Designing for Speech Intelligibility 1397The %Alcons equations work well with single point
or center cluster systems or even split clusters, however,
with distributed systems (especially high-density ceiling
systems for example) determining the (n+ 1) factor
becomes extremely difficult, as it is difficult to appor-
tion what percentage of the radiation from adjacent or
semiadjacent speakers is actually contributing to the
direct field and early fields and which is contributing to
the reverberant.
To a certain extent this is made easier in the more
complex or long form version as a straight apportion-
ment factor can be applied, though some considerable
skill in doing this is required. Because the %Alcons
equations do not effectively account for the early or late
reflected energy, their accuracy needs to be treated with
some caution. Furthermore, the method and equations
are based on statistical acoustics, which at low reverber-
ation times (e.g., <1.5 s) in itself becomes less accurate.
36.7.2 Intelligibility and Reverberation Time
Although, as we have seen, there is a lot more to intelli-
gibility than reverberation alone, knowing the reverber-
ation time of a space is a good starting point for a
system design and immediately allows the potential
difficulty of the task to be quantified. Some general
rules of thumb can be applied in this context as seen in
Table 36-2.
When designing or setting up systems for use in
reverberant and reflective environments, the main rule
to follow is, “Aim the loudspeakers at the listeners and
keep as much sound as possible off the walls and ceil-
ing.” This automatically partially maximizes the direct-
to-reverberant ratio, though in practice it may not be
quite so simple. The introduction of active and phased
line arrays has had a huge impact on the intelligibility
that now can be achieved in reverberant and highly
reverberant spaces. Arrays of up to 5 m (~16 ft) are
readily available and can produce remarkable intelligi-
bility at distances of over 20–30 m even in 10 s plus
reverberation time environments. The use of music line
arrays has also led to a significant improvement in
music/vocal clarity in arenas and concert halls. Whereas
the intelligibility form a point or low Q source effec-
tively reduces as square of the distance, this is not the
case for a well-designed/installed line array. An exam-
ple of this is Fig. 36-23 where it can readily be seen that
the intelligibility (as measured using the Speech Trans-
mission Index—STI) remains virtually constant over a
distance of 30 m in a highly reverberant cathedral
(RT = 4 s).36.8 Some Further Effects of Echoes and Late
ReflectionsAs already noted, speech signals arriving within 35 ms of
the direct sound generally integrate with the direct sound
and aid intelligibility. In most sound system applicationsFigure 36-22. Effect of direct-to-reverberant ratio as a func-
tion of RT 60 on %Alcons.
Probable articulation loss of consonants versus
reverberation time and direct-to-reverberant
sound ratio.Ratio of direct to reverberant sound pressure–dBArticulation loss of consonants–%Time–sTable 36-1. Effect of Reverberation Time
RT 60 Results< 1 s Excellent intelligibility should be obtained.
1.0–1.2 s Excellent to good intelligibility should be achieved.
1.2–1.5 s Good intelligibility should be achieved though loud-
speaker type and location become important.
>1.5 s Careful design required (loudspeaker selection and
spacing).
1.7 s Limit for good intelligibility in large spaces (distributed
systems)—e.g., shopping malls, airport terminals.
>1.7 s Directional loudspeaker required (churches, multipur-
pose auditoriums, and highly reflective spaces).
>2 s Very careful design required. High-quality directional
loudspeaker required. Intelligibility may have limita-
tions (Concert halls, churches, treated sports halls/
arenas.)
>2.5 s Intelligibility will have limitations. Highly directional
loudspeaker required. Large (stone built) churches,
sports halls, arenas, atriums, enclosed railway stations,
and transportation terminals.
>4 s Very large churches, cathedrals, mosques, large and
untreated atria, aircraft hangars, untreated enclosed ice
sports arenas/stadiums. Highly directional loudspeakers
required and located as close to the listener as possible.