C INEXPO WHITEPAPER
Three-way Loudspeakers
for the Digital Age
The Laws of Physics / The Art of Listening
20 100 1000 10000 20000
dB SPL
Frequency (Hz)
Traditional 2-way loudspeaker design
vocal range
woofer compression driver
20 100 1000 10000 20000
dB SPL
Frequency (Hz)
EAW 3-way loudspeaker design
vocal range
woofer
compression drivermidrange cone
INTRODUCTION
The primary requirement of left/center/right main channel
cinema loudspeakers is that they faithfully reproduce the
entire audio spectrum from 20 Hz to 20 kHz. While crisp,
natural reproduction of the vocal range (300 Hz - 2 kHz) has
always been critical to good cinema sound, very high and
very low frequencies (abov e 4 kHz and below 200 Hz)
demand particular attention in the age of digital cinema
audio. The advent of digital technology has brought sound
effects to the very extremes of hearing, filling these ranges
with sonic material which today defines the digital age
cinema experience.
Traditional two-wa y cinema loudspeaker design (a large
cone woofer and a horn-loaded high frequency compression
driver) suffers from an unfortunate 500 Hz crossov er placement that robs dialog of much of its warmth. By requiring
the high frequency compression driver to cover frequencies
as low as 500 Hz, two-w ay design introduces substantial,
audible distortion in the vocal region, leaving voices sounding harsh and hollow .
Similarly , a 500 Hz crossov er compromises a loudspeaker’s reproduction of extreme high and low frequencies.
Transducers (cone or compression drivers) hav e optimal
ranges in which they are most “comfortable” and reproduce
sound most easily . When a driver is asked to go beyond its
“comfort zone ,” distortion and/or inadequate response
result. While it is possible for a two-way design to emplo y
drivers that cover extreme highs and lo ws adequately , it
would require sacrificing response in other frequency ranges
– notably the vocal range.
Finally , for a traditional two-wa y cinema loudspeaker to
achieve adequate pattern control from 500 Hz to 20 kHz,
the horn’s throat must be of such a size that cancellations
within the throat add yet more distortion, typically in the
700 Hz range .
This paper will endeavor to show that b y dedicating a
subsystem to the vocal region, three-w ay loudspeakers are
better able to handle the new demands of digitally edited,
mixed and encoded soundtracks. Compared to two-wa y
designs, three-way loudspeakers impro ve vocal clarity and
warmth, reduce distortion from both horn-throat cancellation and driver excursion, increase the loudspeaker’s pow er
handling capability , and improves performance across the
audio spectrum including extreme highs and lows.
DIVIDING THE AUDIO SPECTRUM
The audible spectrum spans just under 1 0 octav es ranging
from 20 Hz to 20 kHz (20,000 Hz). Frequency is measured
according to a logarithmic scale; for each octave ascended,
frequency doubles. If middle “C” is 300 Hz, then the “C”
above it would be 600 Hz and the “C” belo w , 150 Hz.
Since a single transducer can only effectively cover so
much of this range, two-w ay and three-wa y loudspeaker systems divide the range into high/low- and high/mid/low- frequency sections, respectively .
Our research and experience have shown that a single
transducer’s comfort zone usually spans no more than three
and a half octaves. Yet traditional 500 Hz crossover design
requires the high frequency compression driver to cover w ell
in excess of five octaves .
While compression drivers reproduce frequencies above
1 kHz quite well, much higher distortion result below that
point due to excessive driver excursion. Critical vocal material in the compression driver’s low est octave (5 00 - 1 000 Hz)
falls outside the driver’s comfort zone . The resulting distortion gives voices a harsh, unnatural edge sometimes rendering them unintelligible. One cannot ov eremphasize the
importance this holds for cinema sound; without intelligible dialog, the cinematic experience is meaningless.
By dedicating a subsystem to the mid frequency range ,
three-way loudspeakers effectively eliminate these problems. EA W’s three-w ay designs use large cone woofers , cone
driven midrange subsystems and horn-loaded high frequency compression drivers. W e place crossov er points at specific
frequencies selected to optimize the performance of each
particular loudspeaker system. In general, the low/mid
crossover is set around 350 Hz and the mid/high crosso ver
around 2 kHz.
Simply adding another horn-loaded compression driver
to cover the mid frequency range does not eliminate distortion from the vocal range; in fact, it aggravates the problem.
If excessive driver excursion at 500 Hz causes a compression
driver to introduce significant distortion, distortion at 300
Hz will be even greater . T o eliminate distortion from the