source file: mills2.txt Date: Sun, 4 Feb 1996 10:47:46 -0800 From: "John H. Chalmers" From: mclaren Subject: miscellany --- Congratulations to Steven M. Miller, Harold Fortuin, Matthew Puzan, Enrique Moreno and Neil Haverstick. Even after a month, Neil's amazing MicroStock continues to reverberate in the form of the recordings of the concert. A first-rate event. Utterly superb! Way to go, Neil. Steven M. Miller deserves kudos for reaching outside of academia to solicit tapes for the U. of Santa Fe's upcoming electronic music concert series. In so doing, he discovered a remarkable fact: some of the best music out there is being done by ordinary schmucks with NO university affiliation, NO resources, and NO predelection for elaborate mathematics. Congrats also to Harold Fortuin both for building a xenharmonic generalized MIDI keyboard (see the Huyghens- Fokkers 1994 Yearbook for more info). Any chance of commercializing the widget? Good idea also to offer to exchange copies of your music. All too few forum subscribers seem to be interested in making and sharing MUSIC, as opposed to words about numbers about theories about words about numbers about... Especially praiseworthy: the scholarly labors of Mssrs. Moreno and Puzan, whose theses promise a comprehensive survey of xenharmonic incunabula. Speaking of which-- Several articles touching on xenharmonics have recently been published. Of particular interest is Contemporary Music Review, Vol. 10, 1994. The issue is *entirely* devoted to "composition with timbre." As we all know, composing with timbre is the gateway to non-12.. Thus this issue is of inherent promise to xenharmonists. Also of interest: "Musical Scales In Central Africa and Java: Modelling by Synthesis," by Frederic Voisin, Leonardo Music Journal, Vol. 4, 1994, pp. 85-90. Voisin describes the results of a series of experiments in which indigenous tuning experts were allowed to tune up their own scales on a DX7II: the process was recorded with a MIDI sequencer, and the researchers returned after several days in each case and asked the same tuning expert to evaluate the tuning again (so as to confirm the reliability of the results). When the DX7II was tuned correctly, the researchers report that the tuning expert and a group of bystanders would spontaneously perform a piece from their repertory. This methodology sounds like a significant advance in ethnomusicology. It's orders of magnitude beyond the relatively useless previous practice of asking the indigenous tuning expert a series of questions--if the questions are phrased in terms of Western European musical assumptions, what possible use can the answers be? It's as though a bunch of Javanese music theorists came to the U.S. and asked Western symphony orchestra musicians: "What bem do you use? And how do you arrive at your rasa?" The answers could not possibly make sense. And the previous practice of simply measuring tunings in terms of cents and trying to fit them into Western equal-tempered or ji tuning models was also less than satisfactory. After all, a set of measurements of native instruments does not *by itself* tell us what the local tuning experts were trying to achieve...while a digital record of the tuning process *might just do that* "Applying Psychoacoustics in Composition: 'Harmonic' Progressions of 'Nonharmonic' Sonorities," by Richard Parncutt and Hans Strasburger, in Perspectives of New Music, Vol. 32, No. 1, 1995, is also likely to prove interesting to the prospective xenharmonic composer. Finally--finally!--music theorists are beginning to wake up to the reality Plomp and Levelt revealed in 1965: --An INharmonic series of vertical partials will sound entirely transparent and consonant as long as the distance between each successive partial is greater than the critical bandwidth at that frequency. While visionary composers like Jean-Claude Risset, John Chowning, James Dashow and Jonathan Harvey have long taken advantage of this psychoacoustic fact, for just as many decades the music theorists have ignored this reality. Parncutt & Strasberg's paper is thus a welcome addition to the literature. By basing their compositional theories on Ernst Terhardt's model of hearing, they've gone back to basics--to the way the human ear hears, rather than abstract set theory, or partition functions, or abstract algebra, or combinatorics, or matrix operations, none of which has any necessary connection to what the ear actually HEARS. Among other praiseworthy ideas, P & S suggest: "(i) the model takes as its starting point the spectrum of a sonority, rather than its musical notation." This is a VITAL advance in music theory. Now that arbitrarily complex and inharmonic timbres can be generated by computer, the composer needs to consider the implications of timbre as well as such familiar concepts as root note, voice-leading, the relationship twixt harmony and melody, etc. Moreover, P & S insure that "The model is not octave generalized; it is based on pitch height...rather than pitch class." Another extremely important point. Octaves certainly have their uses, but many of us compose in tunings without octaves. Why ought we to be hamstrung by 600-year-old dogmas obsessed with subdividing a 2:1 interval...which interval does not exist in many of the tunings we happen to use? P & S also point out many subtleties often unrecognized by music theorists: for instance, the psychoacoustic effects produced by mistuned harmonics, etc. Listeners still perceive substantially mistuned harmonics as making a contribution to the formant--and so on. P & S offer C code for an algorithm to evaluate vertical sonorities according to their theory--something sure to be useful to many xenharmonists who compose with timbre. Alas, the article also suffers from a number of omissions and oversights. First and most important, P & S base their compositional theory solely on Ernst Terhardt's theory of hearing. Terhardt is a well-known psychoacoustician who's done excellent work in the field. However, Terhardt's theory of hearing is merely one among many. More: Tehardt is a dyed-in-the-wool place theorist, and this tends to bias some of his conclusions. There is, for example, *no* discussion in any of Terhardt's papers of the contradictions and paradoxes bedeviling place theory--for example, that measured jnd's substantially exceed those predicted by the physics of the place theory of hearing. Thus there is good reason to believe that the model of hearing on which Parncutt & Strasberg have based their compositional theory is incomplete, and does not explain many important aspects of the human auditory system. P & S also make statements which tend to mislead the unwary reader: "The perception of the pitch of a complex tone such as a musical tone (piano, violin, voice, and so on) involves pattern recognition (Goldstein 1973, Terhardt 1972 and 1974)." In fact there is no consensus, nor yet any convincing body of evidence, as to the exact processes involved in the perception of the pitch of a complex tone. While Parncutt & Strasberg claim that musical perception is a matter of "pattern recognition," in actual fact this is merely one of three competing theories of the way the ear/brains system operates. The other theories are that the inner ear performs a mechanical Fourier transform, and that the auditory nerve running from the inner ear to the brain extracts & encodes the underlying periodicity of sound waves detected by the inner ear. While there is some evidence in support of the pattern recognition of hearing (first advanced by Wightman, by the way--not Goldstein!), there is also some evidence AGAINST the pattern recognition theory of hearing. There is also a great deal of evidence FOR the other two theories of hearing, none of which Parncutt & Strsberg seem to be aware of. On top of these lacunae, P & S proceed to collapse their model down to 12 pitch-classes. This is a pretty low blow. After much fine talk about freeing the composer from octave equivalence, etc., they wind up giving another recipe for producing pretty sounds in 12-equal. File that one under the category "The beatings were scientifically designed to enhance creativity." Lastly, P & S betray a profound lack of familiarity with the microtonal and psychoacoustics literature. Their bibliography is full of extremely glaring gaps and omissions: for instance, they omit completely most of the CLASSIC articles on composing with inharmonic sonorities: John R. Pierce's letter "Attaining Consonance in Arbitrary Scales," In JASA, 1966; Mathews' and Pierce's "Control of Consonance and Dissonance With Nonharmonic Overtones" in "Music by Computers," ed. Beauchamp & Von Foerster, 1969; Jean-Claude Risset's "Digital Experiments: 1964...." in Computer Music Journal, 1984; James Dashow's "Spectra As Chords" in Comptuer Music Journal, 1980; William Sethares' "Local Consonance and..." in JASA, September 1992, my own "The Uses and Characterisics of Non-Just Non-Equal-Tempered Scales" in Xenharmonikon 15, 1993, and last (but not least!) Slaymaker, "Chords From Tones Having Stretched Partials," JASA, 1970 and Geary, J. R., "Consonance of Pairs of Inharmonic Tones," JASA, 1980. If a nudnik like myself can rattle off this many obvious references from memory, shouldn't Parncutt & Strasberg have been able to do the half an hour or so of research required by minimal standards of scholarships? Ought not these two distinguished researchers to have been able to dredge up adequate citations for their article? Such negligence bespeaks more than mere laziness; it augurs a total lack of interest in looking outside their own narrow little sphere of interest--namely, the theories developed "by Ernst Terhardt and his colleauges at the Institute of Electroacoustics, Technical University of Munich." Thus, while Parncutt & Strasberg's article is a promising start, it falls short on many counts. Withal, still a worthwhile and useful article. --mclaren Received: from eartha.mills.edu [144.91.3.20] by vbv40.ezh.nl with SMTP-OpenVMS via TCP/IP; Mon, 5 Feb 1996 00:17 +0100 Received: from by eartha.mills.edu via SMTP (940816.SGI.8.6.9/930416.SGI) for id PAA29226; Sun, 4 Feb 1996 15:16:58 -0800 Date: Sun, 4 Feb 1996 15:16:58 -0800 Message-Id: <960204231257_71670.2576_HHB69-6@CompuServe.COM> Errors-To: madole@ella.mills.edu Reply-To: tuning@eartha.mills.edu Originator: tuning@eartha.mills.edu Sender: tuning@eartha.mills.edu