocketPower, the proprietary code that enables a digital audio sequencer to use the Internet recording studios provided by Rocket Network. For this promotion, Rocket Network has set aside public studios that Logic Rocket users can access for free. Rocket Network and Digidesign ( have announced plans to release a RocketPower version of the Pro Tools digital audio workstation. (For more about using the Rocket Network system, see "Web Page: We Have Liftoff!" in the April 2000 issue of EM.) . . . At David Emery's Urban Legends and Folklore site ( library/blhoax.htm), learn the truth about the current rumors and hoaxes spreading like wildfire across the Internet and empty your inbox of all the e-petitions, bogus virus alerts, and other junk-mail clutter.

Indonesia's indigenous music and the instruments on which it is played are called gamelan. The traditional gamelan orchestra includes gongs, drums, plucked and bowed string instruments, flutes, and voices. There are many styles of gamelan music, classifiable by region, ethnicity, and social class. This rich musical tradition has spawned worldwide interest, and gamelan ensembles can be found around the globe. An important source of information about gamelan is the American Gamelan Institute (AGI).

Gongcast ( AGI/gongcast.html), hosted by the AGI and curated by composers Jody Diamond, Sapto Raharjo, and I Nyoman Wenten, is a Web site devoted to gamelan music of all styles. Its offerings include traditional and contemporary music composed and performed by both Indonesians and non-Indonesians, including American composer Lou Harrison.

When you click on the Gongcast button, five sound files (randomly chosen from the site's selections) play. A menu of Gongcast's compositions, titled "What you might hear on Gongcast," appears farther down the page. Gongcast can be heard using RealPlayer 3, 5, 7, or G2, the latter two yielding the best results.

What is MPEG-4? The Motion Picture Experts Group (MPEG) is part of an international standards organization that publishes specifications for compressing and encoding low bit-rate streams of audio and video information. Each phase of the continually evolving standard is numbered and contains different layers. Thus, MPEG-2 Layer III is phase 2 of the standard, and it contains three layers. It defines the popular music-file format colloquially known as MP3. Perhaps to avoid confusion, the next phase of the standard is called MPEG-4.

MPEG-4 Audio was designed with a broadband, multimedia future in mind. It describes audio "objects" such as speech or music and encodes them using a variety of techniques: Code Excited Linear Predictive (CELP), Harmonic Vector Excitation Coding (HVXC), and Transform Domain Weighted Interleave Vector Quantization, also known as TwinVQ.

The specification creates a standardized file format called Structured Audio that can encode and transmit a variety of synthesis algorithms. The idea here is that it's much more efficient to send instructions on how to reconstruct the audio on the user's computer than it is to send the audio file itself. A very simple example is, "Play a sine wave at 440 Hz for 1 second." In 30 characters, you have described 176 KB of Red Book audio.

However, producing music requires a more complicated algorithm. MPEG-4 includes a description of the Structured Audio Orchestra Language (SAOL, pronounced like sail) and the Structured Audio Sample Bank Format (SASBF), which together describe wavetable synthesis in a way similar to that of DownLoadable Samples (DLS) and Beatnik's Rich Music Format (RMF). Using SAOL and SASBF, a computer downloads and decodes Structured Audio files containing samples, instrument definitions, and performance information. The decoded data is then used to re-create the music on the user's system.

But MPEG-4 doesn't stop there: the standard also defines an interface for Text-to-Speech processes that can be used for lip synching and can modify synthetic vocal quality to take into account factors such as age and gender. Various effects-for example, reverb, echo, and 3-D spatialization-can also be applied to the bitstream, providing a mechanism for synching cinematic-style soundtracks to MPEG-4 Video objects (which are defined elsewhere in the spec).

Because it is designed to cover such a broad range of audio and video applications, the MPEG-4 standard is deep, comprehensive, and extremely complicated. It is also important to remember that it does not define how any of the synthesis methods will be implemented. It merely defines a format by which the audio data can be encoded for transmission. The required algorithms are so complex that, unlike MP3 (for which a plethora of free "rippers" is available), MPEG-4 will probably require dedicated hardware to generate the bitstream data (a fact that pleases a number of hardware manufacturers).

In any case, it will be a while before MPEG-4 products reach the general public, although research and development continues in facilities such as MIT's MediaLab. Version 2 of the standard, which adds error-checking and scalability functions, was only released in July 1999. But hold on to your hats, folks-the MPEG committee is already working on the Multimedia Content Description Interface called MPEG-7. Visit the MIT Media Laboratory web site at ( for more information on MPEG-4.-Peter Drescher