The question of what is the best type of speaker system for a house of worship is not a simple one. The answer lies somewhere in amongst the aged questions of: where do all the other socks go, what is your favorite color, and how many licks does it take to get to the middle of a Tootsie Roll Pop?
The fact is the question itself is immensely complex. There are so many factors that go into determining what will be an applicable speaker system that I prefer to call it a process rather than a question. What follows will be a general discussion on the subject and is intended as a primer on speaker systems application as they relate to houses of worship.
The process begins with a string of questions like, but not limited to; what will the system be used for? What is the style of worship? How important is the spoken word? How loud does it have to be able to get? By the way, these questions are independent of the architectural space and the budget. Rather, they are of what it is supposed to do, not how. This requires pulling together the ministries that will be impacted by it and putting together a list of operational requirements for the system. With this in hand you are ready to start to look at the application of technology to meet the goals.
Next, keep in mind that in the church environment there are two major uses of a speaker system: Voice reproduction and music support. Unfortunately the two have rather differing requirements. It goes without saying that either of these systems must be free of hums, buzzes and distortion. They must have a natural full frequency range sound and be capable of reproducing sound pressure levels (SPL) consistent with the type of programs they are supporting. The sound distribution of both systems must be consistent in SPL and frequency response. The maximum SPL variation should never exceed 6dB over the majority of the intended coverage area. There are however, several differences which set voice reproduction and music support systems apart.
In voice reproduction, the system must be particularly articulate and have a strong localization to where the person is speaking from. Articulation and speech intelligibility are the ability to understand the spoken word. Speech intelligibility is directly related to the listeners receiving the reinforced speech directly and with only a very few early reflections no later than 20ms. Late arriving reflections and high reverberant-to-direct ratios will conspire to reduce speech intelligibility. One of the terms we use to quantifiably describe articulation is the Percent Articulation Loss of Consonants or %ALcons. %ALcons is the percent of misunderstood words due to the loss of consonant sounds. For instance a %ALcons of 10 would mean that 10 words out of 100 were misunderstood. By the way, a %ALcons of 15 is really poor and generally considered to be the highest value acceptable in most cases. In regards to point of origin we do not like to see the person talking one place and hear them coming from another. It is most natural to have the two localize together. If a speech system has poor articulation and localization, the listeners will have to force their attention to the subject and will tire in time. This is called listener fatigue and is one of the reasons people sometimes sleep in church.
On the other hand, music support systems have characteristics of wide frequency bandwidth, diverse sound source localization and higher SPL requirements. Articulation is not quite as important as in speech reinforcement systems but must be considered. A music support system must have a frequency bandwidth much greater than a speech reinforcement system. If we look at the fundamental frequencies of voice reproduction you will find that the entire vocal band lies between 80Hz and 1.25Khz. Harmonics and consonant sounds will extend that range up to 6KHz or so.
Music support systems have to support a frequency range of 25Hz out to 16 to 20Khz to include the harmonic content. An interesting point to make is that the highest fundamental note on a piano is 4186Hz. In contrast to an individual speaking to an audience, music is usually created by several people located in different locations on the platform area. If we consider what our eyes see in regards to the overall width of the visual stage we would like to hear a soundstage of similar width. We would even like to hear the individual instruments and vocals relative to their location on the stage if possible as well. This requires a multi-channel system capable of synthesizing the location for each of the music sources. This multi-channel type of system is not particularly suited to voice reproduction.
Another consideration is the SPL required for music. Elevated oratory voice is around 80dB SPL while a loud symphony orchestra passage might peak at around 115dB SPL. The power requirement to reproduce the music support is better than 100+ times the speech requirement. Depending on the type of music program it is not unreasonable to expect a music support system to be able to generate 110dB SPL long term.
So you see, the differences in the two requirements are significant and this is why it is very important to determine your program requirements before looking at a speaker system. So what am I saying?
The point is, to ultimately and most applicably support both requirements there would most likely be two separate speaker systems. Don’t panic! It’s OK! I’m not suggesting that if you don’t do this that you are doing it wrong. Remember that this would be the best, ultimate, optimum arrangement and is predicated on requirement not capability. So let’s talk about application and some of the other considerations that come into play.
In general the room has a lot to do with what type of speaker system will best work in it. Difficult acoustic spaces require much greater attention than spaces that have well-controlled acoustics. Small rooms are different than large rooms. The shape of the room will make a great deal of difference in what approaches work and those that will not. Let’s look at some of the highlights.
In rooms that are deep rather than wide the visual soundstage for music is narrower. In those cases a full bandwidth central cluster will most likely be sufficient to reasonably support the music program without distracting displacement of the musicians. Voice reproduction will be optimum with the best point of focus and articulation. In rooms that are wide in respect to their depth the soundstage is also quite wide, especially out towards the sides of the seating. In these rooms several flavors of multi channel systems could be employed each with their pros and cons.
A properly implemented two channel (stereo) system will work well for the music but will not be the best option for the spoken word. Speech will suffer from time offsets between the two speaker systems that in many rooms could easily fall out of the 20ms time window and degrade articulation.
An arced array, sometimes referred to as an exploded cluster, is another option and may be implemented as a single channel or multi-channel system. These systems are a bit tricky and need to be implemented with care to minimize time offsets while guarding against phase induced cancellations through out the coverage area. A common mistake is to use too few devices of too wide beamwidth which maximizes all of the negative effects.
An approach that is becoming more common is the Left Center Right or LCR array. Actually the LCR array is a combination of the center cluster with stereo left and right clusters. Several audio console manufacturers make specific models that have the routing to properly distribute the various inputs to the appropriate clusters. Of course these are the two different systems that were talked about earlier. There are a couple of special implementations of the LCR concept that can mitigate the need for three complete clusters, however outside of the scope of this discussion. One way to approach an LCR system where budget is limited is to design a LCR system and install it in two steps. First the center cluster followed by the L/R clusters when budget allows.
Rooms that have difficult acoustics are a real challenge. A difficult acoustic space might be one that has all hard surfaces with several of them parallel to each other and has long reverb times. In general it is best that the sound system not unnecessarily excite the reverberant field in any room but especially not in these highly reverberant spaces. Use of tightly controlled highly directive speaker devices is a must in reverberant spaces. What that means is that low Q or wide-angle coverage speakers will not provide satisfactory results in tough rooms. Presently the most common way to control beamwidth or Q in a loudspeaker is to use some sort of waveguide (horn) to control the pattern. The size of the horn will directly relate to its ability to control the lower frequencies. You must control as much of the bandwidth as possible if you are to be successful. So speakers that have horn loaded low-mid and midranges as well as high frequencies are preferable. In some cases a delay system can also be used to keep from planting loudspeaker energy on walls and unnecessarily exciting the room.
A low ceiling is also a constraint to be reckoned with. Loudspeaker height will determine how far out into the room you will be able to cover and still keep within the 6dB change in level. At the point out in the room where the distance from the speaker to the front row (D1) is double (2xD1) you will most likely be 6dB in direct sound level. This also depends on the vertical coverage angle of the speaker. The solution is to add another speaker out in the room, just prior to where the level drops, and apply signal delay to the feed for it. This requires deliberate planning and careful placement of the speakers and could easily occupy an article of its own. If you are not certain on what to do with delay speakers don’t experiment, get help from someone who does. In low ceiling buildings it is not uncommon to have to use several zones of delay.
Sub woofers can be added to any system to enhance the low frequency capability of the system. By planning subs into the design from the start you can alleviate the main clusters from having to produce frequencies below 70Hz or so. This can greatly reduce the size of the devices used up in the air and make for less aesthetic impact on the room.
Two points to keep in mind about subs. They are virtually omni-directional so half-space load them. That is to place them on the floor or some other solid surface that concentrates all of the output into half of the space. You will gain 6dB SPL in the subs level by doing this. The other is to rigidly mount them but not to flimsy structures. Loose hung subs will have a mushy low impact sound and lack efficiency. I prefer to feed the subs from an auxiliary output of the console rather than from the main outputs. By doing this you have control over what goes to the subs rather than using EQ filters to attempt to keep things out of them. You will find that there really are not that many sources that require subwoofer support but they are important to the overall sound.
These are but a few of the myriad of considerations that go into selecting the proper speaker system for your church. This information is intended as a primer on the subject and in no way is the complete and definitive dissertation. If you are interested in learning more on this subject look into attending one of the Synergetic Audio Concepts (Syn-Aud-Con) seminars or others like them and you will have a better basic knowledge to work from. This information will help you in determining if someone is trying to sell you a bag of goods in regards to your speaker system needs. I will always recommend that you locate a professional audio firm who is regularly engaged in the scientific application of technology to assist in providing solutions for your sound system requirements. Mmmmm, sounds like a good subject for another article to me.