Page 168 - The Master Handbook Of Acoustics

143
REVERBERATION

where the random excitation happens to be the instant the switch is
opened. Even though there is a family resemblance between the four
decays, fitting a straight line to evaluate the reverberation time of each
can be affected by the beat pattern. For this reason, it is good practice to
record five decays for each octave for each microphone position of a
room. With eight octaves (63 Hz–8 kHz), five decays per octave, and
three microphone positions, this means 120 separate decays to fit
and figure for each room, which is laborious. This approach is one way
to get a good, statistically significant view of the variation with fre-
quency. A hand-held reverberation time measuring device could accom-
plish this with less work, but it would not give hard-copy detail of the
shape of each decay. There is much information in each decay, and
acoustical flaws can often be identified from aberrant decay shapes.
Four decays at 500 Hz are also shown in Fig. 7-8B for the same room
and the same microphone position. The 500-Hz octave (354–707 Hz)
embraces about 2,500 room modes. With such a high mode density, the
500-Hz octave decay is much smoother than the 63-Hz octave with only
a dozen. Even so, the irregularities for the 500-Hz decay of Fig. 7-8B
result from the same cause. Remembering that some modes die away
faster than others, the decays in Fig. 7-8 for both octaves are composites
of all modal decays included.

Writing Speed
The B&K 2305 graphic-level recorder has a widely adjustable writing
speed. A sluggish pen response is useful when fast fluctuations need
to be ironed out. When detail is desired, faster writing speeds are
required. A too slow writing speed can affect the rate of decay as it
smooths out the trace, as will be examined.
In Fig. 7-9, the same 63-Hz decay is recorded with five different pen
response speeds ranging from 200 to 1,000 mm/sec. The instrument-
limited decay for each is indicated by the solid straight lines. A writ-
ing speed of 200 mm/sec smooths the fluctuations very well. The
decay detail increases as the writing speed is increased, suggesting
that a cathode-ray oscilloscope tracing of the decay would show even
more modal interference effects during the decay.
The big question is: Does writing speed affect the decay slope from
which we read the reverberation time values? Obviously, an extremely
slow pen response would record the machine’s decay characteristic