Page 68 - samyn_ebook2

66
Speech intelligibility under the platform cover
Before describing the selected PA system, we will present a more general
analysis of the speech intelligibility levels under the platform cover. The aim is to
demonstrate that the competition design in itself already fulfils the basic required
acoustic necessities to insure adequate speech intelligibility regardless of the
chosen PA system.
Even though a large open space such as the one under the cover cannot fully meet
statistical room acoustic requirements, it is nonetheless possible to derive a first
speech intelligibility value from the reverberation time. Figure 13 indicates the
calculated average reverberation time for the space under the cover, for conditions
with and without sound absorbing roof. Combined with the graphs in figure 12.1
this generates an average speech intelligibility level of STI≅ 0.40 and STI≅0.50,
respectively. The sound absorbing cover establishes the principal conditions for
good speech quality levels for any given PA system.
Using the calculation model presented in figure 1, a detailed calculation of the
speech intelligibility levels on the platforms was made. A simple pa system design
was laid out: loudspeakers were hung at a height of 4m above platform level, at
regular intervals of 13 m. Figure 14 indicates the positions of the loudspeakers. The
indicated positions were utilized for the general analysis of the speech intelligibility
levels and the evaluation of the main architectural options; they do not match the
ultimate positions of the design which, however, does not affect conclusions.
The loudspeakers are omni directional and project an equal sound energy both to
the platforms and the canopy. This is a hypothetical and rather “poor” PA system,
since the speakers are not aimed at the platforms and fully address the entire
reverberation field. The calculated speech intelligibility level is therefore a lower
value than the levels that can be attained under the canopy.
Figures 15.1 - 15.2 indicate the dispersion of sound pressure levels on the platform,
with all omni directional speakers active, in two situations: a platform canopy with
and without sound absorbing roof. The sound source consists of 75 loudspeakers
(figure 14). The speech intelligibility levels on the platforms (STI, see figure 11) is
measured. The effect of sound absorption is clearly indicated in the illustrations.
Figure 16 summarizes the calculation results from figures 15.1 - 15.2 into a graph
indicating the cumulative distribution of the speech intelligibility levels on the
platforms. Both graphs indicate a percentage of the floor area (y-axis) where a
given speech intelligibility level (x-axis) is not exceeded.
The sound absorption makes that for more than 95% of the positions the speech
intelligibility is at least “fair” (STI > 0.45), even with “poor” sound amplification.
Under a canopy without sound absorbing roof the speech intelligibility for 75% of
platform positions is “poor” (STI < 0.45). The competition design complies with
the room acoustic al prerequisites for good speech intelligibility.