You can tune a trombone, but you can only argue with a sound system.
Tuning a sound system is a figure of speech. What one is actually doing is pushing things around to find a happy medium. If one is constructing a sound system for presenting sound effects, the system equalization can be manipulated to fulfill the artistic requirements of the sound system. This is often much easier and more straightforward than manipulating a system for technical reasons, such as feedback control, coverage extension or system combination.

Complex sound systems can require extensive manipulation of the signals to coax the different pieces of a system into working with each other. If every situation could be covered with one speaker, things would be significantly simpler, however, here in the real world, combining speakers and/or speaker systems to increase coverage or increase sound pressure level is the norm. When more than one speaker carries the same signal, the signals from those sources interact with each other in predictable, but complex ways.In complex speaker systems, the game is all about reducing the damaging effects of signal interaction.

Phase = Time = 20,000 Different Things

Fig. 1 Animated illustration - Phase
Fig 1 Animated illustration of the additions and subtractions that occur when two identical waves are added together with varying phase

Fig 1 shows two waves of equal frequency and level. When the two signals are in phase with each other, they add and produce a wave that is twice the level of the original signals.

When the phase of one wave is changed relative to the second, the addition of the signals produces a result that is less than twice the level. When the two waves reach 90° (1/4 of a period) out of phase with each other, the result wave is equal to the level of the original waves. At 180° of phase difference, the signals cancel each other out completely.

Why is that important?

Every frequency has a different period (the time it takes to produce a single full wave) and therefore a different wavelength

For example:

  • The period of a 1000Hz sine wave is 1 millisecond.
  • The period of a 1500Hz sine wave is 0.6 milliseconds.

Phase is a time unit relative to frequency. The 90° and 180° points for each frequency occur at different times for different frequencies. The damaging effects are compounded when one applies delay to a signal being sent to a speaker. When one chooses to delay a signal the amount of delay is usually based on the time it takes the sound to reach the listeners ear, and that time is determined by the medium through which the waves pass. In this case, air. The speed of sound in air is variable as well, and based on the density of the air, which is determined by density of the air, the largest two variables of which are temperature and humidity. However, that is another story for another time.

Fig. 2 Two sources - single system
Fig 2 Two Speakers to be ‘Married’ into a single system.

Of course, the seemingly simple activity of aligning two speakers is shockingly complicated. Let us look at a 'simple' example. In the picture, you see two speakers. One above the stage is the Main signal source and the one in the house is, in this case, for coverage extension. These two sources are to be 'married' together to work as a single system. (Fig 2) It would seem that alignment of the signals requires the speaker closer to the listeners to be delayed the distance between the speakers relative to the listener's position. While that is mostly true, what is the actual amount of delay that should be applied to effectively marry these two sources? That's a tough question to answer. We can pick a listening position, measure, and set the delay equal to the measured value. (Fig 3) That will get us close, but only for the one single point that was measured. (Fig 4 and 5)

Fig. 3 Two sources - Impluse response
Fig 3 Impulse Response showing the time difference between the two sources at the position of the mic. In this case 13.81ms.
Fig. 4 Two sources - Comb Filtering
Fig 4 Response as measured from the fourth row from the back with no delay applied to the speakers. Shows evidence of comb filtering. Not a nice sound to listen to.
Fig. 5 Two sources - Time Aligned for one point
Fig 5 Response as measured from the FOURTH row from the back with the 13.81ms applied to the speakers. This is a much nicer response, but only for that one point.
Fig. 6 Two sources - Out of position
Fig 6 Response as measured from the FIFTH row from the back with the 13.81ms applied to the speakers. Changing listening position drastically changes the response.
Fig. 7 Two sources separated by a distance
Fig 7 Peaks and valleys at one frequency, produced by two sources separated by a distance

If we move to a different position, say one row back from the first position, and measure again, the relationship between the listener and the speakers is different, if only a little. It’s those little changes from one position to another that make aligning two sources a lesson in compromise. (Fig 6) One might be able to get the alignment right for one person, but it won’t be exactly right throughout the entire coverage area of a particular speaker.

How to hear it…

Listen to some full range music in mono (single channel) to start. This can be a stereo recording mixed to a mono signal. Some recordings are better than others are at summing to mono. Have a listen to several different recordings to see what works best. Using mono music is important because it will allow you to hear more easily the interactions between the speakers. Interactions are most apparent with correlated (the same) signals. A Stereo recording consists of two channels of audio. These two channels are most often different or non-correlated signals, and therefore speaker signal interactions in the room won’t be as obvious to the ear. It is important to use familiar music, perhaps even in the style of the music that the system will amplify or reinforce. Try to steer away from music that purposely uses distortion or noise in its content. These things will make it more difficult to hear the summations and cancellations. While listening to a mono signal, walk the space listening for inconsistencies in the overall sound. There may be places in the room where there is a precipitous rise in low frequency energy, and conversely other locations where there low frequency energy is lower than average. As you walk you may also hear the high frequencies appear to come and go. This interaction (comb filtering) is created when the same sound arrives at your ear at slightly different times. The arrival time differences create some places where the waves add together and others where the waves cancel. Notice that the peaks and valleys in the low frequencies are feet apart and the peaks and valleys in the high frequencies are quite close together. The locations where peaks and valleys are observed are directly related to the wavelength of those frequencies. Fig 7 shows two speakers separated by a distance. Both are producing the same frequency. As one moves from the center line arrival times from the two sources begin to vary producing the pattern of summations and cancellations shown.

Having done this, you will have a much better understanding of the interactions between the speakers.

For more an in depth information on complex sound systems, sound waves and system optimization see Sound Systems:Design and Optimization by Bob McCarthy