Handbook for Sound Engineers

Test and Measurement 1627

orderly approach to achieving meaningful results that
correlate with the perception of live listeners. The
measurer is free to choose alternate cursor placements,
but great care must be taken to be consistent. Also,
alternate cursor placements make it difficult if not
impossible to compare your results with those obtained
by other measurers. In the default %Alcons placement,
the early energy (direct sound field) includes the first
major sound arrival and any energy arrivals within the
next 7–10 ms. This forms a tight time span for the direct
sound. Energy beyond this span is considered late
energy and an impairment to communication. As one
might guess, a later cursor placement yields better intel-
ligibility scores, since more of the room response is
being considered beneficial to intelligibility. As such,
the default placement yields a worst-case scenario. The
default placement considers the effects of the
early-decay time (EDT) rather than the classical T 30
since short EDTs can yield good intelligibility, even in
rooms with a long T 30. Again, the measurer is free to
select an alternative cursor placement for determining
the decay time used in the calculation, with the same
caveats as placing the early-to-late dividing cursor. The
%Alcons score is displayed instantly upon cursor place-
ment and updates as the cursors are moved.

46.3.6.2 Speech Transmission Index—(STI)

The STI can be calculated from the measured impulse
response with a routine outlined by Schroeder and
detailed by Becker in the reference. The STI is probably
the most widely used contemporary measure of intelligi-
bility. It is supported by virtually all measurement plat-
forms, and some handheld analyzers are available for
quick checks. In short, it is a number ranging from 0 to
1, with fair intelligibility centered at 0.5 on the scale.
For more details on the Speech Transmission Index, see
the chapter on speech intelligibility in this text.

46.3.7 Polarity

Good sound system installation practice dictates main-
taining proper signal polarity from system input to sys-
tem output. An audio signal waveform always swings
above and below some reference point. In acoustics, this
reference point is the ambient atmospheric pressure. In
an electronic device, the reference is the 0 VA reference
of the power supply (often called signal ground) in
push-pull circuits or a fixed dc offset in class A circuits.
Let’s look at the acoustic situation first. An increase in
the air pressure caused by a sound wave will produce an
inward deflection of the diaphragm of a pressure micro-

phone (the most common type) regardless of the micro-

phone’s orientation toward the source. This inward

deflection should cause a positive-going voltage swing

at the output of the microphone on pin 2 relative to pin

3, as well as at the output of each piece of equipment

that the signal passes through. Ultimately the electrical

signal will be applied to a loudspeaker, which should

deflect outward (toward an axial listener) on the posi-

tive-going signal, producing an increase in the ambient

atmospheric pressure. Think of the microphone dia-

phragm and loudspeaker diaphragm moving in tandem

and you will have the picture. Since most sound rein-

forcement equipment uses bipolar power supplies

(allowing the audio signal to swing positive and nega-

tive about a zero reference point), it is possible for sig-

nals to become inverted in polarity (flipped over). This

causes a device to output a negative-going voltage when

it is fed a positive-going voltage. If the loudspeaker is

reverse-polarity from the microphone, an increase in

sound pressure at the microphone (compression) will

cause a decrease in pressure in front of the loudspeaker

(rarefaction). Under some conditions, this can be

extremely audible and destructive to sound quality. In

other scenarios it can be irrelevant, but it is always good

to check.

System installers should always check for proper

polarity when installing the sound system. There are a

number of methods, some simple and some complex.

Let’s deal with them in order of complexity, starting

with the simplest and least-costly method.

46.3.7.1 The Battery Test

Low-frequency loudspeakers can be tested using a stan-

dard 9 V battery. The battery has a positive and negative

terminal, and the spacing between the terminals is just

about right to fit across the terminals of most woofers.

The loudspeaker cone will move outward when the bat-

tery is placed across the loudspeaker terminals with the

battery positive connected to the loudspeaker positive.

While this is one of the most accurate methods for test-

ing polarity, it doesn’t work for most electronic devices

or high-frequency drivers. Even so, it’s probably the

least-costly and most accurate way to test a woofer.

46.3.7.2 Polarity Testers

There are a number of commercially available polarity

test sets in the audio marketplace. The set includes a

sending device that outputs a test pulse, Fig. 46-28,