Handbook for Sound Engineers

Sound System Design 1269

will hear the sound coming from over their heads. That
is, the natural localization provided by a central cluster
is not provided by a distributed system. This problem is
usually minor and can be minimized by using an elec-
tronic signal delay in combination with a localizer loud-
speaker, a subject covered in Section 34.3.3.8.1.
Sometimes, a distributed system will be installed in a
very large room with a high reverberation time. A
convention hall exhibit room or high-ceiling airport
terminal is a good example. This type of system will
probably be used for paging and perhaps for back-
ground music. Thus, the localization of a central cluster
is not needed. The high reverberation and high noise
present in rooms like this, however, present problems to
the system designer. These problems can be at least
partially overcome by a dense enough layout and
careful equalization. In these areas N also plays an
important role. The distributed system has an additional
advantage over the central cluster in this case because it
effectively reduces the value of D 2 (the distance from
the loudspeaker to the farthest listener) and, thus,
improves the SNR and the direct/reverberant ratio.
Another reason for using a distributed system, even
in a room that could utilize a central cluster, is that the
distributed system allows a more flexible room layout
than the central cluster. In some multipurpose rooms,
for example, the stage location may be changed from
event to event, and some events may not use a stage at
all. The distributed system allows almost any location to
be used as the stage or primary microphone location
without the distraction that would be caused by having a
central cluster behind the heads of the audience. In addi-
tion, in large, reverberant spaces, like sports arenas,
distributed loudspeakers above unused portions of the
room can be turned off. This helps intelligibility
because it improves the direct/reverberant ratio by
lowering the amount of energy uselessly put into the
reverberant field.
The primary disadvantage of any distributed system
is its (usual) higher cost compared with a central cluster
designed for the same space (assuming that a central
cluster could work in the space).

34.3.3.1 Distributed Ceiling Loudspeaker Systems

Distributed ceiling loudspeaker systems are normally

installed in rooms with low ceilings (low compared to

the length and width dimensions of the room) where a

central cluster or distributed cluster system cannot ade-

quately cover the room. In some situations where a cen-

tral cluster would work from a design point of view, a

distributed system is chosen for its versatility. A distrib-

uted system (without delay) does not have the psycho-

acoustic localization of a central cluster. Thus,

microphone locations can be varied without worrying

about the effects on localization. Loudspeakers can be

turned off above microphone locations making multiple

(or varying) microphone locations possible while reduc-

ing feedback potential. Loudspeakers can also be turned

off in areas not in use to avoid exciting the reverberant

field in those areas. Adding variable signal delay makes

it possible to provide the psychoacoustic localization of

a central cluster from any chosen microphone location,

Fig. 34-24.

34.3.3.2 Central Cluster Plus Distributed System

In some rooms with central clusters and rear balconies,

listeners seated under the balcony are not adequately

covered by the central cluster. In this case, a distributed

Figure 34-24. Ceiling distributed loudspeaker system

with signal delay and localizer loudspeaker. Courtesy

Bosch/Electro-Voice.

40 ms area

60 ms area

80 ms area

No delay

Frontal localizer

loudspeaker

Loudspeaker location

40'

23'

23'