Alessandro Ludovico


* * *


The project Faust Music On line was an open invitation to participate through the Internet in the musical composition of a part of the sound track of the forthcoming show of the Spaniard drama group, La Fura dels Baus, entitled: Faust v3.0, based on Johann Wolfgang von Goethe's work.
This show by the Fura dels Baus had its first performance in Barcelona on 28 April 1998 and it was planned to do a tour through Spain, other European countries and NYC, giving some 200 performances, during 1998 and 1999.
All the music was composed by means of a W95 software you can freely download (2 Mb) at During the months prior to the first performance a web server stored and updated the database with the compositions received and permitted the audition and ongoing development of these compositions, as described below.
All communication with the server took place through this software. At the beginning of April, 1998, La Fura selected some (around 60) of the compositions for the play.

The musical composition of all the pieces for this project has been done by means of this program, responsible for audio synthesis and control of the graphic interface and all the communications with the server. This program is an independent application, which, when executed, is automatically linked to the page set aside for this purpose in the S.G.A.E. server and interacts with it to exchange information.

Composition is done in real time by way of attractive graphic interfaces. The program does not handle traditional musical concepts and it is, therefore, open to people who do not possess this knowledge. The whole control is executed in real time with the aid of the mouse on animated graphic objects that evolve on the computer screen.
However, we have also thought of more advanced and professional musicians and we have tried to ensure that these interfaces should not just be a simple banal interactive game, but potent sound experimentation tools.

Each of the compositions done with the program and stored in the server is the product of co-operation between various authors (normally 4) and not the work of a single composer. We consider that group creation is one of the most important aspects of this project and one of the most interesting capabilities - still unexploited - offered by the net.

When an author retakes the piece produced by another to enrich/modify/distort/deconstruct it, he will not be modifying the original but working on a copy. Thanks to this mechanism, an idea or musical germ generated by one author may evolve in countless directions at the same time, although all of them (including the initial one) are equally accessible in the form of a tree (like the Windows Explorer) in the database.

By means of the Direct Sound and Direct Draw technologies, the program is capable of synthesising in real time eight 16-bit, 22-Khz stereo voices with different fully parametrizable effects (reverb, panning, delay, pitch shift, resonant filters, ring modulation, etc.) for each of these voices, as well as to control and generate all the graphic interface animations at the same time.

Faust Music on Line has been conceived by Sergi Jorda and la Fura dels Baus, and developed by Sergi Jorda (, with the sponsorship of FUNDACION AUTOR.

More info:

* * *


1. How many people concurred to F@ust Music Online?

From 98-01-18 to 98-04-16 (the date on which the reception of pieces for the show concluded), more than 1,100 brief pieces by around 100 composers were submitted, but the core of assiduous participants was no more than 20. These constituted a virtual family, which communicated every night, exchanging sounds, creating music collectively, without knowing each other. They were the FMOL junkies; creators who spent several hours a week during a three months period, and with whom I exchanged numerous technical and esthetical e-mails.
Given the SGAE promotion, this influx may seem somewhat poor, but the existence of this selected and faithful core group was enough to justify all effort involved.

When conceiving FMOL, one of my personal aims was to conceive an engine and a graphical interface that could be attractive to both trained and untrained electronic musicians. And I think that this goal was fully attained. We now know that several of these "faithful" participants had not a prior contact with experimental electronic music; a few were even composing for the first time, but all of them, took it, however, as a really serious game, and the final quality level of the contributions was really amazing. On the other side, I would like to say that during this period, I did also receive many e-mails from more orthodox composers, puzzled and confused by that crazy piece of software, but who still wanted to collaborate with La Fura in a more traditional manner.

2. 2. Anyone try to hack it, or use it in a non-standard way?

I have not heard anything about it. It has to be considered that the software was completely free, so usually people do not care very much about manipulating something that, to their eyes, doesn't cost anything. Some care, it is true, was taken (following the SGAE 's suggestions) to avoid unallowed or anonymous manipulation of the submitted and stored pieces. What I mean is that I don't really care about anyone manipulating the software, but the SGAE cared a lot, about anyone manipulating the music. For that reason, the program itself had many protections against accessing the database in an unpredicted manner. It has to be said that FMOL is a stand-alone software. All TCPIP communications between the synthesizer and the server was done through the program, and not through a normal browser. Nevertheless, of course, any advanced hacker could have easily accessed and manipulated the database. Considering that no one complained about his/her music being manipulated I suppose no one tried to; which is quite logical, considering how little the "prize" was (being selected for La Fura's soundtrack).
While trying not to be very technical, I think some words about the FMOL submission process can help understand all this mechanism. While any internaut can download the application, listen to stored pieces or rehearse with the instruments, users have to become registered before sending a contribution. This simple process is done from the
application by filling a form with the name, an alias, and an e-mail address (optional check-boxes allow everything except the alias, to be hidden to other users); a few seconds after this information has been posted, the host sends back a password which the registered composer will have to use at any time he submits a new piece.
But again, your question can have many different lectures or answers. For me, one of the most important points of the project was to encourage collective creation; to allow germinal compositions, musical ideas brought by one composer, to grow and evolve, through
the participation of new authors, in many different ways, unpredicted by the first author. So the program permits (and promotes) the manipulation/distortion/deconstruction of previous pieces, by new authors. That may seem a non-standard approach, but it is FMOL's
approach. The only point is that all these manipulations were allowed and registered in the database!
And last, but not least, there's another point related to your question. As a personal musical instrument, FMOL had some cut-offs that were implemented because of the submission-protection I've mentioned before. For instance, multitrack recording is not possible without going through the server (after recording a track, this one has to be uploaded and downloaded back, before a new track can be overdubbed). In addition, La Fura wanted very short pieces, so recording is only allowed for 20 seconds! Obviously, once La Fura's project was done, many users asked me for a new "unprotected" version, more oriented to personal (vs. collective) composition. I haven't had the time to do it yet, so I'll be glad if someone hacked the code and did it for me!

3. Did you apply any selection to the work received? If yes, what's the criteria on which you based it?

Of course we did! La Fura wanted 50 pieces, and we had 1,100! Therefore, a complicated election process took place. It's always subjective. We discarded pieces for not being too interesting. We discarded many pieces for being quite similar to others. Some were good but didn't bring anything new to existing ones. When we had around 250 pieces, we classified them in groups based on their sonorities, atmospheres, etc. La Fura had 20 or 25 parts of the play were they wanted to include FMOL pieces, and each part had some adjectives describing the atmosphere required. So we started distributing the preselected pieces into these "boxes" and then we choose the best (2 or 3) from each box. All this process was done by Carlos Padrissa, the show's director and promoter of the project, and me.

4. Did any later collaboration start after getting in touch for FMOL?

During this selection process, it became clear that a great number of interesting microcompositions -and the good job done by many unknown authors- had to be left aside. Moreover, many pieces were begging for expanded development, beyond the 20" limitation. What were we going to do with the thousands of hours composers had spent playing with our toy? What would happen to the 7 hours of music stored in the SGAE's host? For all these reasons, I made a new ecological proposal both to La Fura and the SGAE: why not produce a CD with new compositions, but all of them being based on the audio material stocked at the database?

In September 98, I asked some of the top (my favorite) FMOL composers (together with new guests), to contribute, and each one received a CD-ROM with more than 200 pieces of 20" selected from the database (already rendered to audio to simplify the job). The only rule: "everything is allowed as long as the raw material is FMOL". So the composers started to cut, paste, overdub and process their favorite bits, composing totally new and longer pieces (but keeping track of the fragments used!). Two months later, in November 98, the CD, with
17 new compositions, was released (La Fura dels Baus - Sergi Jord?: F@ust 3.0 - FMOL, Fundación Autor 02801).

I'm really convinced that I have "discovered" several great composers who had never had the chance of publishing before, and I keep E-mailing a lot with some of them. I also know that several collaborations between some authors have started outside from FMOL.

5. If you could create a hardware device to improve the composing on
the Internet, which could be the features that you'll want to

That is a complicated question. The problems with the Internet are the problem with communication: (a) transmitter and receiver have to understand the same codes, and (b) bandwidth. For point (a), on a first, simple, non sci-fi level I would say that a very basic problem is the need for more standards. Standard MIDI files proved to be very successful, a decade ago, for exchanging files between different sequencers and therefore different users. Later, General MIDI appeared, allowing a common sonic palette between users. But obviously, General MIDI cannot make any good to creative music; it is a very impoverishing standard. A very simple next step should be a sample format standard. The Akai is a standard de facto between professional musicians, but not many hobbyist have an Akai sampler!
This problem -the lack of a sampler standard format among all platforms- was, in fact, one of the reasons I decided to implement real-time synthesis, instead of MIDI, in FMOL. Creative Labs owns a SoundFont ® format which is very widespread, considering the large number of existing cards (SB32, AWE32, AWE64, Live, etc.) that use it. It is clear that this problem is crucial not only for musicians, but specially for all the computer games industry (which, as we know, moves much more money). That's why last year, Microsoft did establish a new format file, DLS (downloadable sound) which will probably make it. It is a shame that this format is so much more restrictive than Creative's one.

This format question is technologically very simple. The problems related with (b) bandwidth are a bit more complicated. I'm talking about top quality streaming MIDI and streaming audio (no perceptible loss, no perceptible -or not too much- delay). That can be done right now with RDSI. In fact, many important audio recording studios around the world are doing recordings with musicians on the other side of the ocean. However, even with RDSI the delay is too important for a real jam session. The idea of musical computer networks is by no means original; earlier implementations (on a local area scale) date back to the late 1970s with performances by the League of Automatic Music Composers. But twenty years later, real-time collective improvisation keeps posing serious timing difficulties at a global scale.

I think, in fact, that maybe these points don't really respond to your question. What kind of hardware device would improve the composing on the Internet? If you are talking from an artistic or aesthetic (more than technical) point of view, then I have no answer.
Any device can be useful to someone, but I don't see now The composing device, and least the Internet-specific one.

The big aesthetic Internet question, is, in my opinion, a mentality question. The Internet not only favors the omnidirectional distribution of information; it also promotes the dialog among its users with services like E-mail and chat. If we imagine some kind of chats or E-mails that would use music instead of written words, then both collective creation and the production of open and continuously evolving works, would doubtlessly be two of their major and more appealing artistic consequences. In this context, concepts like authorship and copyright will necessarily have to evolve and adapt to a new reality.

The prevalent use of digital tools and the Internet's communication facilities may be changing the rigid authorship concept that still prevails in music and other arts. The use of sampling technology has boosted an appropriationist approach in music (although related possibilities have always been available) that carries both legal and aesthetic ramifications. Digital data is clonable, which leads to piracy, but it is also manipulable to any degree; and if this data is
thrown into a public and easily accessible medium such as the Internet, its chances of development expand, allowing any kind of structural, formal or content redefinitions. When this inevitable process is unwanted by its original creator, the secret evolution of any piece may be hard to follow; but on the occasions when it is desired or encouraged, the resultant collective creations exhibit an artistic expressivity that comprises both the sum of all the participants' work and an expression of all the social interchanges that took place during this interactive process. In this context, copyrights become difficult to define and control, if not questionable.

You see, that is my point of view, but then things took a perverse gloss when La Fura, who was looking for FMOL sponsors, convinced the Sociedad General de Autores y Editores (SGAE-the Spanish authors' association) to produce it. Not only did the SGAE finance and promote
it, but they also worked on the database server programming, whom they hosted (together with the downloadable program) on their website for more than one year. Changes were made in the program's conception, concerning authors' registration. While any "internaut" can download the application, listen to stored pieces or rehearse with the instruments, users have to register with SGAE before sending a contribution. This simple process is done within the application by
filling in a form with fields for name, alias, and E-mail address (optional check-boxes allow everything except the alias to be hidden to other users); a few seconds after this information has been posted, the host sends back a password that the registered composer must use whenever submitting a new piece.

As a collective composition simulation game, FMOL is a closed ecosystem in which all of its inhabitants (few and identified) agree to play by the rules. The full control exertable over the FMOL world obviously has nothing to do with any real and open Internet situation. However, the SGAE (and some authors) may be satisfied: all the composers of the pieces that were selected and are now part of F@ust's score are receiving their lawful rights.

6. If you could project a theatre environment with the sound shaped on the audience's humour, what'd be its main characteristics?

I have conceived several virtual instruments and interactive music systems since 1989. Some of them were thought for myself or for trained musicians, while others like in EPIZOO, had to be controlled by members of an audience in public performances. The demands for both genres are obviously different. For the first group, complicated tools, which offer great freedom, can be built. The second group demands simple but appealing tools which, although giving its users the feeling of control and interaction, should not produce "too bad" outputs. Both classes are often mutually exclusive; the musician gets easily bored with the "popular" tool, while the casual user gets lost with the sophisticated one.
Despite this conviction, in FMOL I tried to conceive a graphical interface which could appeal to both sectors; a tool that would not dishearten hobbyist musicians, but that would still be able to produce completely different musics, allowing a rich and intricate control, and offering various stages of training and different learning curves. According to the users' feedback I think we did it pretty well, but I still believe that it's a rather high goal trying to please everyone -- not speaking of sound installation-like interfaces, which are to be used only for a small amount of time.

In EPIZOO, one member of the audience at a time conducts the all show, "torturing" the performer, controlling the animations projected on a big screen, the lightshow and the interactive music. Epizoo is an interactive performance that requires a rock-concert-kind-of trance atmosphere; the show must not decline, which means necessarily that the interactive freedom has to be restricted. The user must feel the power of the conductor, but at the same time, the orchestra must play with conviction whoever the conductor is, whatever he or she does. And we are talking about persons that pick up the mouse and have to face a given program for the first time.

Therefore, in this case (and in many others) the illusion of interactivity is therefore more important than interactivity itself. And you are asking me about a whole audience, not just one member.
Then, I believe, we want it or not, only the illusion stands. How can one individual perceive its grain of sand contribution? I don't want to be misunderstood: I believe that interactivity may be the most important point in computer art, but there are limits to it, and music is a very complicated thing. In Siggraph, for some years, the theatre audience was divided into two teams, each having a different color paddle, and then 1,000 people are asked to play a pong game. And after some seconds needed for the audience to get tuned, to behave as one entity, it works; but music is more complicated than a rising and falling paddle. It could surely be always fun to play, but not necessarily to listen to.

Before we manage to shape the sound according to the audience's humour, we should be able to shape it according to one individual humor, and that's not an easy task either. In my opinion, the Brain Opera, will still have to wait.

7. The ongoing development of sounds added to the FMOL database isn't it a bit like a sort of a casual human-driven sonic garden that realises a restless evolution?

I had never thought in gardening terms, but it is in fact absolutely like that! As I already said before, I believe that collective creation and the production of open and continuously evolving works are two of the major and more appealing artistic breakthroughs the Internet has to offer to creators, and both concepts were really the starting points for that project. In that sense, the most important decision we took, my collaborator Toni Aguilar and I, was to organise the server's scorefiles database not as a simple list, but rather as a tree. This would allow germinal compositions -musical ideas brought by one composer- to grow and evolve, through the participation of new authors, in many different ways, while still permitting, at the same time, access to all the existing pieces/nodes. In gardening terms, it means that new plants can come up at any time, but also, that the existing ones can keep growing and branching.

In FMOL, pieces can have up to four tracks, each one possibly by a different composer. Each time a user accesses the database, the program receives and updates the compositions' tree, allowing the user to see all the compositions' genealogies. This user is then able to download and listen to any of these composition and decide to enrich/modify/distort/deconstruct any of them (as long as its four layers are not filled), sending it back to the server. This new node will then take its place in the tree as an offspring of the downloaded piece. Although the maximum number of users (or layers) per piece is reduced to four, there is no limit to the number of variations that each layer may generate. This means that the tree has a maximum depth of four generations but unlimited offspring at each one of them (four-layer pieces can be listened to but not "augmented", though its parents can keep engendering). That way, a musical idea brought by one composer can develop in multiple and unexpected directions; pieces become entities with lives of their own, capable of evolving out of the control of their original creators.
I have found this whole process a great inspiration booster: you throw one idea to the server, and come back some hours later to see what different evolutions it has undergone. Some may be completely unexpected and also maybe quite challenging.

* * *

Neural Online -
Suoni Futuri Digitali -