Loudness Levels and Ensemble Room Acoustics

Ellzabeth A. Cohen
Cohen Acoustical, Inc.
Los Angeles, Ca. 90004

0.0 Abstract

A previous paper (92nd AES meeting) presented musicians' preferences for the acoustical design of music practice rooms. Acoustical design criteria and priorities have evolved from that information. The requirements for large scale ensemble spaces will be discussed. The loudness levels of a typical brass & percussion ensemble were measured. These levels in turn can be used refine the sound isolation requirements of ensemble practice rooms and the room volume/height requirements.

1.0 Introduction

The purpose of the study was to establish the loudness lelvels of typical brass & percussion ensembles of an orchestra. These levels in turn can be used to reflne the sound isolation requirements for large ensemble practice rooms and the room volume/heigh t requirements.

A common problem for large rehearsal rooms for instrumental groups and for choruses is that they often suffer from inadequate volume(Gerdes,1989). Inadequate volume can generate:

1. Reverberation times that are too short to meet a primary musician requirement of simulating performance hall acoustics as closely as possible.
2. The need for placing absorption close to performers which stimulates forced tone production.(Gerdes,1989)
3. Loudness Levels that create discomfort and potentiai occupational hazards for rehearsing musicians.

• 1.1 Recent Literature Review Royster, Royster, & Killion summarized our knowledge about sound levels and sound exposure(Royster, Royster, & Killion,l991):
  • Both the instantaneous sound levels measured for a musician and the resulting equivalent continuous sound leVel(Leq)for the measurement period depend on the instrument played and the music being performed, as well as the position of the musician on the stage in relation to other players. The microphone location is also important, especially in instances where one ear receives a higher sound exposure from the player's own instrument(such as the left ear for a violinist) or from other instruments. Lit erature reports of sound levels and exposure o'f musicians have not aiways described their measurement conditions adequately to permit comparison of results among authors.

  Royster et. ai describe combined rehearsal & performance Leqs for ,,J musician's ranging from 78.5 dB(A) to 88.4 dB(A). They review Camp &~ Horstman's(1987) paper showing personal dosimetry results on musicians that gave average Losha values of 82. 4 dB(A) for bass players to 96.0 dB(A) for horn players. They also describe Schake's (1987) dosimetry study showing average sound levels for brass and wind, players ranging form 87-96 dB(A), with the typical day calculated a~s ah 8-h Leq of 87.7 dB(A).

  In summary, they have found that, " at least some orchestra members have typical daily 8-h Leq vaiues exceeding 85 dB(A)...therefore it would be expected that some musicians may face risk of hearing damage from on-the-job sound."

  For the purpose of large ensemble room design refinement, we need to determine the levels found in the ensemble rehearsal condition.

  For confirmation of sound isolation criteria, we are interested in the maximum sound level generated by the ensembles and perceived loudness. For musician comfort, both the maximum and the average levels are relevant to the determination of rehearsal/prac tice room comfort and safety.

2.0 Rehearsal "Room 2 & Room 3"

Initially, our measurements were scheduled for "Room 2", however we were informed about 15 minutes before the scheduled rehearsal that the ensemble and the conductor had managed to change the rehearsal to "Room 3 "because they found the "Room 2", "too sma ll and that it would be too loud in the room (for this slze of an ensiemble)." Figure 1 shows the instrumental seating charts to illustrate ensemble size.

The decision to switch to "Room 3" is not surprising as Room #2 is a smaller and more reverberant room, and therefore is louder. We include partial descriptions of "Rooms Two & Three" (room dimensions, background noise, & reverberation time) for compariso n purposes in Figure 2, Figure 3, Figure 4,Figure 5.

Establishment of the background noise level allows for calculation of masking levels which are critical to determining the perceived dynamic range of the room and the potential intrusion of noise from adjacent spaces.

3.0 Measurement Procedures

Sound Pressure Level Measurement Method Measurements were made at aDproximated edge of diffuse fleld We confirmed that peak levels from the brass instruments were relatively constant throughout the room except close to the source. Therefore our microphone locations represent good average Sound Pressure Level readings.

The Average A-weighted levels (slow scale) were recorded directly onto the scores. The average was read off a Bruel & Kjaer 2203 with 4145 1 inch microphone. The peaks were measured with a Genrad 1982 set on the 120 dB scale, both A-weighted and peak were noted on the scores.

Additional measurements were made with a Bruel &Kjaer 2230,110 dB full scale reflection with 1/2" microphone and stored on a Panasonic SV3700 DAT on Channel B. A Bruel &Kjaer 4145 1/2 inch capsule on a Genrad preamp was connected directly to Channel A of the Panasonic SV3700 recorder.

The Bruel &K.jaer 2230 microphone was located 1.65m off the floor and 2.5m to the west of the conductor's podium. The Bruel &K.jaer 4145 microphone was located 1.5m off the floor and 3.20m to the west of the 2230.

Three pieces of music were performed: Deserts, Equatorial, and lonization durin.. which sound levels were recorded.

4.0 Loudness Levels of Orchestral Instruments

The previous section dealt with the measuring the sound pressure levels generated by the ensemble. The subsequent sections attempt to relate the physical variables to their sub.jective loudness. In addition to changes in sound pressure level, variations o f sound spectrum contribute to changes of loudness(Zwicker 1957). Zwicker demonstrated that an increase in frequency bandwidth results in growth of loudness(Zwicker 1957). Changes of sound spectrum related to the playing level of musical instruments resul ts in an increase of the dynamic range of the instrument( Miskiewicz & Rakowski,1992).

Figure 6 shows calculated loudness levels based on Stevens magnitude estimation techniques as described in ANSI S 3.4. However there is widespread criticism of this loudness scaling technique(Moore,l991).

Miskiewicz has shown that Zwicker's method yields higher values of loudness level than both Steven's Mark 6 & 7 procedures and that the differences between Zwicker's and Steven's phons are more pronounced in low pitch registers than high ones. Furthermor e, Miskiewicz's research revealed that, "in the brass lnstruments and the clarlnet, the dynamlc range of loudness level mar}edly exceeds that of sound pressure level."

Therefore, we calculated the Zwicker loudness levels, based upon the 1/3 octave band levels (Leq and Lmax), for two four minute periods during the rehearsal of 17 May 1992 and compare them with the levels generated by the Stevens method.

The differences between calculating loudness levels for instrumental ensembles between the Zwicker and Stevens method is consistent.with that found by Miskiewicz for individual orchestral instruments: The difference in calculated levels ranges between 3 a nd 7 Phons, with the Zwicker method predicting perceptually signiflcant increased loudness.

The standard method of determining necessary sound isolation is comparing the noise criteria of critical ad.jacencies (source/receiver) with the Lmax levels(source/receiver) using Sound Pressure Level data not Loudness Levels. However, the above measured loudness levels appear perceptually signiflcant. This suggests that loudness levels can serve as a guide when a consu'ltant is involved in value engineering. In thls case, it would indicate that the use of conservative sound pressure levels in estimating the necessary sound isolation between ensemble rooms and other adjacent critical listening spaces is advisable.. Furthermore, the perception of loudness generated by high SPLs (especially at low frequencies) suggests possiblv adding a few dB as a safety f actor in sound isolation calculations.

There is no loudness meter standard calculation process to date and the safety factor technique is only recommended at low frequencies due to both the relative ease of mitigating high frequency noise and the weakness of the Zwicker Method at high frequenc ies (Hellman,1987: Quinlan,1992).

5.0 Concluslons

1. The measurements indicate the importance of adequate volume for ensemble rooms.
2. Rooms 2 & 3 are considered too loud by the musicians and too small for large, loud ensembles, therefore, we expect laree ensembles will not use the smaller rooms V-Z shown in Table 2. Musicians commented that they contemplated wearing earplugs during the rehearsals. Issues of Senza Sordino were distributed by an orchestra member at the rehearsal on earplug types and hearing loss among orchestra musicians.
3. Rooms that are dedgnated for brass and percussion ensemble rehearsal should have dimensions on the order of Room 3 or greater because the Loudness Levels generated by such rehearsals in small rooms can create discomfort and potential occupation al hazards for rehearsing musicians .
4. If space is limited and large amounts of absorption are necessary to assure proper dlstributlon of room modes and reduce detrlmental loudness then ensemble practice rooms should be made cable/conduit ready to allow for future adustabillty of acoustic s by electronic means(such as the LARES system). Electroacoustics can produce the needed baiance between reflections necessary for ensemble hearing conditions and loudness levels. This approach is not expected to satisfy aTI musicians.
5. The Sound Pressure Level data obtained in this study should be used to verify that typical maximum levels generated by brass and percussion ensembles will not be audible, to an intrusive degree, into any critical adjacent spaces such as practice rooms.

6.0 Acknowledgements

The author wishes to thank George Augspurger, Pierre Boulez, Robert Greenhood, Ara Guzelemian, David Kahn, Rachel Murray, & The New Music Ensemble.

7.0 References

Geerdes, Harold P. "Music Facilities: Building, Equipping, and Renovating" Music Educator's Conference Publishers, 1989. DD.66

Miskiewicz & Rakowski, Loudness Levels of Orchestral Instruments JASA, 1992.

Moore, BCJ, " An Introduction to the Psychology of Hearing" Academic

Nobile,M.A., Bienvenue,G.R., & Conahan,S.G. ,'The Critical Band Spectrum", Internolse, 1992,pp. 111 1-15.

Quinlan,D.,''Sublective Experiments on ISO 532-B,"Method for calculating Loudness Level'."Internoise, 1992. pp. 1103-07

Royster, Rovster, & Killion, JASA, June 1991.