User Guide

Sound System Design Reference Manual
Another identifiable characteristic, particularly
of small rooms, is the presence of identifiable
resonance frequencies. Although this factor is
ignored in our statistical model, a room is actually a
complicated resonant system very much like a
musical instrument. As mentioned previously, if
individual resonances are clustered close together in
frequency the ear tends to average out peaks and
dips, and the statistical model seems valid. At lower
frequencies, where resonances may be separated by
more than a critical bandwidth, the ear identifies a
particular timbral characteristic of that room at a
specific listening location.
Since the direct sound field is independent of
the room, we might say that the “three R’s” of room
acoustics are reverberation , room resonances and
early reflections .
The distinction between early reflections and
the later reverberation is usually made at some point
between 20 and 30 milliseconds after the arrival of
the direct sound. Most people with normal hearing
find that early reflections are combined with the
direct sound by the hearing mechanism, whereas
later reflections become identified as a property of
the enclosed space. See Figure 5-14. The early
reflections, therefore, can be used by the brain as
part of the decoding process. Late reverberation,
while providing an agreeable aesthetic component
for many kinds of music, tends to mask the early
sound and interferes with speech intelligibility.
One final characteristic of sound is ignored in
all standard equations. Localization of a sound
source affects our subjective assessment of the
sound field. In the design of sound reinforcement
systems, localization is largely disregarded except for
a few general rules. It achieves critical importance,
however, in the design of multi-channel monitoring
and mixdown rooms for recording studios.
Direct and Reverberant Sound Fields
What happens to the inverse square law in a
room? As far as the direct sound is concerned (that
which reaches a listener directly from the source
without any reflections) the inverse square
relationship remains unchanged. But in an enclosed
space we now have a second component of the total
sound field. In our statistical model we assumed that
at some distance sufficiently far from the source, the
direct sound would be buried in a “soup” of random
reflections from all directions. This reverberant sound
field was assumed to be uniform throughout the
enclosed space.
Figure 5-15 illustrates how these two
components of the total sound field are related in a
typical situation. We have a sound source radiating
uniformly through a hemispherical solid angle. The
direct energy radiated by the source is represented
by the black dots. Relative energy density is
5-12
Figure 5-14. Early reflections in relation to direct sound