Ahh the center channel debate... Here's a can of worms which can certainly make for some interesting discussion!
Right on. The front center speaker is neglected and should receive more attention. My take on this is that the horizontal design used for most front center speakers doesn't provide a perfect match for the front left and right. There are a few reasons for this, such as different woofer sizes, and the tendency of dual woofers to cancel each other out, creating an unevenness known as comb filtering.
The first thing which needs to be set straight is that there is no such thing as a "no compromise" home theater design. If we did everything possible to optomize the audio performance within a given space, we would then end up messing things up when we have to place that pesky screen at the front of the room.
Conceptually we would want the sound to eminate from the location of the image, but a perforated screen allowing this brings in other compromises which have to be dealt with. Once we accept that no ideal will exist, then we are able to better evaluate our options, and decide on a solution.
The common center channel for home theaters started life as a solution to an application, not as an ideal. The problem is simple. We needed a speaker which can fit above or below a video display. The first option was to just turn a speaker on its side, but then things feel rather unbalanced visually, and in many cases were sonically unbalanced as well. Enter the horizontal MTM.
The first question should be "can this be made to work?" Yes, but driver sizes and spacing make things difficult and most solutions require compromises. The most significant problem which a designer needs to work with is the spacing of the woofers and the tweeter in these systems. Two horizontally spaced woofers CAN work just fine, but the distance between the two woofers dictates how high in frequency they can be used. While small problems are observed directly on axis, the majority of problems occur off axis in the horizontal. At crossover frequency, any dome tweeter will have very wide dispersion. The top end of the the horizontally spaced woofers will have a response which changes with horizontal angle above some frequency related to their spacing. Uniform and smooth response can only be achieved in the area where both the tweeter and bass drivers have constant response.
While graphs of this type are not commonly viewed in home audio applications, it quickly lends a visual to what is going on with the interaction between two spaced drivers. While the below example is for woofers at 9' spacing, the effect is directly scalable to a 9" spacing by multiplying the frequencies by 12. The separation distance dictates the frequency at which these effect occur, not whether or not they occur.
Subwoofer Spacing by Charlie Hughes
http://www.mississippi.net/~charlie/Art ... acing.html
If you study the above plots enough it can be understood that for an MTM speaker, the woofers need to be as close as possible, and the tweeter to be crossed over as low as possible. Both of these goals have consequences. Tweeters will have increasing distortion and lesser output capability as the crossover is lowered. Placing drivers close requires modest diameter drivers which will have bass output limitations, and a small face tweeter which further limits options. I recall back when I was still doing the retail thing that some of the first set of front speakers that seemed to actually "gel" together had similar qualities like the NHT VS-2 which used a small face tweeter with the woofers crammed as close together as possible.
As home theater grew attention was finally focused on the performance of a center channel, and we saw the entry of Aerial Acoustics and Revel with larger, 3 way center channels. I do recall the first time I heard a CC-3 with a matched right and left, and thinking "now this is what it is supposed to sound like!" One of the strongest benefits of these 3 way designs is the ability to use the same/similar tweeter and midrange compliment which are also oriented the same as the matching left and right speakers. This way both the on axis sound along with the off axis are quite similar. In these 3 way designs the woofers are only used below a few hundred Hz, where we can see from the graphs referenced earlier that they operate essentially as one source. What we find is that a horizontal woofer configuration results in little difference from a vertical configuration aside from the interaction of each with the room. The room interaction difference cannot really be solved unless you place the speakers at similar locations(relative to height for example).
Every time I review a set of surround speakers, I ask for both the horizontal center plus an extra clone of the front left/right. When I compare the two in the center position, the clone always sounds better, providing a more uniform and involving soundfield.
While two speakers can be "matched" to sound quite similar with careful design, for many speakers this only holds true for a limited angle or area in front of the speakers. Off axis there will be differences, and the differing box sizes cause reflections to occur at different frequencies. The more reflective your listening room, the more significant the off axis response becomes. Using 3 identical speakers insures that the on and off axis response of each speaker is the same. There are still problems that arise from the locations we end up placing the speakers, yet the effect is still much more convincing.
The best way to improve an average surround system is not to add more speakers in the rear but simply to match speakers all around.
When I design a surround sound system I usually first consider the quality and robustness of the center channel and then make sure the right and left match appropriately, followed by examination of the subwoofer which can relieve the workload of every main speaker when of sufficient quality. From there then I will see what sort of solution I can come up with for the surrounds.
In defense of using more surrounds, we should acknowledge one definite advantage. Our surround system strives to create the acoustics of a space within our theater. The most significant hurdle to accomplishing this is that the system has to operate in the acoustics of your own space. More speakers in different locations will give the system more dominance over the acoustics in the rear of the room. Of course a little room treatment would probably still provide even better gains... for less money.
Regards,
Mark Seaton
Sound Physics Labs/ ServoDrive