Handbook for Sound Engineers

60 Chapter 3

3.11.4 Precedence (Hass) Effect

When two clicks are presented simultaneously to a lis-
tener, one on the left and one on the right, the listener
would perceive a click in front—i.e., average the local-
ization cues of the two clicks. However, if one of the
clicks is delayed (up to 5 ms) compared to the other, the
listener still perceives them as one fused click but will
localize the fused image with cues of the first click only
and ignore the localization cue of the later one. For
delays longer than 5 ms, the listener will hear two dis-
tinct clicks instead of one fused click. For speech, music
or other complex signals, this upper limit can be
increased to about 40 ms. This phenomenon that the
auditory system localizes on the first arrival is called
precedence effect, or Haas effect.54,55

The precedence effect has very practical uses in
audio. For example, in a large church, it may not be
possible or practical to cover the entire church from one
loudspeaker location. One solution is to place a primary
loudspeaker in the front of the church, with secondary
loudspeakers along the side walls. Because of the prece-
dence effect, if the signal to the loudspeakers along the
walls is delayed so that the direct sound from the front
arrives at a listener first, the listener will localize the
front loudspeaker as the source of the sound, even
though most of the content will actually be coming from
the loudspeaker to the side, which is much closer to the
listener, and may even be operating at a higher level.
When such systems are correctly set up, it will sound as
though the secondary loudspeakers are not even turned
on. Actually turning them off demonstrates exactly how
important they are, as without them the sound is unac-
ceptable, and speech may even be unintelligible.

3.11.5 Franssen Effect

The Franssen Effect^56 can be a very impressive demon-
stration in a live room. A pure tone is played through
two loudspeakers at two different locations. One loud-
speaker plays the tone first and is immediately faded,
while the same pure tone is boosted at the other loud-
speaker, so that the overall level is not changed signifi-
cantly, Fig. 3-23. Although the original loudspeaker is
not playing at all, most of the audience will still believe
that the sound is coming from the first loudspeaker. This
effect can last for a couple minutes. One can make this
demonstration more effectively by disconnecting the
cable to the first loudspeaker, and the audience will still
localize the sound to that loudspeaker. The Franssen
effect reveals the level of our auditory memory of
source locations in a live room.

3.11.6 Cocktail Party Effect and Improvement of

Signal Detection

In a noisy environment, such as a cocktail party, many

people are talking simultaneously. However, most peo-

ple have the ability to listen to one conversation at a

time, while ignoring other conversations going on

around them. One can even do this without turning his

or her head to the loudspeaker. As we mentioned earlier,

one benefit of binaural hearing is the ability to spatially

filter. Because the talkers are spatially separated, our

auditory system can filter out unwanted sound spatially.

Patients with hearing difficulties usually suffer greatly

in a noisy environment because they are unable to pick

up an individual’s conversation out of the background.

Because the background noise is normally in phase

between the two ears, in electronic communication, one

can reverse the phase of a signal in one ear and make it

out of phase between the ears. Then signal detection is

much better due to spatial filtering. So, in general,

binaural hearing not only gives us localization ability,

but also improves our ability to detect an acoustical

signal, especially in a noisy or reverberant environment.

3.11.7 Distance Perception

Distance cues are fairly difficult to replicate. In free

field conditions, the sound pressure level will decrease

6 dB with every doubling of the distance between a

point source and an observer. Thus reducing the vol-

ume should make us feel the source is farther away. In

practice, however, we tend to underestimate the dis-

tance: the level has to be attenuated by 20 dB in order to

give us the perception of a doubled distance.^57 Of

course if we do not know how loud the original source

is, we do not have an absolute scale based on level.

Figure 3-23. Franssen effect (Reference 56). The figure

shows the level of two loudspeakers at two difference loca-

tions in a live room. Loudspeaker One plays a pure tone

first, and is immediately faded. Meanwhile, the same tone

played by Loudspeaker Two is boosted, so that the overall

level in the room is not changed significantly. After Loud-

speaker One stops playing, listeners will still perceive the

sound originated from Loudspeaker One, up to a couple

minutes.

Loudspeaker

One

Loudspeaker

Two

0

Time