jwilzewske: Undeutlicheit: Acoustic Surprise in Extemporaneous Performance

DSM: Pomerium performed in The Cathedral of the Immaculate Conception in Kansas City last night. That cathedral sanctuary is beautiful but cavernous—a rectangular shape about 40 meters long, by 20 meters wide, by 20 meters high, with a horizontal-half-cylinder plaster ceiling. It’s a cruciform building plan with a nave crossed by a stubby transept, and the stone altar is in the center of the large nave, raised several steps up from the main floor. The members of Pomerium assembled themselves on those steps in front of the altar. The audience in their winter-time clothing filled the pews in the long part of the nave, so there was an acoustically absorptive-dissipative effect over there. And the upper part of the nave, beyond the transept, was empty pews all the way to the apse—a highly echoic effect in that direction.

CMT: Yes, I was there. Several times during the performance it seemed to me that Alexander Blachly, the director, was re-gauging the reverberation in the cathedral—making adjustments in his direction for attacks, releases, dynamics, and so on, based on his evolving impressions of the acoustics and how the acoustics of the sanctuary was affecting the sound of the various parts and registers. Despite Pomerium having performed in venues like this since the group’s founding in 1972, he and the singers seemed last night to be genuinely and continuously ‘surprised’ by the characteristics of this particular cathedral. I could only imagine that he was wishing he’d had a chance to more extensively check the acoustic properties of the church in advance of the performance!

DSM: Well, there’s checking and then there’s checking. Only a standing ensemble who routinely performs in a familiar, consistent venue really has a chance to truly know and adapt to the strengths and limitations of that venue. What could you do, besides a rehearse in the room? A well-known and basic method for characterizing the acoustical characteristics of a room is to measure the response at a particular microphone position to a brief source of energy, such as a “blank” pistol shot (not usually done in churches) or a hand clap. The use of a deterministic signal (e.g., long-length sequence, sine-wave sweep) is also possible. From the perspective of room acoustic quality, the result usually involves visual inspection of a graphic display of the “room impulse response”, that is, the root-mean-square pressure of the real part of the analyzed signal, measured in decibels as a function of time. A similar sort of graphic can be obtained from a modeling program that uses ray tracing or other techniques for predicting, rather than measuring, the room impulse response. This information can be used for both analyzing the acoustics of a real room or for simulation of the acoustics of a virtual room. You wouldn’t find regular musicians doing that sort of thing—especially not on the budgets that typify chamber music and early music. But, if you had your ‘druthers’, that’s the sort of thing that you’d like to do…

CMT: And then post-analysis of the reflection amplitudes relative to the level of the direct sound would determine their significance in terms of audibility and clarity. Early reflections are well-known to be potentially detrimental to timbre reproduction, speech intelligibility, and the formation of spatial images in a sound field.

DSM: You remember in Bang&Olufsen’s old ‘Archimedes’ project, the hypothesis to be tested in the project was that the interaction between the sound source and the room has an influence on the perceived sound quality (timbre)? The hypothesis was verified in a Round Robin Test (RRT) in different listening rooms, which included as variables the room, the directivity of the source, and position of the sound source in the room. The results showed that the room has a significant influence on 1) the reproduction of timbre in all positions for all sources, 2) timbre differences between different sources in the same position, and 3) timbre differences between the same source in different positions.

CMT: Yes, in the B&O RRT, the purpose of the project was to examine the influence of individual reflections on the timbre or localization of sound reproduced by a single source in a listening room. The setup simulated the physics—mathematically speaking, the ‘transfer function’—of individual reflections in addition to the attenuation due to distance. The simulated transfer functions included the frequency response of the off-axis angle of the reflection path from the original loudspeaker and the frequency dependent attenuation of the reflection from the simulated room surfaces. The first-order floor reflection contributes on an individual basis to the timbre of the sound source for a sound field that has many harmonics—a noise signal, say; or an ensemble with lots of polyphony, many players and parts. An increase in the level of the first-order reflections either from the floor, the left-hand wall or the wall behind the listener will influence the timbre of later-arriving reflections and noise.

DSM: The threshold of detection for timbre and a noise signal seemed in that study to be determined by the spectral changes in the frequency range 500 Hz —2 kHz The threshold of detection for localization of the later reflections and noise seems to be determined by the spectral changes in a frequency region above 2 kHz.

CMT: I was thinking about that last night. Despite the fact that I sat in Row #2 not 10 meters from the center of Pomerium members, the soprano and alto and countertenor voices and their reflections were far more clearly audible than the bass. And as the singers changed positions with almost every piece, we had an unusual—Unintended!—opportunity to observe the effects on the different parts and registers. The baritones seemed to be immune to positional effects. The tenors were, at various points, the hapless victims.

DSM: You, pitying tenors? Hah! Although it’s common to characterize hall acoustics in terms of a single number—reverberation time—there’s much more detail that affects how listeners perceive the music, and how the performers and the director interact with each other. In a large room, listeners hear the direct sound of the performers, followed by many variably-delayed reflections of the sound—bouncing off many different surfaces that are at different distances from the performers and different distances from the listener. Early-arriving reflections within about 50 msec after the direct sound are especially important—our brain tends to associate them with the direct sound, and we perceive it as louder rather than being diverted by a “different” sound. (This was published by Haas in 1951, but it was something others had known about for almost 100 years before that.) Later-arriving reflected sounds are not processed by the brain in this way, and cause one sound (note) to mask or interfere with the next. I think this is partly what was going on among the members of Pomerium last night.

CMT: Lochner and Burger developed the concept of “useful-to-detrimental” sound ratios, where ‘useful’ is the sum of the intensities of direct and early-reflected sounds and ‘detrimental’ is the sum of the late-arriving sounds plus background noise. Obviously, the judgments ‘useful’ and ‘detrimental’ are too simple for music—the ratios were developed initially for studying intelligibility of speech, not music. But other not-so-simple but more appropriate measures have been proposed—ones that take into account some of the values in musical aesthetics. Gade’s ST1 metric, for example. That one is basically a metric that quantifies a musician’s ability to hear himself/herself and maintain a graceful balance of his/her own instrument or voice with that of the others.

DSM: You could see in the singers’ eyes the fact that they were exploring that cave, that sanctuary—discovering as they sang their way through it just what acoustical support the chamber was giving (or not giving) to their performance. The subjective impression you could read in their eyes was “…I need to have some sensation that my voice has filled this chamber … I am trying to find out when my voice has reached the last row of seats … I want to feel that my voice has slapped back to the altar here.”

CMT: Yes, really, what impressed me last night—apart from the beauty that Pomerium achieved in their singing—was their esprit de corps in adapting to the idiosyncratic acoustic properties of the cathedral. The uncertainty of the acoustics—in a chamber where the performers didn’t have the luxury of being able to rehearse and calibrate themselves to the reverb or other effects in advance, and therefore have to discover and adjust their performance on-the-fly—this didn’t daunt them. In fact, it’s a type of “inadvertent improvisation” that’s typical of early-music performances in cathedrals or other historical venues. Some performers are annoyed by it, I know. But, facing the challenges head-on, Pomerium seemed to find the “discovery” aspects of extemporizing their performance in that cavernous chamber illuminating. Sort of like camaraderie in rock-climbing or other “extreme sports”. When successful, this does tend to bring the aesthetic-athletic feeling of “winning”. Maybe early music in period-appropriate settings is inherently ‘extreme’ chamber music!

DSM: So is it the shape of the chamber that matters most? Is it the sheer volume or size of it that matters? Is it the position of the sound source and the position of the listener that dominate these things? Are some particular registers of the instruments or voices affected more than others, based on the geometry of the room and some resonances or harmonics that the room supports? Is it the “people density”—the room’s volume in cubic-meters-per-person?

CMT: All of those things, in various complicated ways. You might want to have a look at one or more of these recent books on acoustical physics and architectural acoustics.

jwilzewske

Friday, December 15, 2006

Undeutlicheit: Acoustic Surprise in Extemporaneous Performance

No comments:

Post a Comment