Effects of Choir Spacing and Riser Step Heights on Acoustic and Perceptual Measures of SATB Choir Sound Acquired From Four Microphone Positions in Two Performance Halls

Abstract

We assessed the effects of singing in two venues on acoustic and perceptualmeasures of performances by an SATB choir (N = 28) as it sang in three spacing conditions, on two different riser unit heights, with data acquired from four microphone locations in each hall. Among primary acoustic results, long-term average spectra measures yielded a systematic pattern of significant contrasts between close and spread spacing conditions, with decreasing high-frequency energy as space between singers increased, regardless of venue, riser unit, or microphone location. Unlike previous findings with 12-in. (0.30-m) riser step heights, a doubled 16-in. (0.41-m) height did not enhance this pattern. Primary perceptual results were as follows: (a) All choristers (100%) thought spread singer spacing facilitated best hearing of their own voices in relation to the rest of the ensemble. (b) Nearly all singers (96%) reported their most comfortable vocal production in spread spacing. (c) All choristers thought the doubled riser step heights influenced choral sound, but 78% described vocal production as most comfortable with spread spacing on the regular riser unit. We discuss results in terms of implications for choral pedagogy and directions for future research.

Keywords

choir spacing choral risers room acoustics choral pedagogy singing

As choral teacher-conductors go about the common task of positioning singers on portable standing risers, they might thank Nebraska music educator Magnus Paysen, who patented his design for such risers in 1930, commencing mass manufacture shortly thereafter (Daugherty, Manternach, & Brunkan, 2012). One aspect of Paysen’s continuing legacy informs the present investigation.

Paysen advertisements (e.g., MENC, 1943, p. 52) depict riser units accommodating the weight of as many people as possible with a simultaneous claim that choirs positioned on these risers “sound better.” Typically, contemporary manufacturers provide consumers with two occupancy calculations: one based on “performers [standing] shoulder to shoulder” and the other designed to accommodate still more singers with “performers facing center, shoulders turned in” (SICO Europe Ltd., 2017; Wenger Corporation, 2017). Both calculations, like Paysen’s advertisements, attend solely to maximum occupancy.

Three complementary lines of investigation have addressed possible effects of proximity of choir singers to one another. Foot (1965) found that adult vocalists sang with increased intensity in proportion to decreases in the intensity of auditory feedback from direct and reflected sound pressure waves to their external ears. He reported that these singers seemed unaware that the intensity of their singing changed as the acoustic environment changed. Tonkinson (1994) examined the Lombard effect among singers in choir-like conditions. Results indicated that singers, unless otherwise instructed, tended reflexively to raise their voices in the presence of loud sounds from neighboring singers. Both studies suggested the possibility of pressed-edgy oversinging when singers stood in close, shoulder-to-shoulder configurations.

Ternström (1989, 1994, 1999) explored self-to-other ratios among choir singers. He determined that choir singers exhibited defined individual preferences for balance between the aural feedback received from their own voices (self) and the sound received from the rest of the choir (other).

In studies by Daugherty (1999, 2003), the amount of intersinger space afforded choristers on portable risers significantly affected singer and audience perceptions of singing. Choristers and auditors alike preferred performances with spread singer spacing over close, shoulder-to-shoulder spacing, regardless of the standing formation employed (sectional, mixed, or synergistic). Singers in each study reported more efficient vocal production and better ability to hear their own voices with spread spacing. Daugherty (2005) found also that choir performances with spread spacing stayed significantly more in tune than with close spacing.

Ensuing studies in this line of investigation examined horizontal and vertical aspects of singer riser spacing acoustically as well as perceptually. Daugherty et al. (2012) positioned an intact SATB choir on two types of riser units: one with regular, 0.20-m (8-in.) step heights and the other with 0.30-m (12-in.) step heights. Long-term average spectra (LTAS) analyses of recordings acquired from one audience-position microphone showed significant mean amplitude reductions with spread singer spacing. These reductions appeared more robust (M = 2–3 dB SPL) in 2.4–4.3 kHz spectrum in and around the “singer’s formant” frequency region. Moreover, the taller, 0.30-m (12-in.) risers enhanced this effect by an additional 1- to 2-dB reduction, including a direct comparison of performances in close singer spacing on both riser units. This finding suggested that vertical space between and among singers might also affect choral sound phenomena.

In a follow-up study, Daugherty, Manternach, Coffeen, and Brunkan (2011) repeated this two–riser unit protocol with another SATB choir singing another musical excerpt in the same hall, this time with two microphone locations (conductor position, audience position) and focusing solely on close versus lateral singer spacing. LTAS and smoothed one-third octave band analyses showed a similar pattern of reduced spectral energy with laterally spread chorister positioning at both microphone locations. Once again, the 0.30-m (12-in.) step heights modestly enhanced this effect.

Listeners in both these studies readily perceived nuanced differences in choral sound attributable to varied intersinger spacing and riser step heights. They described the choral sound as more blended and balanced in spread spacing than close spacing. Similarly, Ford (2003) found that auditors significantly preferred a choral sound wherein singers intentionally employed less resonant energy in the 2- to 4-kHz frequency region.

Two researchers (Hom, 2013; Ternström, 1989) documented by controlled experiments what choir singers already may sense intuitively—namely, that differing room acoustics can affect choral sound by strengthening or attenuating the amplitude of high-frequency energy. These studies acquired recordings from one or two microphone locations within the venues examined, with a single, consistent intersinger spacing. Sugrue (2013) compared ensemble singers’ perceptions and preferences in three performance venues with various acoustic venue measurements. He found moderately strong correlations between singer preferences and reverberation time, early decay time, and overall strength of room response.

Because multiple recording and listening fields (near field, mixed field, diffuse field) may exist within the same room and because assessing intersinger spacing conditions in these varied sonic fields may yield somewhat differing spectral features, analyses of recordings acquired from four consistent microphone positions in contrasting venues could interest choral teachers. Moreover, some choirs, because of their size or the dimensions of their performing venues, cannot readily implement horizontally spread spacing on portable riser units. Given the modest yet discernable changes in spectral energy obtained when riser step heights increase from 0.20 m (8 in.) to 0.30 m (12 in.; Daugherty et al., 2011; Daugherty et al., 2012), we wondered if choirs with limited horizontal “real estate” might benefit from doubling available vertical space from 0.20-m (8-in.) to 0.41-m (16-in.) riser step heights.

The purpose of this study was to assess possible effects of singing in two different venues on acoustic and perceptual measures of performances by an SATB choir (N = 28) as it sang the same musical excerpt in three chorister spacing conditions (close, lateral, circumambient) on two portable riser units (standard: 0.20-m [8-in.] step height; tall: 0.41-m [16-in.] step height), with recordings acquired from four consistent ear-height microphone locations in each hall (conductor position, front-row audience, midaudience, back audience).

The following research questions guided this investigation:

Research Question 1: Are there significant acoustic differences among this choir’s performances (N = 12) in three singer spacing conditions on two types of choir risers as measured by LTAS analyses of recordings obtained from four microphone positions in two performance halls?

Research Question 2: Do choristers express preferences for singing according to the varied horizontal (close vs. spread spacing) and vertical (regular riser units vs. risers with doubled step heights) intersinger space afforded?

Method and Procedures

Participants

An intact university choir served as the choral ensemble for this study. Choristers (N = 28) ranged in age from 18 to 33 years and included 20 females (11 sopranos, 9 altos) and 8 males (4 tenor, 4 bass). For this investigation, we randomly assigned positions to the singers within a block sectional formation. Singers remained in their assigned rows throughout the study.

Sung Excerpt

The choir performed from memory an excerpt from “Domine Tu Mihi Lavas Pedes” (duration: 1 min 57 s), a four-part cappella motet by Brazilian composer Jose Mauricio Nunes Garcia (1789; 1760–1830). In consultation with the choir’s conductor, we chose this excerpt from repertoire that the choir was rehearsing. It was largely homophonic and, within the 24 measures chosen, exhibited some variety in vocal ranges and suggested dynamics. We desired an excerpt of approximately 2 min for LTAS analyses. We ensured that the choir practiced this motet for like amounts of time on the two riser units in the three intersinger spacing conditions of interest in the week prior to the recording sessions.

Recording Venues

Two rooms (Room A, Room B) used for university ensemble performances served as recording venues for this investigation. Room A, a dedicated 350-seat recital hall, had a raked floor, upholstered theatre chairs, and a stage. Distance from the upstage wall to rear auditorium wall was 24.38-m (80-ft), with a 14.94-m (49-ft) distance from the downstage edge of stage to the rear wall. Average room width at audience seating was 14.94 m (49 ft). Average ceiling height at audience seating was 6.40-m (21-ft). Room A contained wood panel walls and a slanted ceiling that provided some useful sound reflections from the stage to the audience seating.

Room B was a dual-purpose rehearsal-performance space in the shape of a trapezoid. Parallel walls measured 25.91 m (85 ft) and 19.81 m (65 ft), the perpendicular wall measured 18.23 m (60 ft), and the angled wall was 21.34 m (70 ft). The room had a flat floor, a 7.62 m (25 ft) high dropped ceiling suspended from a venue height of 10.67 m (35 ft), and no permanent seating, fixed stage, or performance platform. Most of the walls in Room B were constructed of sound scattering and absorbing concrete block. The dropped ceiling consisted of 0.61 - × 0.61-m (2 - × 2-ft) grids with sound scattering panels in some locations.

Room A evidenced a midfrequency reverberation period of 1.5 s. Room B was somewhat less reverberant than Room A, with a midfrequency reverberation time of 1.2 s. We calculated an equivalent absorption area (A) for each room by computing room volume (V) in cubic meters and applying Sabine’s formula (RT = 0.161V/A) using the obtained reverberation times (RT) reported here. We then compared the difference between the acquired equivalent absorption area for the two venues, obtaining a gain factor. By these procedures, the equivalent absorption area of the two venues differed by a factor of 2.28.

Noise criterion (NC) ratings acquired with conformity to the American National Standards Institute’s Standard S12.2-2008 were as follows: NC = 24, 32 dBA for Room A and NC = 34, 40 dBA for Room B (for graphs, see Appendix A in the online version of this article). Room B evidenced more ambient noise than Room A, likely due to a HVAC system that could not be turned off. The varied dimensions and characteristics of Rooms A and B thus offered two contrasting yet real-life venues for this study.

Performance Protocol and Equipment

During the recording sessions, the choir sang in the following six conditions: regular risers with (a) close, (b) lateral, and (c) circumambient singer spacing and tall risers with (d) circumambient, (e) lateral, and (f) close singer spacing. We randomly selected beforehand the initial riser and spacing condition. However, logistical management of riser and singer movements thereafter dictated the subsequent order.

For the regular riser unit, we used Wenger Tourmaster three-step risers as manufactured. The height between riser steps was 0.20 m (8 in.). The step width was 0.46 m (18 in.). Risers conjoined to form a modest semicircular curve. For the taller riser unit, we maintained all other dimensions of the regular riser unit but added an additional 0.20 m (8 in.) to the height of the two upper riser steps. Thus, the height between each riser step on the tall riser unit was 0.41 m (16 in.), double the step height of the regular unit.

Horizontal intersinger spacing conditions conformed to those used in previous studies (e.g., Daugherty, 1999, 2003). For close spacing, singers stood in a comfortable shoulder-to-shoulder stance with <1.0 in. (0.03 m) between the upper arms of contiguous singers. A consistent horizontal distance of 0.61 m (24 in.) between singers, as measured with dowel rods prior to each performance, constituted the lateral spacing condition. For circumambient spacing, singers retained the 0.61-m (24-in.) lateral distance and, in addition, left vacant the equivalent of a riser step width (0.46 m [18 in.]) among the three rows of singers. We accomplished this configuration by having the first row of singers remain in place while moving the riser unit back 0.46 m (18 in.). Thus, the second row of the choir stood on the first riser step, and the third row stood on the third riser step. At no point, however, did the location of the first row of singers change.

Videotaped conducting ensured that singers responded to the same conducting stimuli in each performance. It thus served to control for possible confounding variables due to any changes between performances in conductor gesture, facial expression, and tempo.

We used four Earthworks precision-instrumentation, omni-condenser microphones (model M30, n = 2; model QJC50, n = 2) to capture each performance at a sampling rate of 44.1 kHz (16 bits) in .wav format. We placed the four microphones in the center of the room at the same distances from the center front row of the choir in both rooms, as follows:

Conductor position microphone: 3.07 m (10 ft 1 in.) at a height of 1.62 m (5 ft 4 in.) or standing conductor ear height

First row audience position microphone: 7.32 m (24 ft) near a middle section center seat at a height of 1.12 m (44 in.) or seated listener ear height

Midaudience position microphone: 12.29 m (40 ft 4 in.) at 1.12 m (44 in.) height

Back audience position microphone: 50.58 m (50 ft 7 in.) at 1.12 m (44 in.) height

Because Room B was longer than Room A, the back audience microphone was 3.68 m (12 ft 1 in.) from the rear wall in Room B, while it was 2.44 m (8 ft) from the rear wall in Room A. Audience position microphones were in the center of the room (Room A, center seat within the row; Room B, equidistant from the side walls).

We calibrated microphones immediately prior to recording in each hall using a Cirrus CRL 511E microphone calibrator against a known sound source of 1000 Hz at 94 dB. We used KayPentax Computerized Speech Lab Model 4500 software to examine recordings acquired from the four microphones. To obtain LTAS data, we analyzed each recording from each microphone position using a window size of 512 points with no preemphasis or smoothing, a bandwidth of 86.13 Hz, and a Hamming window.

Singer Questionnaire

Choir members completed a short questionnaire at the conclusion of the first recording session. This questionnaire replicated language used in previous studies (e.g., Daugherty, 1999, 2003) and ascertained singers’ perceptions of singing in the three spacing conditions on the two riser units. The questionnaire consisted of six items: (1) “What effect, if any, do you think spacing between singers had on the sound of this choir?” (2) “What effect, if any, do you think taller riser steps had on the sound of this choir?” (3) “I thought my own vocal production was MOST efficient in . . .” (selected spacing condition, selected riser condition). (4) “I thought my own vocal production was LEAST efficient in . . .” (selected spacing condition, selected riser condition).” (5) “In which arrangement were you generally able best to hear/monitor the sound of your own voice?” (selected spacing condition, selected riser condition). (6) “In which arrangement do you think this choir sounded best?” (selected spacing condition, selected riser condition).

Results

We present results according to the research questions posed for this study. For acoustic results, we applied a twofold standard: (a) findings appeared not to have occurred by chance, as indicated by a predetermined alpha level of .05 for statistical tests, and (b) any timbral differences met or exceeded 1 dB SPL, the just noticeable difference (JND) necessary for human hearing to detect changes in complex sound (Howard & Angus, 2006). We reasoned that any statistically significant differences in choral sound would have little practical import without some expectation that human ears might detect them.

Research Question 1: LTAS

LTAS measurement provides frequency and sound pressure density data across a given spectrum as averaged over a period of time. Graphs of LTAS present sound pressure level (SPL dB) as a function of frequency (Hz). These data supply a quantifiable index of sound quality, useful for detecting persistent spectral features during a specified time period.

Our first research question asked if LTAS data would indicate differences between the choir’s recorded performances in the two venues. Due to the volume of data collected (3 spacing conditions × 2 riser conditions × 4 microphone locations × 2 rooms = 48 data sets, each set with 117 data points), we report here grand means and grand mean ranges of signal amplitude changes according to spacing and riser comparisons across the four microphone locations per room.

Figure 1 displays grand mean LTAS acquired from four microphone locations per riser units, intersinger spacing conditions, and venues across the 1.8- to 3.9-kHz spectrum, a frequency region to which human hearing is particularly sensitive (Fletcher & Munson, 1933). Figure 1 presents a comparison of Room A and B grand mean LTAS for each spacing condition delineated by riser height (regular height, tall height). Appendix B (in the online version of article) presents LTAS graphs for the 0- to 10-kHz spectrum.

Figure 1.

Riser height comparisons (regular height, left; tall height, right) of grand mean LTAS (1.8–3.9 kHz) acquired from four microphone locations per intersinger spacing conditions and venues.

As can be observed, LTAS contours showed an overall pattern of intersinger spacing differences, wherein signal amplitude decreased as horizontal space between singers increased. This pattern persisted regardless of venue or riser unit.

Tables 1 and 2 show per venue grand mean SPL differences and grand mean ranges for each comparison of choir spacing according to each riser height in the 0- to 10-kHz and 1.8- to 3.9-kHz frequency regions. In all instances, the sound level of the more-spaced condition was subtracted from the less spaced condition in that pairing.

Table 1.

Room A Grand Mean SPL Differences and Ranges for Each Comparison of Choir Spacing According to Riser Height in the 0- to 10-kHz and 1.8- to 3.9-kHz Frequency Regions.

		0- to 10-kHz Frequency Region, dB		1.8- to 3.9-kHz Frequency Region, dB
Riser Height	Choir Spacing Comparison	Grand Mean Differences	Range	Grand Mean Differences	Range
Regular	Close vs. circumambient	2.2	0.5–3.8	3.1	2.4–3.8
Regular	Close vs. lateral	1.2	0.4–2.6	1.7	1.3–2.1
Regular	Lateral vs. circumambient	1.0	0.0–1.8	1.4	1.0–1.8
Tall	Close vs. circumambient	1.7	0.1–2.6	2.2	1.6–2.6
Tall	Close vs. lateral	1.0	0.2–2.3	1.6	0.7–2.3
Tall	Lateral vs. circumambient	0.7	0.1–1.2	0.6	0.2–1.2

Note. SPL = sound pressure level.

Table 2.

Room B Grand Mean SPL Differences and Ranges for Each Comparison of Choir Spacing According to Riser Height in the 0- to 10-kHz and 1.8- to 3.9-kHz Frequency Regions.

		0- to 10-kHz Frequency Region, dB		1.8- to 3.9-kHz Frequency Region, dB
Riser Height	Choir Spacing Comparison	Grand Mean Differences	Range	Grand Mean Differences	Range
Regular	Close vs. circumambient	2.3	0.3–6.1	3.8	2.1–6.1
Regular	Close vs. lateral	0.5	0.0–2.1	1.2	0.0–2.1
Regular	Lateral vs. circumambient	1.8	1.1–4.4	2.7	1.4–4.4
Tall	Close vs. circumambient	1.7	0.0–3.7	2.3	1.5–3.7
Tall	Close vs. lateral	1.2	0.1–2.6	1.8	0.5–2.6
Tall	Lateral vs circumambient	0.5	0.0–1.7	0.6	0.1–1.2

Note. SPL = sound pressure level.

We conducted a repeated measures multivariate analysis of variance using full spectrum (0–10 kHz) LTAS data with chorister spacing (three levels), riser units (two levels), and microphone locations (four levels) as within-subjects factors and with venue (Room A, Room B) as the between-subjects factor. All assumptions of the multivariate analysis of variance were met. Pillai’s trace multivariate test yielded significant main effects for spacing, F(2, 231) = 1,019.561, p < .001, partial η² = .898; risers, F(1, 232) = 481.270, p < .001, partial η² = .675; microphone locations, F(3, 230) = 2,532.878, p < .001, partial η² = .971; and venue, F(1, 232) = 3.991, p = .047, partial η² = .017.

As indicated by the ranges reported in Tables 1 and 2, all comparisons of performance pairs (n = 12) showed differences that exceeded the 1-dB SPL JND. Grand mean differences per pairs of contrasting performances met or exceeded the 1 dB SPL JND with three exceptions: (a) Room A tall riser lateral versus circumambient performances in both spectral frequency regions, (b) Room B tall riser lateral versus circumambient performances in both spectral frequency regions, and (c) Room B regular riser lateral versus circumambient performances across the 0- to 10-kHz frequency region (though not in the 1.8- to 3.9-kHz region). These grand mean exceptions warranted further inspection of venue-specific, riser unit, and microphone location data.

Venue differences

Disaggregation of LTAS data according to venue indicated increased relative sound pressure levels (M = 3.6 dB SPL) in performances sung in Room B as compared with performances sung in Room A. This increase was apparent regardless of riser unit employed (M = 3.57 dB SPL Room B increase with regular riser performances, M = 3.63 dB SPL Room B increase with tall riser performances). However, given the noted difference between the equivalent absorption areas of the two venues (2.28), the room acoustics, rather than singer output power, likely accounted for the 3.6-dB sound-level difference between Rooms A and B (Nordenberg & Sundberg, 2004).

Riser unit differences

Across spacing conditions in both venues, grand mean spectral energy (0–10 kHz) of performances on the tall riser unit increased or remained about the same when compared with regular riser unit performances, as follows: close spacing (Room A: M = 0.1 dB increase, range: 0–1.6 dB; Room B: M = 0.6 dB increase, range: 0–1.9 dB), lateral spacing (Room A: M = 0.3 dB increase, range: 0–1.3 dB; Room B: M = 0.1 dB decrease, range: 0–1.3 dB), and circumambient spacing (Room A: M = 0.6 dB increase, range: 0-1.4 dB; Room B: M = 1.2 dB increase, range: 0–2.5 dB). Only with the Room B circumambient comparison did the tall riser performance yield a significant mean increase (1.2 dB) in spectral energy that exceeded the JND standard. Comparisons isolating the 1.8- to 3.9-kHz region did not differ substantially from results obtained with the 0- to 10-kHz region.

Microphone locations

The LTAS examined from each of the four microphone locations in each hall indicated, as expected, that sound power decayed as distances from the choir increased. A single exception to this trend occurred in Room A, where there was a lesser reduction between conductor and back-audience microphones than between the conductor and the two other audience microphones, likely a function of the sloped floor in that room. At each microphone location in each room, the overall pattern of LTAS differences according to spacing condition and riser step height remained consistent.

Research Question 2: Singer Perceptions

We invited participants to share their perceptions upon completion of the first recording session. All but one chorister completed a singer questionnaire, yielding a 96.43% response rate.

Responses to Item 1 (“What effect, if any, do you think spacing between singers had on the sound of this choir?”) indicated that 100% of respondents thought that intersinger spacing had some effect on the choir’s sound, with 93% describing this perceived effect as moderate to much (no effect, n = 0, 0%; a little effect, n = 2, 7%; moderate effect, n = 14, 52%; much effect, n = 11, 41%; not sure, n = 0, 0%).

Item 2 (“What effect, if any, do you think taller riser steps had on the sound of this choir?”) responses indicated that 100% of participants thought that the taller riser step heights affected the choir’s sound, with 81% of choristers describing this effect as moderate to much (no effect, n = 0, 0%; a little effect, n = 5, 19%; moderate effect, n = 20, 74%; much effect, n = 2, 7%; not sure, n = 0, 0%).

Most respondents (96%, n = 26) answered Item 3 (“I thought my own vocal production was MOST efficient in . . .”) by indicating some version of spread spacing (close spacing, n = 1, 4%; lateral spacing, n = 17, 63%; circumambient spacing, n = 9, 33%). Most respondents (78%, n = 21) thought that their vocal production was most efficient on the regular riser unit, while six respondents (22%) favored the tall riser unit.

A majority of respondents (89%, n = 24) described their vocal production as LEAST efficient (Item 4) in close spacing (close spacing, n = 24, 89%; lateral spacing, n = 2, 7%; circumambient spacing, n = 1, 4%). Most choristers (70%, n = 19) thought that their vocal production was least efficient on the tall riser unit. Eight respondents (30%) thought that their vocal production was least efficient on the regular riser unit.

All respondents (100%) thought that they were best able to hear/monitor the sound of their own voices (Item 5) in spread spacing (lateral spacing, n = 5, 19%; circumambient spacing, n = 22, 81%). Twenty-one respondents (78%) thought the tall riser unit enabled them best to hear their own voices, while six participants (22%) thought that they could hear their own voices better when singing on the regular riser unit.

Item 6 asked, “In what arrangement did you think this choir sounded best?” Twenty-five participants (93%) responded with some form of spread spacing (close spacing, n = 2, 7%; lateral spacing, n = 15, 56%; circumambient spacing, n = 10, 37%). Most respondents thought that the choir sounded best on the regular riser unit (regular riser, n = 16, 59%; tall riser, n = 11, 41%).

Discussion

This study assessed the qualities of conglomerate choral sound when singers in an SATB choir stand close together or farther apart as they perform in contrasting venue and riser step height conditions. It also solicited singers’ perceptions of these spacing and riser variables.

Although limited to the methods, participants, and venues of this study, results of this investigation relate suggestively to findings from previous research, raise matters for future studies, and invite reflection on aspects of choral pedagogy theory and practice.

Intersinger Spacing

Primary acoustic findings with respect to close, lateral, and circumambient chorister spacing conditions indicate a patterned similarity of LTAS contours. Specifically, high-frequency energy largely decreases in proportion to increased horizontal inter-singer spacing. This effect is particularly notable from roughly 2 to 4 kHz, a frequency region to which human hearing is most sensitive. This finding confirms results from previous acoustic investigations (e.g., Daugherty et al., 2012) conducted with other choirs performing other literature in single venues with fewer microphone locations and with or without variations in riser step heights. In the present study, this acoustic pattern persists within each venue, on each riser unit, and at each of the four within-venue microphone locations, which suggests stability of the phenomenon across the independent variables examined.

Such stable sound-level differences are consistent with a reduction in singer power as the spacing between and among singers increases. This reduction, in turn, is consistent with the notion that singers tend to increase the level of self when the level of other increases (e.g., Foot, 1965) to maintain a useful self-to-other ratio (e.g., Ternström, 1999) and, conversely, to reduce the level of self if the level of other is lowered for economizing vocal effort. Perceptions of singers in this study appear to confirm this explanation, with nearly all singers (96%) reporting their most efficient, least effortful singing in spread intersinger spacing.

We note that these sound-level differences eventuate absent verbal instruction to manipulate vocal production. We suggest, therefore, that spread chorister spacing may set up within the immediate soundscape of the choir a Lebensraum environment that facilitates cooperative negotiations among human sound sources in ensemble. This environment—as opposed to close, shoulder-to-shoulder standing—may enhance the ability of singers to hear themselves appropriately in relation to the sound that they hear from the rest of the choir by reducing the more immediate sound absorption potential of closely situated neighboring bodies, improving the clarity of available early sound reflections, and perhaps conveying a mental sense of freedom rather than constriction. Within such an environment, singers, individually and jointly, appear to negotiate some nuanced changes in choral sound on their own, reflexively and quickly.

Prior studies indicated that auditors significantly prefer choral tone produced with less singer’s formant energy (Ford, 2003) and choral sound achieved by spread inter-singer spacing (Daugherty, 1999, 2003), describing it as blended or balanced. We choose on that basis to interpret in a positive light the amplitude decreases observed here with more spread spacing. However, this interpretation does not affect the acoustic pattern per se. Should one prefer choral sound with lesser diffusion of high-frequency energy, one could interpret positively the choral tone quality achieved with shoulder-to-shoulder singer spacing.

Reference to an ongoing more or less cordial debate among studio voice teachers and choir directors affords an instructive way to explain why nuanced reductions in the amplitude of high-frequency energy in choir sound may be important. Voice teachers—particularly those teachers who train singers in Western, classical vocal production—value increased amplitude in higher frequency energy among solo singers because it may indicate the presence of desirable robust resonance strategies that lend richness and color to solo vocal sound. However, choir directors, especially those who work with choirs whose members include trained and untrained singers, have to find ways for these disparate sound sources to “blend,” or perceptually sound as one conglomerate voice without individual voices sticking out.

To the dismay of many studio voice teachers, some choir teacher-conductors address blend by issuing one-size-fits-all verbal instructions that reduce the amplitude of higher frequency energy by directing choir singers to unnecessarily and artificially manipulate vocal production. For example, verbal directions to lip-round all vowels, maintain a consistently dropped jaw, or open the nasopharyngeal port by imagining a “ball” in one’s mouth can indeed reduce amplitudes of high-frequency energy. Yet, voice teachers might argue that this type of reduction comes at a cost to efficiently coordinated vocal technique. Less experienced singers who may rely on choir participation for learning about vocal production might assume that such lip rounding, jaw dropping, or mouth space is necessary for singing. Singers who pay good money for private voice instruction aimed in part at strengthening the amplitude of high-frequency energy may be exasperated by such instructions. Thus, a nonverbal, nonintrusive instructional strategy, such as an invitation to stand farther apart while singing, may be a particularly promising way to enable singers, especially trained or in-training vocalists, to individually adjust and achieve some small higher frequency amplitude reductions without consciously or artificially altering vocal technique. That most singers in intersinger spacing studies to date (Daugherty, 1999, 2003, 2005; Daugherty et al., 2011; Daugherty et al., 2012), which include trained and untrained vocalists or more experienced and less experienced choristers, reported their most efficient vocal technique in spread spacing lends additional credence to this possibility.

Chorister spacing studies to date employ high school, university, and adult choirs (Daugherty, 1999, 2003, 2005; Daugherty et al., 2011; Daugherty et al., 2012). Future investigations might employ ensembles of prepubertal children to test if the pattern thus far observed occurs among singers with smaller vocal folds and tracts. Chorister spacing studies to date examined mixed SATB choirs. Subsequent studies could include sex-specific ensembles of men’s or women’s voices, where source sound waves would be more similar in length.

It can be difficult in naturalistic performance situations to credibly isolate and objectively examine the vocal output and hearing of individual singers within a choir. Yet such data may be instructive. Subsequent investigations might feasibly equip some choristers with portable, battery-powered phonation and noise dosimeters. It might be possible with such instrumentation to correlate perceptions of individual singers with data about how they phonate and hear in various intersinger spacing conditions.

Riser Units Comparison

We suspect that Magus Payson’s decision to construct portable choir risers with 0.20-m (8-in.) step heights may stem more from building code regulations for staircases than from acoustic forethought. Moreover, some choir conductor-teachers do not have sufficient space or “real estate” in their given performance and rehearsal venues to implement spread spacing with their larger ensembles. Others may not have sufficient budget to purchase the additional riser units required for more spread ensemble spacing. Hence, our interest in testing riser units that afford more vertical space between rows of singers—that is, on units with step heights exceeding the 0.20-m (8-in.) elevations—stems from a practical problem confronting some choir teacher-conductors.

In the present study, performances on riser units with doubled, 0.41 m (16 in.) step heights exhibited a similar acoustic pattern among choir spacing conditions as performances on the regular riser unit. However, the taller unit in this study, unlike results obtained with 0.30-m (12-in.) step heights (Daugherty et al., 2012), did not appear to enhance this pattern with further reductions of mean signal energy. Indeed, when compared with performances on the regular unit with 0.20-m (8-in.) step heights, singing on the taller unit seemed to make little difference or it produced slight increases in mean SPL.

Most choristers (78%) reported their most efficient vocal production with the regular unit, while most (70%) perceived better hearing of their own voices on the taller unit. These perceptions may offer a possible explanation for what occurred acoustically with riser unit comparisons regardless of venue. It could be that doubly elevated riser rows disrupt singer expectations of airborne, reference sound received from the choir at large. With doubly elevated riser steps, in other words, sound pressure waves emanating from the mouths of singers in upper riser rows on the taller unit may not reach the ears of choristers in lower rows as efficiently or directly as the sound that comes from immediately neighboring singers on the same row, while the bodies of singers on adjacent tall unit rows absorb less vocal output than what may be the case with regular riser performances.

Although the choir’s conductor and each researcher remarked anecdotally upon the stunning visual appeal of the choir singers standing on doubly elevated riser rows, the 0.41-m (16-in.) step heights appear to function as a bridge too far acoustically and psychoacoustically. Subsequent research may investigate various riser step heights. If 0.30-m (12-in.) heights might enhance diffusion of high-frequency energy and 0.41-m (16-in.) heights appear to attenuate that effect somewhat, then perhaps 0.36 m (14 in.) or 0.25-m (10-in.) elevations might yield a happy medium.

Venue Comparison

The noteworthy finding of this investigation with respect to the two venues employed is that the overall observed pattern of LTAS attributable to intersinger spacing conditions remained stable across Rooms A and B. A limitation of the present study is that we did not calibrate sound levels within each venue. Although our calculation of equivalent absorption levels functions as an acceptable measure to examine and explain the mean 3.6-dB-level differences between the choir’s performances in Rooms A and B, subsequent investigations should employ calibrated sound levels.

Thus far, studies of intersinger spacing have occurred in venues regularly used for university and high school choir rehearsals and performances. Although these venues have differed in their particular acoustic characteristics, it would be instructive to ascertain not only whether the persistence of LTAS differences according to intersinger spacing observed here continues in other venues regularly used for choir performances but also whether these LTAS differences remain stable in more extremely reverberant or absorptive room environments occasionally used for choir performances—for instance, school gymnasia or carpeted, thickly draped hotel banquet rooms. In other words, what may be the room-imposed limits of singers to negotiate cooperative, nuanced changes in choir sound via changes in intersinger spacing distances?

Implications for Choral Pedagogy

Choruses chorus. That is, within any chorus multiple singers phonate simultaneously. This basic act of simultaneous phonation automatically engenders negotiation of acoustic and psychoacoustic relationships with other singers phonating within a shared venue environment. However, unlike ensembles of mechanical or manufactured instruments, choral singers bring to this negotiation process living, neurobiological instruments with a sound source and the organ of hearing embodied in the same apparatus.

Choir singers lack recourse to manual regulation afforded by valves, keys, drilled holes, or bowed or fretted strings. Instead, choristers depend primarily on airborne and bone-conducted phonation feedback to sense how to regulate internally the intensity and frequency of their pitched sound in relation to the sound that they hear from others. Data from this study suggest that attention to intersinger spacing may be one tool available to choral teacher-conductors for enabling singers to negotiate some of the challenges inherent to simultaneous phonation.

At the same time, data from this study suggest to us that the ultimate cause of singers responding the way that they did to the intersinger spacing conditions tested here may be human physiology, as informed by the computational power of human brains to assess and respond to environmental conditions. That contention—if borne out by subsequent studies exploring singer-centered, vocally nonmanipulative, and largely nonverbal approaches—may have vaster implications for the overall theory and practice of choral pedagogy.

In this respect, choral teacher-conductors and researchers alike may benefit from reflecting on the notion that choir sound is preeminently a context dependent event. Choirs produce sound from varying numbers of singers of different ages, experiences, and abilities, who may be in various stages of vocal health and voice change and who may exhibit differing degrees of hearing acuity. These singers join to perform a variety of scored and improvised literature in varied voicings with or without conductors, and they sing in a remarkable variety of venues. All these divergent, contextual variables appear to argue against large-scale imposition of one-size-fits-all pedagogical prescriptions for achieving nuances in choir sound, especially prescriptions derived solely from score-based data (e.g., Ehmann, 1968, among others, claimed that the musical score should determine the standing position of choir singers) or untested, anecdotal beliefs that sometimes inform choral practice.

One basic fact, however, appears constant: Human beings who populate choirs, although they may differ according to age, sex, nationality, and singing experience, bring to the negotiation of choir tone quality given parameters of anatomy, physiology, phonation, hearing, and human agency. We suggest, therefore, that choir conductor-teachers should be conversant with human vocal anatomy, hearing anatomy, physiology, and acoustics—that is, with how human voices and ears actually work. Such understanding can aid choral teacher-conductors in diagnosing choir sound more accurately. Even more, it may afford warrant and opportunities for vocal learning and problem solving that go beyond primary reliance on maestro-centered verbal instruction. Continued research of how basic shared biological and neurological parameters may interact to inform and improve choir singing in varied contexts seems appropriate.

Supplemental Material

Daugherty_et_al_Riser_Supplementary_Appendices_FINAL – Supplemental material for Effects of Choir Spacing and Riser Step Heights on Acoustic and Perceptual Measures of SATB Choir Sound Acquired From Four Microphone Positions in Two Performance Halls

Supplemental material, Daugherty_et_al_Riser_Supplementary_Appendices_FINAL for Effects of Choir Spacing and Riser Step Heights on Acoustic and Perceptual Measures of SATB Choir Sound Acquired From Four Microphone Positions in Two Performance Halls by James F. Daugherty, Melissa L. Grady and Robert C. Coffeen in Journal of Research in Music Education

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Supplemental Material

Appendixes A and B are available in the online version of the article at

Author Biographies

James F. Daugherty is associate professor emeritus of choral/vocal pedagogy at the University of Kansas. His research interests include life span voice pedagogy, acoustics of choir sound, conducting behaviors, and the philosophy of music and music education.

Melissa L. Grady is assistant professor of music education at the University of Kansas. Her research interests include effects of conducting gesture on choral sound, choral warm-ups, choir spacing, and adolescent female changing voice.

Robert C. Coffeen is a fellow of the Acoustical Society of America and now retired from the University of Kansas School of Architecture and Design after 25 years of teaching courses relating to architectural acoustics and building noise control. He continues to provide acoustical consulting services after 50 years in this field. His primary research interest is in room acoustics for music presentation and speech venues.

References

American National Standards Institute. (2008). Criteria for evaluating room noise. Washington, DC: Author.

Daugherty

J. F.

(1999). Spacing, formation, and choral sound: Preferences and perceptions of auditors and choristers. Journal of Research in Music Education, 47, 224–238. doi:10.2307/3345781

Daugherty

J. F.

(2003). Choir spacing and formation: Choral sound preferences in random, synergistic, and gender specific placements. International Journal of Research in Choral Singing, 1(1), 48–59.

Daugherty

J. F.

(2005, October). The effects of choir spacing and choir formation on the tuning accuracy and intonation tendencies of a mixed choir. Paper presented at the Acoustical Society of America National Conference, Minneapolis, MN.

Daugherty

J. F.

Manternach

J. N.

Brunkan

M. C.

(2012). Acoustic and perceptual measures of SATB choir performances on two types of portable riser units in three singer-spacing conditions. International Journal of Music Education, 31(3), 359–375. doi:0255761411434499

Daugherty

J. F.

Manternach

J. N.

Coffeen

R. C.

Brunkan

M. C.

(2011, June). Effects of two singer spacing condition and two riser step heights on acoustic and perceptual measures of choir sound acquired from conductor and audience position microphones. Paper presented at the Voice Foundation Symposium, Philadelphia, PA.

Ehmann

(1968). Choral directing. Philadelphia, PA: Augsburg Fortress Press.

Fletcher

Munson

W. A.

(1933). Loudness, its definition, measurement and calculation. Journal of the Acoustical Society of America, 5, 82–108.

Foot

E. L.

(1965). The effect of the intensity of auditory feedback on the loudness-intensity, quality-wave form, and intelligibility of the singer’s voice (Unpublished PhD dissertation). University of Kansas, Lawrence.

10.

Ford

J. K.

(2003). Preference for strong or weak singer’s formant resonance in choral tone quality. International Journal of Research in Choral Singing, 1(1), 29–47.

11.

Hom

K. S.

(2013). The effect of two different rooms on acoustical and perceptual measures of SATB choir sound (Master’s thesis). Retrieved from ProQuest Dissertations and Theses database. (AAT 154543)

12.

Howard

D. M.

Angus

(2006). Acoustics and psychoacoustics (3rd ed.). Amsterdam, Netherlands: Focal Press.

13.

MENC. (1943). Paysen’s collapsible chorus stands: Paysen’s choir risers [Advertisement]. Music Educators Journal, 29(6), 52.

14.

Nordenberg

Sundberg

(2004). Effect on LTAS of vocal loudness variation. Logopedics Phoniatrics Vocology, 29(4), 183–191. doi:10.1080/14015430410004689

15.

Nunes Garcia

J. M

. (1789). Domine Tu Mihi Lavas Pedes. Retrieved from http://cpdl.org/wiki/index.php/Domine_tu_mihi_lavas_pedes_(Jose_Maurricio_Nunes_Garcia)

16.

SICO Europe Ltd. (2017). Three and four step harmony choral risers [Advertising brochure]. Retrieved from: http://www.sicoinc.com/wp-content/uploads/2015/12/Harmony-Choral-Riser-InfoSheet.pdf

17.

Sugrue

(2013). Singing and space: A comparative study of choral singers and perceptions of auditorium acoustics in three venue types (Unpublished thesis). London South Bank University, London, UK.

18.

Ternström

(1989). Acoustical aspects of choir singing (PhD dissertation). Royal Institute of Technology, Stockholm, Sweden.

19.

Ternström

(1994). Hearing myself with others: Sound levels in choral performance measured with separations of one’s own voice from the rest of the choir. Journal of Voice, 8, 293–302.

20.

Ternström

(1999). Preferred self-to-other ratios in choir singing. Journal of the Acoustical Society of America, 105, 3563–3574.

21.

Tonkinson

(1994). The Lombard effect in choral singing. Journal of Voice, 8(1), 24–29.

22.

Wenger Corporation. (2017). Wenger equipment catalog. Retrieved from https://www.wengercorp.com/risers/tourmaster-choral-risers.php

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.38 MB