Originally Posted by Toole
Although reflections appear not to be great problems, it
is reasonable to think that there must be a level above
which the good attributes are diminished and negative attributes
grow. Obviously an empty room is not a comfortable
listening environment, even for conversation. The
furnishings and paraphernalia of life tend to bring normal
living spaces into familiar acoustical territory. Custom listening
spaces need to be treated. In all rooms absorption,
scattering or diffusion, and reflection occur, and devices to
encourage each are commonly used by acousticians.
It appears that much of what we perceive in terms of
sound quality can be predicted by the anechoic characterization
of loudspeakers. Because most of these data pertain
to sounds that reach listeners by indirect paths, it is proper
to suggest that nothing in those indirect sound paths
should alter the spectral balance.
For example, a 1-inch
(25.4-mm) layer of fiberglass board at the point of a strong
first reflection is effective at removing sound energy
above about 1 kHz. From the perspective of the loudspeaker,
the off-axis response of the tweeter has just been
greatly attenuatedit will sound duller and less good.Obviously if the purpose of the absorbing material is to
attenuate the reflection, the material should be equally effective at all frequencies.
Given the duplex nature of
sound fields in small rooms, it seems reasonable to expect
similar performance at all frequencies above the transition
In their examination of the audibility of reflections, Olive
and Toole looked at detection thresholds as high frequencies
were progressively eliminated from the reflected
sounds, as they might be by frequency-selective absorbers.
They found that only small to moderate threshold elevations
occurred for low-pass filter cutoff frequencies down
to about 500 Hz, where the investigation ended. Removing
the high frequencies alone is not sufficient to prevent audible
Finally there are the indications that the precedence effect
is maximally effective when the spectra of the direct
and reflected sounds are similar , , . If the spectrum
of a reflection is different from that of the direct
sound, the probability that it will be heard as a separate
spatial event is increasednot a good thing.