Bringing Architectural Acoustical Standards to a One-Off Event

As acoustical consultants and sound system designers, we at SIA Acoustics we believe that the design of special events requires the same standards of skills, tools and attention to detail essential for architectural projects. Special events pose many of the same problems as developing acoustical and electro-acoustic designs for performance spaces, but happen on a much shorter time frame.
As part of a special event, SIA Acoustics was asked to temporarily improve the acoustics of The New York State Theatre Promenade at Lincoln Center. The program was a celebration of mentors and their protgs in the fields of dance, theater, film, music, literature and visual arts. It was recorded for future television broadcast or distribution. The event featured several short films and rich soundtracks illustrating the work of each artist. Following the presentation mentors and protgs shared remarks from three podiums located on three sides of the audience seating area.
The State Theatre Promenade is known to be acoustically challenging. To achieve the high intelligibility required by the client, significant effort, both acoustically and in terms of sound system design and operation was required. As a result, both the sound system design and acoustical treatments were developed in tandem.
To be effective in both architectural and special event designs, we find it essential to become involved early in the design process. In this case, we were able to participate in the development of sound critical items including the podium designs, location of loudspeakers and schemes to introduce visually acceptable acoustical elements. Further, we became involved with decisions that impact the over all sound quality, such as: the format the sound tracks were mixed to during post production (in preparation for presentation at the show) and the selection of vendors.
As with the design of permanent venues in our consultant work, our work on this special event focused on acoustical and sound system issues, but also included concerns such as room layouts, mechanical noise, and acoustic isolation. Working closely with the design team and event producers, acoustical details were incorporated into the event design and were often invisible to audience members.
To help determine our goals and define our challenges, we began by studying the acoustics of the space using SIA-SMAART. The extent and configuration of acoustical treatments and the levels of speech intelligibility that could be achieved were determined by studying impulse responses taken with SIA-SMAART software. Impulse response measurements reveal acoustical characteristics of the room including both the frequency dependent rates of the reverberant decay and the presence of discrete reflections of sound. By comparing the arrival time of direct sound, the arrival time of reflections and the promenades geometry, we were able to determine which surfaces in the facility are reflecting the most sound energy. We analyze these reflections to determine which reflections would degrade perceived sound quality and where to position treatments to either absorb or diffuse (i.e. scatter) these reflections.
For the Promenade of the New Your State Theater, our measurements established that the level of broadband reverberation was in excess of 2.4 seconds and tonally unbalanced. Tonal unbalance is an undesirable ratio of low frequency to high frequency reverberation decay rates. Our concern was that the music would loose much of its impact and definition without sufficient treatment and that the intelligibility of a presenters amplified voice would be degraded by the "boomy" and "echo-y" characteristics of the room. Typically we like to see low frequency reverberation that is about 1.35 times the reverberation time of the 1kHz 2kHz Hz frequency range.
Based on the measurement data, we developed a series of acoustical treatments designed to reduce the promenades overall amount of reverberation and to help control discrete reflections of sound. Acoustical panels were widely used. More than 550 2" thick 3 pound density Fiberglas panels helped to control off axis energy from each loudspeaker array and effectively absorb both reverberant decay and discrete reflections in the critical speech intelligibility range. The 550 panels added approx. 18,929 sabin units of absorption to the promenade for the gala.
The amount of effort and labor required to install the acoustical treatments was a major consideration. With acoustical panels positioned throughout the promenade and balconies, we planed to keep the installation simple and require only a few moments per panel. The balcony panels did not require attachment or suspension from any surfaces. Instead, the panels were hidden out of direct visual line of sight, behind the balcony railing.
The panels did not interfere with the windows of the Promenade, and major visual design element for the show - a sheer fabric allowing attendees to see the shimmering lights on Lincoln Centers Josie Robertson plaza. While heavy velour drapes absorb significant sound energy very efficiently, the large sheer fabric drapes that were used allowed sound energy to easily pass through. This allowed loudspeakers to be effectively hidden behind the acoustically transparent material and acoustical panels to be exposed to sound energy from the room.
To ensure the materials selected by the events designers would perform acoustically (either as absorbers or transparently) we conducted measurements to understand the acoustical properties of key fabrics planed for use at the event, such as those intended for use in front of loudspeakers. Samples of the proposed materials were tested in order to understand the acoustical impact and modify our design accordingly. Using SMAART, it is possible to determine their acoustical transparency.
One often overlooked reason for degraded intelligibility is excessive ambient noise. To address this we recommended that all sources of noise, not critical to the event, be either removed or shut down. An example of this is refrigeration units in the catering service area that were part of the facility but not needed for the event. Shutting them down eliminated a significant source of rumble.
About half, the acoustical acoustical panels were suspended with "S" hooks behind catering screens or pipes and drapes located at the promenade level. These panels added significant absorption to areas where light weight fabrics are typically used. Light weight fabrics would not have absorbed any significant sound energy and would contributed very little to reducing noise generated from within the catering areas.
Another common acoustical problem is the reflection from the podium surfaces used by presenter. To address this, the podium countertops were covered with sound absorbing panels. Adding acoustical treatment to top surfaces, an area typically invisible to the audience, reduced the reflections in close proximity to the microphone, improving sound quality. The podium construction was reviewed and recommendations were made to increase the stiffness of the floor areas to prevent them from resonating like a drum and reducing the sound quality of the speeches. Each podium was on a riser deck covered in carpet to further reduce the possibility of structure borne noise from entering the microphones.
One common sound control element we did not recommend is the use of carpet for the entire promenade space. Carpet would have been somewhat acoustically beneficial, however we feel it can be replaced by more acoustically effective, strategically placed acoustical treatments. There are several reasons not to use carpet: high cost of installation, long load in time, visual impact on the space, and typically limited acoustical performance. Also it is not uncommon for carpet to absorb too much high frequency and not enough mid-low or low frequency energy leaving the room tonally unbalanced.
The sound systems were designed to be highly-directional and configured into 16 zones. We were able to control where sound energy was radiated and allow special directional sound cues. Not all loudspeakers were used simultaneously. Audio signals were distributed to over 75 Meyer self-powered (M1D and M1D Sub) loudspeakers via an Aviom audio network connected directly to a Yamaha M7CL digital mixing console console on CAT 5e cable. Audio inputs to the Yamaha M7CL console from videotape decks were connected via a second Aviom digital snake. Using self powered speakers, and distributing signals over cat 5 cable, simplified the installation process and reduced the amount of set up time required to just 5 hours.
The venue had a limited amount of power available and the events design called for extensive video and lighting systems. As a result we operated all the loudspeakers at 208 v to reduce the overall current draw. All sensitive electronics were supplied from uninterruptible power supplies for added safety.
Loudspeaker selection and positioning is critical to achieving intelligibility. The primary loudspeaker arrays were distributed along the promenades perimeter, on the first balcony. To suspend the arrays, "Goal posts" were constructed from schedule 40 pipe and heavy bases. No existing attachment points were available and ground stacking was not possible. Suspending the arrays was required to achieve the optimum overall angle of tilt for the array while allowing clearance over a hand rail that could not be removed. A small 30" acoustical panel was attached to the rear of each line array to help absorb off axis mid and high frequency energy. Because the large sheer fabric was lit from the front, the loudspeakers were hidden from view. The use of a distributed sound system will allowed directional perspective to be maintained, so sound appeared to come from the direction the presenter was speaking from.
Due to the wide audience seating area, a low frequency array of 20 Meyer M1D ultra compact subwoofers was used to achieve consistent coverage for listeners. M1D subs were selected because they are so small; many units were needed to achieve the desire directivity but high SPL was not required, making larger products unnecessary. Positioned on the floor near the northern glass wall, the subwoofers were unobtrusive and remained hidden behind the scrim materials. Each unit was spaced by 7 6 to achieve the desired coverage with a significant reduction in level to the sides of the seating area. The subwoofers were only received program from the soundtracks, not from live mics.
With the arrays on the balconies and orientated to the rear of each podium, gain before feedback was addressed by creating intentional holes in the coverage area and taking advantage of the podium mics pick up pattern and off axis attenuation. We used Schoeps MK41 supercardiod capsules with 420mm goosenecks for each presenter. We diligently took measurements during rehearsals to allow the correct height of each microphone for a specific presenter to be repeated. A TC Electronic system 6000 processor was used operating 6 channels of Dolby cat 430 emulation, one for each mic. This program includes multi-band compression and allowed us to subtly provide a very clean podium feed. Breath pops were removed, louder voices were controlled and ambiance was reduced.
The areas not covered by the main arrays in the vicinity of each podium were served by a front fill system consisting of 3 Meyer MM-4s at each podium; a total of 9 total MM-4 units were used. The MM-4s were externally powered, and their power amps positioned on the first balcony behind each podium. Each MM-4 unit operated on a separate amplifier channel allowing maximum control over level and delay. No stage monitors for the presenters were required.
Equalization and delay for each loudspeaker zone was achieved via 5 BSS Soundwebs distributed around the balcony level, near the loudspeaker arrays. We designed a standard set of control panels for use on live events. This greatly simplifies the management of a large number of zones during the sound system optimization process by incorporating a SMAART routing scheme into the Soundweb design. It is possible to move quickly from zone to zone during the sound system optimization process quickly changing reference and measurements signals needed from each zone under test. Without Soundweb, external wiring, "Y" cables and patching would have been required and there would have been a significant chance of making a mistake. With Soundweb, we were able to configure all the necessary routing and controls into to a quick, reliable user interface that was proven to work even before we arrive on site. Were looking forward to the day we can take care of this on the output side of a console using its own internal signal processing.
Scharff Weisberg was selected to provide the sound equipment and their facility was used for preparation. All loudspeakers, equipment, equipment racks, subassemblies and cables were tested before arriving on site.
APPENDIX #1: SELECTED MEASUREMENTS, RESULTS AND ANALYSIS
Figure #1: An impulse response of the New York State Theatre promenade This measurement result describes the promenades acoustical characteristics including details about the reverberant decay and discrete reflections of sound. Time is on the X-axis, decibels are on the Y-axis. By comparing the arrival time of direct sound, the arrival time of reflections and the promenade geometry, we can determine which surfaces in the facility are reflecting the most sound energy. We analyze these reflections is to determine which reflections degrade sound quality and where to position treatments to either absorb or diffuse (i.e. scatter) these reflections.
The inset scale shows both the speech transmission index (STI) and clarity factor (C50) results for the promenade. The speech transmission index (STI) is measure of intelligibility whose value varies from 0 (completely unintelligible) to 1 (perfect intelligibility). The clarity factor, as shown here, is the ratio of sound arriving before and after 50 milliseconds. Both measurements are fair indicators of subjective intelligibility. The promenade scores fair to poorly on the indexes and will benefit form acoustical treatment.
Figure #2: A spectrograph of the data shown in figure # 1. In a spectrograph, the measured impulse response is displayed with time in the x-axis and frequency on the y-axis and the magnitude of sound energy is shown by color. Reds are the highest values, followed by shades of yellow, green and blue. In this plot you can see that once the direct sound is generated, the decay includes several strong reflections. Energy below 1000Hz decays more slowly as shown by the amount of yellows shown overtime on the graph. The promenades tonally unbalanced characteristic indicates this room requires added low frequency frequencies absorption.
Figure #3: A rainbow plot of the data shown in figures #1 and #2. This plot shows the decay rate of sound in the promenade for the 250, 500 and 4000 Hz octave bands. Unlike an impulse response (figure #1), the rainbow plot shows the sound energy in specific octave wide bands. We can see that the 250, 500 and 4000 Hz octave bands all have different decay rates.