Inter-Cluster Misalignment
In multiple-component loudspeaker clusters constructed of self-contained loudspeakers (two-way, three-way, etc.) or individual drivers and horns, there is another source of misalignment not always acknowledged and/or often confused by marketing hype.
All loudspeakers provide varying degrees of directivity or pattern control.
But what is often referred to as a “so-many-degrees by so-many-degrees” loudspeaker can only provide this stated coverage pattern within the frequency range that is within the pattern control provided by the high-frequency horn, and to a lesser extent, the shape of the low-frequency cone driver.
In many designs, the direct radiating woofer provides pattern control that extends from the crossover point down to the frequency range where the wavelengths become longer than the size and shape of the low-frequency cone.
The result is that the specified pattern control occurs, at best, within the upper five octaves of the total of eight octaves of output supplied by typical two-way “full-range” loudspeakers (See Illustrations 4A and 4B).
Thus any cluster consisting of multiple devices covering the same frequency range will have an overlap of coverage. In a well-designed cluster, mid and high frequencies will not overlap significantly.
But even a very well designed cluster, using premium loudspeakers, will present overlap at lower frequencies.
Within the coverage zone of one loudspeaker, overlapping output from one or more neighboring loudspeakers will arrive slightly later. The resulting combined coverage will be altered due to comb filtering, also known as destructive interference.
However, during the process of loudspeaker system optimization, it is often possible to delay one loudspeaker in relation to another in order to minimize this interaction within the primary coverage zone.
Part of this optimization process includes observation of the effects of such delay on each loudspeaker’s coverage.
As delay is applied to a loudspeaker to optimize the response of its neighbor, it is important to insure that the response of the primary loudspeaker (providing primary coverage in its zone) has not been worsened.
Another source of time-based signal corruption is the diffraction that occurs at lower frequencies.
These longer wavelengths literally “wrap around” a loudspeaker enclosure, and in clusters consisting of several splayed loudspeakers, this diffracted energy enters the cavities between the enclosures and is re-projected back out towards the primary listening area (See Illustration 5A).
The interference that results from the combining direct energy with diffracted and redirected energy (that is now lagging behind) can result in severe dips in the lower-mid frequency range.
The size and shape of these dips is dependent on the physical dimensions of the loudspeaker enclosure.
This anomaly cannot be equalized nor can it be corrected with delay, but fortunately, it can be treated physically by inserting appropriate panels that close off the openings to the cavities (See Illustration 5B).
