Newsgroups: comp.text.sgml Date: 23 Dec 1994 20:29:46 UT From: "Steven R. Newcomb" Message-ID: <199412232029.AA18633@bruno.techno.com> References: (7152) Subject: Re: Q: Info on SMDL or HyTime for Music? [Dave Williams] | I am told there is an SMDL, Standard Music Description Language, that | is a derivative of SGML and a more powerful markup language called | HyTime. | | Can someone point me to the right direction to get more information on | this? Since I served on that committee, I have a paper copy of the last iteration of the SMDL standardization effort. I do not have the electronic version. This work should be resumed; there is really not too much left to be done, and it would be a great pity to waste the effort that has already gone into it, not to mention the enormous potential benefit to the music industry of having a music representation that is based on the world's abstract timed-based hypermedia standard, HyTime (ISO 10744). As it happens, I recently wrote down some words about how I would sell the importance of ISO CD 10743 Standard Music Description Language to the entertainment industry (if I could ever get access to someone with any imagination, that is). I'm attaching them below. Please feel free to share the below material to anyone at all. It is going to require the cooperation, diligence, and selflessness of certain people to make it real. I know already who some of them are; I just don't know who the rest of them are yet. Are you one of the key people? Do you know where some money is available to lay the foundation of several new kinds of music industries? Best regards, Steven R. Newcomb | TechnoTeacher, Inc. direct +1 716 389 0964 | (courier: 3800 Monroe Avenue, main +1 716 389 0961 | Pittsford, NY 14534-1330 USA) fax +1 716 389 0960 | P.O. Box 23795 Internet: srn@techno.com | Rochester, New York 14692-3795 USA ******************************************************************************** Why Should There Be a Standard Music Description Language? ---------------------------------------------------------- Music is an integral part of almost all entertainment. It's not just for entertainment, though: music is found everywhere anyone wishes to make an appeal to others, or to influence their behavior, or to enhance any special occasion. Almost every situation is a situation in which music can play a role: brain surgeons use it in the operating room, supermarkets play it for customers, manufacturers play it for workers, and most people can sing, play, or whistle it themselves to at least some extent. I'm not saying these things because I think anyone needs to be sold on music, per se; most people already _are_ sold on music. I'm pointing it out because the music industry has yet to wake up to the fact that new information technologies pose opportunities and challenges that will fundamentally and permanently change the way all those who own music information assets will exploit them. What is a music information asset? It's the music itself, in the abstract. Regardless of who sings or plays the Lennon/McCartney tune, "Yesterday," and regardless of how it is sung or played, we recognize it as "Yesterday." What is it that makes it recognizable as "Yesterday"? It's not the words, because we can recognize it when we hear it in the supermarket as Muzak(TM), in which nobody ever sings anything. It's not the pitches, because it could be played in any key (i.e., started on any pitch) and still be recognizable. It's not the duration of the notes in real time, because it could be played fast or slow and it would still be recognizable as "Yesterday." Regardless of the tempo or the key, and regardless of whether it is sung or played, and regardless of the instrumentation, and regardless of the context in which we hear it, we can recognize "Yesterday." However, oddly enough, there is as yet no standard way to represent music which is as abstract as the way we hear it. You can write it down using ordinary music notation of the kind that Western civilization has been developing for about 1,000 years now, but the act of writing it down always instantiates it to some extent (i.e., it fixes the exact pitches that will be played, etc.). One of the side-effects of this situation is that when infringements of music copyrights are claimed to have occurred, the only way to settle the dispute is to ask a jury whether the two pieces of music sound enough alike that an infringement can be adjudged to have occurred. There is now an international standard at the Committee Draft stage that is waiting to be completed; it is known as ISO/IEC CD 10743, "Standard Music Description Language (SMDL)". SMDL is designed to represent music so abstractly that any printed score, and any performance, can be explicitly and comprehensively demonstrated to be an instance of it. The implications of the existence of such a standard, if it could be completed and adopted, are both profound and awesome. Music assets could be exploited very rapidly and efficiently for very diverse purposes. New technologies and markets based on allowing untrained people to have experiences of involvement with music assets become quite feasible. The franchise of musical training itself could be extended to children who cannot afford tutors, but who can afford to be taught by computer; such children could learn any music they wanted to learn. Every small town could support an amateur musical theater. The music publishing industry would be transformed: there would be no need for warehousing, and today's rampant use of photocopy machines by musicians would become far less worthwhile for them. Novel kinds of rights could be licensed. Music could be automatically rearranged for available performing resources, such as various proprietary music synthesizers, or actual performing groups. Do you need a convincing arrangement of "Yesterday" for glockenspiel and kazoo in five minutes? It can be done only if we have a computer-processable, totally abstract representation of "Yesterday." Do you want to experience Richard Wagner's opera, "Siegfried," in virtual reality, perhaps as Siegfried himself, with the Sword leitmotif worked into the music whenever you pick up your sword? Having Wagner's music in abstract form will be essential to making this possible. Have you always dreamed of conducting Mahler's Eighth Symphony, the "Symphony of a Thousand Musicians," because you feel you have a better idea about what the composer intended (or perhaps you are an egomaniac)? Would you like to experience what it's like to be Neil Diamond's bass player? Have you written some music that is absolutely great but no one will publish it because no market for it has been established, and you yourself are not yet marketably dead? Standard Music Description Language (SMDL) can open up all these possibilities, make all of the musical legacies of all cultures and ages available to everyone, create a marketplace of musical ideas, and put money into the hands of those who create musical assets. The SMDL standard was developed over a period of years under the auspices of the American National Standards Institute (ANSI) beginning in July, 1986, and under the Chairmanship of Charles F. Goldfarb and the Vice Chairmanship of Steven R. Newcomb. The project was later transferred to the International Organization for Standardization (ISO) and it became an international effort. However, the SMDL project was to be both energized and waylaid by the needs of the nascent hypermedia industry. Led by the U.S. military training and technical information industries, the hypermedia industry needed an abstract way to represent hypermedia documents, and SMDL had the solution to critical parts of the puzzle. As a result, SMDL was divided into two separate standards: one still called Standard Music Description Language (SMDL), ISO/IEC CD 10743, and the other called Hypermedia/Time-based Structuring Language (HyTime), now ISO/IEC IS 10744. HyTime, officially adopted in 1992, is a more generalized form of most of what used to be SMDL. SMDL is what was left after the HyTime constructs were excised from it. The plan was (and still is) to make SMDL an application of HyTime, thus ensuring the complete compatibility of music information with all other kinds of hypermedia information. However, because of the extreme demands of promulgating and implementing HyTime on the original developers of SMDL, and because of a continuing lack of understanding and interest on the part of the music and entertainment industries, SMDL still languishes unfinished, even while HyTime is being implemented and used by many information-intensive non-music industries worldwide. The expertise needed to complete the design and editorial work exists and the right people are willing to do the work. However, times have changed for all of them and they are no longer in a position to fund their own participation in the work. The meetings are expensive to attend, and the workload both during and between meetings is heavy. The job could be completed if a comparatively small grant could be made available, by relying primarily on the proven dedication and expertise of certain volunteers. What are you waiting for? Pick up the phone and inform Some Very Important Person (SVIP) about this situation, so we can start moving toward completion. Thanks! Steven R. Newcomb | TechnoTeacher, Inc. direct +1 716 389 0964 | (courier: 3800 Monroe Avenue, main +1 716 389 0961 | Pittsford, NY 14534-1330 USA) fax +1 716 389 0960 | P.O. Box 23795 Internet: srn@techno.com | Rochester, New York 14692-3795 USA