Handbook for Sound Engineers

1274 Chapter 34

speaker. In the high-ceiling areas this approach covers
each listener with additional loudspeakers. While the
design is simplified and the layout more symmetric, this
approach results in more loudspeakers, which will
increase the cost.

34.3.3.8 Designing the Distributed Cluster System

A distributed cluster system consists of two or more
clusters separated in space. Two clusters, for example,

might be used to cover a long, narrow religious facility

where a single cluster could not provide acceptable

intelligibility in the rear seats. A second ring of clusters

may also be installed in a large, fan-shaped room where

the main system is an exploded cluster design. In this

case, the second ring of clusters is installed on radii

from the main clusters.

An examination of the Alcons equation (Eq. 34-18)

shows that either increasing Q or decreasing D 2 will

improve intelligibility. In the case where a loudspeaker

with high enough Q will not provide wide enough

coverage or where a loudspeaker with high enough Q is

simply not available, adding one or more additional

clusters that are closer to the far listeners may be the

answer since this decreases D 2.

Design the first cluster (or exploded clusters) to

cover the seating areas out to (and slightly beyond) the

position of the second cluster (or second ring of clus-

ters). All other design criteria for the first cluster remain

the same as if it were the only cluster. The value of N

(the Dc modifier) for either cluster must include the

effects of both clusters, however. Design the second

cluster to cover the remaining seating area to the

farthest listener. In many systems of this type, the

second cluster can have a reduced low-frequency

section for frequencies below about 200 Hz. This is

because the frequencies below 200 Hz do not contribute

to intelligibility and because the reverberant field in

most rooms requiring a second cluster will carry the low

frequencies to the farthest listener with no need for rein-

forcement from the second cluster.

In calculating Alcons (or EPR or PA G and NAG) for

either cluster, the value of N must take into account the

loudspeakers in both clusters, although the value of D 2 ,

of course, will be shorter than it would have been for a

single cluster in the same room.

34.3.3.8.1 Signal Delay in a Distributed Cluster System

The second cluster in the previous example must

receive a signal that is electronically delayed from the

signal sent to the first cluster. See Section 34.3.3.8.1 for

an example of calculating this delay.

34.3.3.8.2 Distributed Clusters with No Delay

In a circular stadium or on long, narrow bleachers such

as at a race track, a system of distributed clusters may

provide the best coverage and may not require elec-

tronic delay because the sound reaching each listener is

primarily from one cluster and any nearby cluster is

Figure 34-33. Basic concepts of distributed placement in
rooms with sloped floor and/or ceiling. Courtesy
Bosch/Electro-Voice.

Power levels

relative to S 1

for equal direct SPL

S 1

S 2.

S 3

S 4

S 5

S 6

S 7

S 8

S 9

S 1 S 2 S 3 S 4 S 5 S 6 S 7

A 1 A 2

A (^3) A
(^4) A 5
A 6
A 7
S (^1) S 2 S
(^3) S 4 S 5
S (^7) S
8
S 6
S 9
0.0 dB
2.50
3.33
4.78
6.44
8.07
9.64
11.21
12.80 dB
A. Longitudinal section of room with
loudspeaker row positions.
B. Plan view of room with loudspeaker positions.
S 9
S 8
A 9 A 8
S 1 S 2 S^3
S 4 S^5
S 6
S 7
S 8
S 9