Public Address Systems

In every locality, where music or speech is to be transmitted to a larger public, a more or less extensive sound system will be employed. While for many events a small and basic set-up is absolutely sufficient, spacious venues call for the use of complex systems, which need to be accurately positioned and adjusted to each other.

For planning such systems, simulation tools like EASE are made use of, which allow the prediction of parameters like sound pressure level and speech intelligibility in the listener areas. Once the equipment has been installed on site, the respective subsystems need to be adjusted to the room and merged to a coherent whole. This requires extensive measurements. In case the public address system acts as an emergency warning system, e.g. in stadiums or road tunnels, the speech intelligibility for announcements has to be verified according to the relevant regulations.

Compulsory for successful commissioning measurements is, that the adjustment of the subsystems in free field conditions is accurate. The generation or correction of controller setups after the equipment has been installed, is virtually impossible.

Generally the calibration starts by checking the operation of all the components (alignment, curving, amp gains, limiter, routing, polarity, backup systems). This is where most of the error sources hide, which have to be eliminated prior to the actual measurements.

At 20-50 measuring positions per loudspeaker or array the frequency response on the main axis is recorded, in large venues via a wireless system. These measurements contain any reflexion that occurs in the room and their energy average results in a representative characteristic. From this curve the EQ setting for the respective loudspeaker, also called Hauskurve, is deduced.

Afterwards the levels of all the individual systems or clusters are adjusted to each other.

Eventually from delay measurements at dedicated positions the delays of the subsystems are tuned, so that the localization of the main system (stage) remains unaffected.

The procedure is concluded with a listening test to verify the settings made.

Additionally further measurements can be necessary:

• Determination of loop gain and adjustment of filter settings for feedback elimination
• Measurements of STI, signal level, noise level

The demands for emergency warning systems can be found in the following standards:

• DIN EN 60849 (IEC 60849 and VDE 0828): Sound Systems for Emergency Purposes
• IEC 60268-16: Sound System Equipment, Objective Rating of Speech Intelligibility by Speech Transmission Index
• DIN VDE 0833-4 Appendix G: Measuring Methods for the Determination of the Speech Transmission Index STI

The intention is to capture the speech intelligibility in a single parameter (STI) by means of objective measurement.

Speech intelligibility can be affected by outside influences (noise), room acoustics (reverb, echoes), as well as sound system deficiencies (shortened frequency response, too high or low level, distortion).

Speech intelligibility is measured either by feeding a dedicated test signal into the system while evaluating the premises with a mobile NTI Acoustiyzer, or by analyzing the impulse response, recorded with Monkey Forest, according to the Schroeder method.

A sound system and the surrounding room are two inseparable components, which define the acoustic properties of a venue. This is why planning large PA systems is only possible in consideration of the room acoustics. Simulation software like EASE is a powerful tool, which allows a good prediction of acoustic properties of the system before it is installed.

Before planning the system itself, one should establish the technical criteria it has to meet. Depending on the application this can be the sound pressure level, which needs to be attained in a defined frequency interval, an especially linear transfer characteristic for high class music reproduction, a prescribed speech intelligibility for emergency warning or the immission control in the outside areas. Furthermore aspects like the protection of historic monuments, vandalism resistance or a specific optical design can become relevant.

Prerequisite for the simulation are high resolution balloon data of the loudspeakers including phase distribution in at least a 5° raster.

The simulation procedure starts by constructing the room, which the equipment is to be installed in. Here all the acoustically relevant elements have to be included, at the same time irrelevant details remain out of consideration, in order to keep the computing time within a limit. As precisely as possible, all the elements are now given parameters for absorption and scattering, which depend on the material used and its structure.

With this model already, the reverberation time according to Eyring or Sabine can be calculated. Comparing the result to reverberation measurements of the room allows readjusting in the model.

Now the loudspeakers are inserted into the model. In case line arrays are going to be used, tools for determining the basic configuration are useful. The objective is a constant coverage of all the audience areas, while avoiding highly reflective surfaces.

Next, the distribution of direct SPL is simulated. Even for a high resolution the computing time for this is rather short. The results in the audience areas as well as outside can be used to adjust the arrangement of the loudspeakers.

Afterwards the impulse responses at selected positions are simulated according to the mirror image source method, using the Aura module in EASE. From here, more parameters can be calculated, like the precise reverberation time with the model sources, total SPL and speech intelligibility.

Ultimately the auralization is generated by convolving arbitrary signals with the impulse responses, to allow an impression of the system’s sound in the room.

Beside traditional loudspeaker clusters, powerful line arrays have made their way on stage, and make possible an even coverage of wide listener areas. Recently a directional bass reinforcement, implemented with cardioid configurations or other bass arrays, is getting increasingly popular. The effort for planning these loudspeaker arrangements is higher than for conventional arrays, and even intensified by delay lines and specific solutions for boundary areas or side rooms.

Sound systems for speech reinforcement have a wide field of application ranging from conventions, press conferences and lecture halls to general meetings or party conferences. Common to all these applications is the need for an excellent speech intelligibility, even levels in the audience as well as a high resistance to feedback.

Convolution with the simulated impulse response allows the auralization of any signal:

Audio Sample 1: Auralization Stadium with Line Arrays without Audience

Audio Sample 2: Auralization Stadium with Line Arrays with Audience

Audio Sample 3: Auralization Stadium with Horn Clusters without Audience

Audio Sample 4: Auralization Stadium with Horn Clusters with Audience