Speakers and Mixing Position, Is Equilateral Triangle a Must?

Before going any further in this blog lets make it clear that the purpose is not to say placing the mix position in an equilateral triangle with the speakers is bad or not beneficiary. Like everything else in acoustics, there is always a compromise. We need to remember that we’re dealing with a fluid medium (air) responding to a HUGE range of energy level difference present within the range of the high frequencies and the low frequencies we can hear (it’ like the difference in weight of a hair and a 15-story building). Understanding the impact of speaker positions and angles with reference to each other and to the room boundaries is key to deciding how to place them and there is always a compromise.

Now that this is out of the way, lets dig in starting with what most people in the audio industry know, the speaker polar coverage pattern. It is important to pause at these basic terms in detail because it plays a big role in how we understand the following part. Due to design limitations and the character of sound waves, speakers do not generate equal sound in all directions. Audio Engineering Society standard number AES-56 standardizes the process of testing and reporting this specification. Without getting into complicated details, in a nutshell, the sound level (in decibels) is measured in a circle around the speaker and the results are plotted on a circular graph. Zero level is in the center of the graph. For example, for the JansZen speaker plot shown (normalized for 1khz); 10 Khz is loudest on axes (to the right of the plot) and drops 15 dB when you stand 30 degrees off access.

The second important aspect to discuss is phase relationship between the sound waves. Since sound energy is cyclic then time become extremely significant, and not just the amount of energy produced. We hear sound because the atmospheric pressure is disturbed by the sound source. The atmospheric pressure increases or decreases following the sound source movements. A frequency wave that is 20Hz means that the atmospheric pressure goes up a little then down a little every 1/20th of a second and this cycle keeps happening until the sound energy is wasted as heat friction. For this wave, time within 1/20th second period decides what “phase” is it in; Just about to cause the increase in the atmospheric pressure, halfway, just beginning to decrease it, at maximum reduction, back to where it started, starting all over again … etc. Phase is reported in degrees because its easier to talk about it that way, since that is relative to the wave itself. A 500 Hz cycle length (period) is 2 msec long, so 180 degrees out of phase is half of that (=1 msec). But 180 degrees out of phase for a 100 Hz wave is 5 msec. We can see that combining full spectrum sound waves that are out of phase from each other will mean quite different things to different frequency ranges in terms of phase. This is significant because the result of the addition of the sound energies coming from the combined waves will be impacted. If we combine two signals with 1milisecond time delay, the 500Hz waves will be 180 degrees out of phase and cancel each other out but the 100Hz waves will only be 36 degrees out of phase and add up energy.

The third and last thing to discuss before discussing speaker positioning is how room boundaries plays into this. The energy coming from the sound source is transferred to air molecules around it and gets relayed from a molecule to the next one to the next which is how the energy moves in the room. Meanwhile the sound source is still cycling back and forth adding more energy to the air in the room. When the source makes one full cycle back to where it started from the sound energy which has been moving into the room is now at a distance from the source which is called the wavelength. Since lower frequencies take longer time to complete a full cycle then the sound energy being relayed has more time to travel and that is why low frequencies have longer wavelengths. As the sound energy arrives to a room boundary, most of it gets bounced back into the room and interacts with the incoming energy from the next wave cycle. To simplify this interaction, we can imagine that the room boundary is basically a speaker located flush with that boundary and playing the same signal (less the amount that is absorbed).

In a room where a sound source is a pair of a stereo speakers; Let us look at sound waves interaction in the horizontal plain only. We will divide the space in front of the speakers into 3 areas, using the polar pattern of the speakers. One, areas on speaker axes where only one speaker is providing sound. Two, areas between the speakers where both speakers are providing sound, off axis so the high frequency levels are reduced but the two-source combination brings it back up. Three, areas near the walls where the sound is provided by the speakers and the wall (the wall acting like a speaker as mentioned earlier). The perceived phantom locations of stereo sound images between speakers are a result of our brain analyzing the time and level differences in sound arriving from each speaker and from boundary reflections. If we entirely kill all the boundary reflections, we will eliminate the interference mentioned in area 3 along with the phase cancelations that come with it, but we also loose a major part of the interaural sound information that gives us stereo imaging. A room with too much absorption has poor stereo quality. When designing a control room, we take a close look at the polar coverage pattern of the speakers and the acoustical characteristics of the boundaries (since it is acting as speakers) and make decisions on the compromises that works best for this room. We place the speakers and mix position where the arrival from both speakers happen at the same time (distance from each speaker is equal) and the room reflections are low enough to not cause major cancelations. As shown in the polar coverage example above, roughly speaking, the coverage looks like half an ellipse which narrows as the frequency goes higher. Pointing the speakers with an angle at the listener will create a diamond shaped area where all frequencies from both speakers are equally produced. Changing the speaker’s angle will only move this “sweet spot” closer or further away along the center line between the speakers. So if that means keeping them away or close to the wall to manage reflections so be it. For as long as the mixing position is within the sweet spot then insisting on an equilateral triangle setup even in cases when it has negative impact in relationship to the acoustical factors will do more harm than good.

Adel Hinawi

Principal Consultant,

Acoustic Consulting Services LLC, NY