28 December 2007

Humidity and Performance Venue Timbre

Bohn, Fig. 6
How might chamber music presenters adjust their HVAC systems to optimize the acoustics of the performance hall?

This question occurred to me when I attended an ‘early music’ concert in Minneapolis. The humidity in the church was very low, probably because of the wintry air temperature outside—and the fact that the church heating plant and thermostats were set so as to conserve energy. Too, there probably was no humidifier in the airflow ducting of the furnace, although I don’t know for certain whether this was the case or not. Within 10 minutes of taking my seat, my eyes felt dry. The performance was delayed because of snowy road conditions (much like the situation on 13-DEC-2007 in Boston). Within 40 minutes, still waiting for the performance to begin, the inside of my nose was crispy and sore. The room was like the Sahara, maybe 5% to 10% relative humidity!

But eventually the concert started. And, oh, the acoustics! The acoustics were unusually crisp. Scintillating! The extraordinary timbre of the room was all the more noticeable because I was seated about 30 meters from the performers—far enough away for the reverberation and acoustic absorption properties of the arid air between the performers and my ears to have a big effect on the sound waves.

By contrast, Boston and Kansas City and New York are temperate cities and always fairly humid (40% to 100% relative humidity outdoors, and 30% to 70% humidity indoors). The altitudes of all of the cities where I’ve ever attended performances or played are low, so I think we can safely exclude barometric pressure differences as the source of the differences I’m referring to.

Chamber music performance venues in these humid cities all sound warm and ‘liquid-ey’. There’s a high-frequency ‘roll-off’ on everything—starting, I think, at about 1 KHz (C6) and dropping maybe 3 dB per octave above that.

The warming, fluid effect at increased humidities that I’m referring to is not a pan-spectrum ‘deadening’ of the sound. (The effect of new-fallen snow, like we have in K.C. today, is deadening—a severe attenuation across all frequencies; no distinct low-pass cut-off frequency or ‘shoulder’—and that’s not what I’m talking about at all. ) The high-humidity timbre is just sweet and warm.

Reverberation is actually enhanced in humid environments. The reflected sounds are louder. And this enhanced sound pressure level of the reflected sounds seems to be what is mostly responsible for the warmer sound in a humid environment.

I  retrieved the papers by Harris and Bohn, extracted the data contained in them, and performed log-normal statistical regressions on the data. (The data are for ‘direct’ reception of ‘flat’ wavefront sound propagating horizontally from a single localized sound source to a receiver at the same elevation above the floor as the source. In other words, the data—and my regressions—do not consider any reflections/reverberation, do not take into account the height difference between the performers and the listener, and do not make any assumptions about hall dimensions, geometry, wall-coverings, etc.) For your interest, I’m making the results of the regressions available in the following spreadsheet.

Spreadsheet to Calculate Sound Pressure Loss Spectrum as a Function of Relative Humidity
If you download the spreadsheet and play with it (adjust the relative humidity and your distance from the performers with the scroll-bar control sliders), you’ll find that the low-pass shoulder frequency shifts to the right (to higher frequencies) as the humidity increases. You’ll find that the slope of the sound pressure level (SPL) roll-off in dB/octave is maximum around RH=25%. You’ll find that the slope is about 2 dB/octave at very low and very high humidities, but can be 7 dB/octave or more (for listeners seated far from the stage in a typical-size chamber music concert hall) when RH is between 20% and 30%.

What this suggests to me is that, to optimize the listening experience for as many concert-goers as possible, presenters should try to avoid humidities in the 20% to 30% range—tough to do in wintertime in climates with outdoor ambient temperatures below –10 ºC.

Hypothetically, it would be wonderful to perform in very low-humidity air—say, chamber music festivals in the mountains at 4,000 meters or higher; or venues (Arizona? New Mexico? inland California? Alberta?) having peculiarly dry air like what I encountered in Minneapolis. Or, conversely, the timbre could be similarly sweet at low altitudes, in air that has been conditioned to between 55% and 70% relative humidity.

(For humans, relative humidity less than 25% feels too dry. The normal recommended comfort range is 25% to 65%. Above 65% many people feel clammy or uncomfortably moist. Many concert halls and other public buildings are conditioned so as to stay close to 45% relative humidity.)

An unexpected surprise from my statistical modeling and analysis is that tropical locations with relative humidity in the 70% to 90% range should have even sweeter, more lively timbres. Yes, such high humidity levels would have some adverse effects on tuning for stringed instruments and pianos, but the liveliness of the sound might be pretty interesting—noticeably warmer than usual, especially for listeners 20 meters or more away from the stage. And, of course, choral and wind instrument performance—no problem at all at such high humidities.

I’d be grateful to hear from you regarding your own acoustical experiences in unusually dry or unusually humid indoor performance venues. If you live or perform in equatorial locations, your impressions would be especially interesting. Any chamber music presenters or organizers of festivals in warm, humid locations out there with comments to share? Do any architects out there reading this blog post have advice based on your concert hall design projects?

With global warming, I suppose the future for most of us will tend to be more humid on average, compared to conditions to which we’ve been accustomed…

NOAA Natl Climate Data Center

25 December 2007

Yotcheva, Franck, Musical Quotation, Minimalism and MIR?


A  n assumption that Minimalism and Baroque are both basically musics of tightly patterned cyclic repetition ... led the director of the Eos Ensemble to
devise a concert program under the rubric ‘Perceptible Processes: Minimalism and the Baroque.’ Scheffer’s belief in his ear’s ability to forge trans-historical links is absolute: ‘Think of the Bach Suites for solo cello and their long series of arpeggios through the circle of fifths. They are at once expansive in time and rigid in gesture. Now think of Philip Glass, of ecstatic arpeggios and simple harmonic changes, and you may begin to hear a great unison sounding throughout the history of music.’ ... ‘Bach the humble craftsman’ is, of course, no more historical a reading than ‘Bach the protominimalist.’ But it does move us from the domain of technique to the domain of music’s societal function. Perhaps the link between Minimalist and Baroque [or Romantic] music is not
how they are produced, but how they are consumed.”
  —  Robert Fink, p.171-2.
Vélitchka Yotcheva’s new recording of Franck’s ‘Sonate en la majeur pour violon avec piano’ FWV 8 (1886) on XXI-21 Records is wonderful—and it contains some novel and convincing interpretations. (Patrice Laré’s piano playing on this disc is admirable as well.) Yotcheva, founder of Trio Rachmaninoff, performs regularly in Canada, Russia, Bulgaria, France and the U.S. Her exuberant style epitomizes the Russian school of cello playing, which she comes by naturally (via her training at the Tchaikovsky Conservatory, with Marina Tchaikovskaya and later Valentin Feigin and Alexandre Korchagin). She completed her doctorate in 2000 at the University of Montreal, with Yuli Turovsky.

Franck wrote this sonata as a wedding gift for his fellow Belgian, the violinist Eugene Ysaÿe, who gave the premiere in Brussels in November 1886. This cello sonata is an arrangement of the violin version, made shortly after Franck’s death. Franck is usually thought of in terms of harmonic textures and compositional idioms more complex and multi-dimensional than those of other Romantics. One of Franck’s snarkier students (Debussy) termed him a chromatic ‘modulating machine’ (‘Une machine à moduler! La theme, c'est comme la confiture, moins on en a, plus on l'étale. À chaque jour suffit sa peine! Il a une araignée au plafond! Chantez à l'âne, il vous fera des pets! Grosse Corvette, petite cervelle!’), and over the years that characterization has somewhat ‘stuck’. But from Yotcheva we hear new possibilities in this sonata, with more emphasis on its cyclical structures. Her articulations reveal a new logical sense of Franck’s modulations and rhythmic patterns—reveal how simplistic attempts to pigeon-hole Franck’s compositional style do him (and us) great disservice.

Themes from one movement in this sonata are transformed and used over subsequent movements. The whole thing devolves from the opening of the first movement. In a manner that anticipates 20th Century minimalism, every new theme in this sonata turns out to be a subtle variation on a previous theme. For example, the piano’s chords at the beginning of the Allegretto introduce a theme that the cello takes up, and this becomes a recurring motif throughout the whole sonata. The mood changes completely at the passionato second movement, but the original theme reappears here, too, as it does again in the Recitativo–Fantasia movement. The finale is a canon in octaves, with one voice following the other at the interval of a measure, but it, too, is derived from the sonata’s opening theme.

B  ach and Mozart were [‘summational’ artists] dealing with known vocabularies and an accepted body of aesthetic principles.”
  —  Gene Lees, The Will to Swing, p. 192.
A  composer’s choice of time signature is an essential element of the compositional process. When Franck chose to use the 9/8 time signature in this sonata, he did so under the influence of the notational and temporal conventions he had inherited from composers before him. But he also adapted it to his own purposes, with some new compositional procedures to create novel expressive effects, including projection of affect and emotion. Franck’s characteristic harmonic language (such as his iambic melodic rhythms in this sonata) is an example of this. Take a look at what we find by running some match-by-incipit queries in ThemeFinder, a joint music information retrieval (MIR) project of the Center for Computer Assisted Research in the Humanities (CCARH) at Stanford University and the Cognitive and Systematic Musicology Laboratory (CSML) at the Ohio State University.

 Franck, Sonata in A maj, Mvt. 1, mm 1 - 11
I t is a mistake to consider this meter (9/8) as a 3/4 meter whose beats consist of triplets. He who has only moderate command of performance knows that triplets in 3/4 meter are played differently from eighths in 9/8 meter. The former are played very lightly and without the slightest pressure on the last note, but the latter heavier and with some weight on the last note…if the two meters were not distinguished by special qualities, all gigues in 6/8 could also be written in 2/4; 12/8 would be a C meter. How senseless this is can easily be discovered by anyone who rewrites, for example, a gigue in 12/8 or 6/8 meter in C or 2/4 meter.”
  —  Johann Kirnberger, p. 396.
The opening phrase of the last movement in this sonata is surely original. But in its development, when it strays from the major into the minor, does it sound like Schumann’s Arabesque? Was Franck hoping to draw upon his audience’s likely familiarity with the Schumann Arabesque? Was he hoping to increase his ‘page rank’ by referencing the Schumann Arabesque?

Schumann, Arabesque, Op. 18, Minore I
Or, instead, ought Franck to have expressly avoided developing his idea in A major the way that he did, lest he evoke an arm’s-length resemblance to Schumann?

No. This sonata is genuinely Franck through and through and not plausibly a ‘borrowing’, regardless of any coincidental resemblance. Should van Gogh have refrained from painting yellow flowers, in deference to other painters who’d chosen yellow flowers as a motif previously? Of course not. Brent Smith, writing in Music & Letters in 1931 (12:71-7), felt that two or more composers might naturally be inspired by the same subject, and that similar motifs might be expected to arise independently. In works that are minimalistic—or in performances like Yotcheva’s that enable us to uniquely detect minimalistic tendencies in works where we’d never noticed them before—such similarities are surprisingly common.

Of course, ‘musical similarity’ metrics are a hot topic in music information retrieval (MIR). It’s notoriously hard to measure musical similarity in a fair and robust way, or to compare the classification accuracy of different MIR systems and MIR software algorithms. The best that can be done is to assemble panels of musicians who represent a range of musical expertise and points of view, and ask them to classify pieces according to a set of standardized attributes, as to whether pairs of pieces are ‘like’ each other—create a gold-standard reference database of pieces of expert-consensus similarity. Then do Cohen kappa, Kendall tau, Cronbach alpha, and other statistical tests to see how much concordance or disagreement there may be between different MIR systems being compared. Look at ‘false-discovery rate’ (FDR), as a genomics software package would do. Look at false-hits and false-misses. Statistical sensitivity and specificity.

Text IR has an easier time compared to MIR insofar as there is—at least at a basic (‘dictionary’) level—a high degree of agreement about the meanings of individual words, despite the obvious problems caused by ambiguities, mis-spellings, contextual alterations of meaning and all the other things that make human language such a rich means of communication. But music has no semantic unit which corresponds in any useful way to a word, and certainly no agreed vocabulary or ontology of ‘meaning’. While it’s possible to decouple the individual ‘atomic’ elements that make up a musical work, none of those elements ‘makes sense’ on its own. The sense of each derives from the relations to the other elements around it, plus ‘summational’ references to other works before it. Yotcheva manages to create an astonishing new ‘sense’ of the elements in this Franck sonata. You think you know this sonata? You may want to listen to Yotcheva and think again. Deeply moving, original!

T his blog is, admittedly, an atypical classical music site. CMT’s not a tlog, and it’s not a splog. I’m grateful for the people I’ve been messaging with here ... and for those who share these interests, including seemingly-bizarre conjunctions (like performance practice, composition methods, musicology, computer science (MIR), musical genres (Baroque, Romantic, Minimalist), musical borrowing, sociology of Web 2.0 cross-links, and search-engine idiosyncrasies—as in this post). I wonder whether Debussy might’ve had Asperger’s or Tourette’s? ... Happy Holidays! Peace!”
  —  DSM.


22 December 2007

La Pasión de Rubén Lorenzo: El Sentimiento Falla’s Fantasía Bética Dentro y Fuera de España

Rubén Lorenzo

I  f Europe as a whole distinguishes today between Norwegian and Russian music it is because many years of practice in their grammar have taught us to distinguish them; yet, it would be easy to demonstrate that what Europe understands as ‘Spanish’ ... is nothing but a very limited aspect ... no more than a certain idiom fashioned principally on an Andalusian base ... Formal schematicism and brevity are essential qualities in the Concerto ... Not only is there never any filler or adornment in the instrumental parts, but these appear with abstemiousness in the extreme. Even the softness of the most idiomatic ranges is avoided, resulting in a timbre like that of old music played by primitive instruments, rude and astringent.”
  —  Adolfo Salazar, El concerto, 1927.
I  n the present Blair-ite age of cultural thinness, it’s all too easy to feel depressed at the comparative indifference shown towards serious music and the arts, and deplore their lack of purchase on our national life. Imagine such contemporary figures as Birtwistle, Benjamin and Ades being conscripted into the search for a redefinition of Englishness in the face of the threatened breakup of the UK and further integration into Europe... By contrast, Carol Hess’s excellent study, ‘Manuel de Falla and modernism in Spain, 1898-1936’, reveals the composer in the context of the Generation of ‘98. Appalled by their country’s shattering defeat by the USA in the Cuban War, these passionate intellectuals proceeded to conduct intense self-examinations on questions of national identity and foreign influences. Musical controversy at first reigned over the merits of the indigenous zarzuela—its conventionalised nationalism (españolismo) appealing to isolationists as opposed to the more Eurocentric cult of Wagner which gripped cosmopolitan Barcelona. … With an agenda of his own, uniting the Latin cultures of Spain and France, [Falla] defended European music’s ‘racial borders’ from the hegemony of the German tradition, and joined Unamuno in condemning the rationalistic legacy of Protestantism with its lack of ‘sensory grounding’… In the eyes of his numerous detractors, however, the modern French school had now replaced Wagnerism as the musical threat to the Spanish sense of ‘race’.”
  —  Andrew Thomson, Musical Times, Autumn 2002.

Mañuel de Falla
I  n the reception of Falla’s works in Spain, self, circumstance, and music intersect in compelling patterns that expose the innate idiosyncrasies in the
‘narrative urge’. It also reinforces the idea that music cannot exist as ‘the score itself’, reified in some utopian preserve where aesthetics alone holds sway, but as an utterance whose meaning will shift depending on the premises of a particular environment. The political Right promoted both Falla’s music and his character, emphasizing the period from 1932 on. In concentrating on this period, the Franco government was able to find powerful ammunition for its propaganda machine so that, with censorship wiping out all other points of view, franquismo singlehandedly laid the foundation for Falla’s legacy in Spain. There were, of course, some major discrepancies between this Rightist orientation, with its fustian diction and warped nostalgias, and the reality of Falla’s life... There is nothing to suggest that Falla actively opposed the principle of religious freedom. He was willing to criticize the administration and practices of the Church as he was the vandalism against holy sites. The latter he saw not so much as an attack on a corrupt power as a direct assault on God, and he feared that the Republic was bent on destroying not only the external trappings of the Church but the religious impulse itself... We do not know the composer’s understanding of the darker aspects of the Nationalist enterprise, namely, the role of its German and Italian allies, the effects of fascism throughout Europe, and Hitler’s designs. Nor could Falla have had foreknowledge of the virulence of Franco’s post-war reprisals. Having gone to Latin America for professional reasons and with every intention of returning, he was not a political exile, as is sometimes suggested. His rejection of Franco’s enticements no more indicates opposition to the regime than his donations to Republican camps and advocacy for friends on the ‘wrong’ side imply support for the vanquished Republic.”
  —  Carol Hess, Sacred Passions, p. 285-7.
Rubén Lorenzo provided an intriguing and passionate account of de Falla’s Fantasía Bética as part of his recital on Monday at Carnegie Hall. Lorenzo is noted for devising programs that he calls ‘monographic’ in their anthology-like scope, surveying the Spanish piano repertoire. But Monday’s recital was no dry academic monograph. Instead, it had more the impulse of an exciting suspense novel!

According to Carol Hess, the rejection of ‘localist andalucismo’ and adoption of a ‘universalist neoclassicism’ was a strong and recurring theme in Spanish modernist compositions in the first half of the 20th Century (2002, p. 4) and Falla was one of Spanish modernism’s leading exponents. Hess takes universalism to be stylistically and philosophically equivalent to neoclassicism, in part based on the force of Stravinsky’s enthusiastic endorsement of Falla’s late works and Stravinsky’s analysis of the compositional structure of Falla’s oeuvre. Like Hess, Falla biographer Tomás Marco (pp. 31-42) also notes Falla’s use of neoclassical compositional techniques in place of strictly nationalistic, andalusian ones—especially in his later compositions.

De Falla’s output ranges from late-romantic pieces to evocations of flamenco to the later neoclassicism. Lorenzo’s Bética emphasizes the piece’s flamenco features, maybe with more force and less restraint than is typical of flamenco. It seems that Falla was reluctant to let go of idiomatic flamenco elements, even if he eventually felt hemmed in by them. Maybe his life also manifested psychological ‘retentiveness’ in other ways, too: his conservative-flavored Roman Catholicism reportedly brought him both comfort and trouble throughout his life.

Incidentally, Bética is the Spanish name for the Roman province that is now Andalusia, transliterated from ‘Baetis River’ (now called the ‘Guadalquivir’). Fantasía Bética was dedicated by Falla to Artur Rubinstein, who used to play frequently in Spain and liked to perform vernacular Spanish music, including the Ritual Fire Dance. Bética’s nominally the final piece in Falla’s flamenco period, flanked by ‘El Amor Brujo’ and ‘Three-Cornered Hat’.

I  t is easy to let Falla’s ‘complete respect for and loyalty to his Excellency, the Generalisimo’ and praise for Franco’s ‘undefeated soldiers’ slide into complete oblivion. But we must ask to what extent we ought to condemn a sickly [and, by then] old man whose faith, although it blinded him to many bad realities, caused him to make remarks that now seem tragically flawed. Unlike Picasso, who refused to exhibit Guernica in Franco’s Spain, or Casals, who turned down invitations to play there as long as Franco remained in power, Falla could have spoken out after the War. That he did not do so may make us uneasy today. But one who sees God before all else may become curiously tongue-tied over temporal matters.”
  —  Carol Hess, Sacred Passions, p. 288.

Piano mechanism
Q  ué estimulante pensar en el futuro! Porque la música comienza a avanzar ahora mismo; la armonía aún debe ser un medio artístico. Por ejemplo, las canciones de Andalucía están basadas en escalas con intervalos más sutiles que las escalas en las que la octava es dividida en doce semitonos. Como compositor, todo lo que puedo hacer por el momento, es crear la ilusión de esos cuartos de tono, por medio de la superposición entre acordes de una tonalidad y otra.

[How stimulating to think about the future! Although music begins to advance right now, harmony still should be an artistically legitimate means of expression. For example, the songs of Andalusia are based on scales with more subtle intervals than the scales in which the octave is divided into twelve semitones. As a composer, all I can do for now is create the illusion of [reconciliation between] those realms of tone, by superimposing harmonies of one tonality on another. ] ”
  —  Manuel de Falla, 1919.
It was a revelation for me to discover that in the 1920s and 1930s Falla had remarked on aspects of his works that exhibited what today would be termed ‘microtonality’. Way ahead of his time! The biographers seem either not to have noticed those remarks, or else did not think the concept of microtonality sufficiently important to merit coverage in the context of Falla’s many accomplishments and overall significance in generating a Spanish national identity. Or else the biographers simply weren’t familiar with microtonality per se and didn’t know what to make of the remarks. Hmmm.

Piano mechanism
And how often do we think of microtonality in the context of solo piano works? Not often! The default prejudice is to think of a piano as tuned to a conventional temperament. It is not like a fretless stringed instrument or a wind instrument that’s amenable to an infinite range of pitch bending. When we think of piano, we think of note heads sitting precisely on staves, of white and black keys whose depression neatly yields sounds that have highly predictable pitches. We do not ordinarily expect a piano to be capable of microtonality. And yet here it was! Had the interpreter been one whose technique and athleticism and emotional exuberance were any less than those of Lorenzo we would not have heard the microtonality that this Falla piece contains. Had the recital hall been one whose acoustic properties were inferior to those in the Carnegie Weill venue we, likewise, would not have heard it. But the confluence of this Lorenzo, this Steinway, and this nice recital hall gave us this unique revelation. I had come with the intention of hearing Lorenzo educate me about Spanish culture and politics, but instead was delighted to receive this unexpected ‘bonus’—a revelation about piano physics and a revelation that Falla had been deliberately exploring microtonality decades before it became fashionable to do.

The Russian physicist Alexander Galembo has done considerable research on microtonal capabilities of the piano. If you are interested in the microtonal possibilities of conventional (‘unprepared’) piano, you may like to try to locate a copy of Galembo’s book. It is out-of-print and it is in Russian. But it’s a relatively slender volume, and Galembo’s writing is clear and concise, so with even a modest ability to read Russian it is possible to comprehend Galembo’s points.

Galembo book
T he microtonal effects are a result, mainly, of velocity effects and both vertical and horizontal motions of the piano strings after they have been struck by the hammers—and a result also of the detailed vibrational modes of the soundboard. The most interesting factor contributing to the existence of aftersound is the presence of more than one string for each piano note, and the consequent dynamical coupling that occurs among the strings struck by the same hammer. The data indicate clearly that we are dealing with two independent modes of vibration, which are producing sound waves through two separate radiating ‘antenna patterns.’ It is a dynamic thing: at various times after the attack, one or the other mode dominates—that is, near the beginning the dominant mode is different than it is near the end of the note. In fact, it is quite reasonable to suppose that the vibration pattern of the soundboard in response to a vertical force at the bridge is quite different from what it is for a horizontal force.”
  —  Gabriel Weinreich, KTH, Stockholm.

[Lorenzo is 48. Besides his extensive repertoire of Spanish music he also has performed and recorded Makrokosmos by Crumb. He is currently professor of the Upper Conservatory of Music of Zaragoza.]

Carnegie Hall, Weill Recital Hall

Carol Hess book