The very best center speakers are rarely horizontal woofer-tweeter-woofer. THAT is for marketing (low and wide). If you don't need that low wide look, three matching bookshelf speakers across the front would be far better.
Ya, jd said it right. Horizontal MTM configurations run into horrible/massive cancellations/nulls as you move off-axis to the left or right, since now one of the mid/woofer drivers is closer to the listener than the other. Ironically the whole point of a center channel speaker is to anchor the dialogue for people sitting off-axis. So a classic vertical design is superior, its just not attractive or fitting for many systems. But the marketing guys keep selling'em. You'll see a few designs were the designer was aware of this and tried a little. Dickason's cookbook has a design where he gets the two mid/bass drivers almost side-by-side with the tweeter tucked into the space above where the two drivers begin curve away from each other. But even then its not ideal. So to answer your first question, firstly, they aren't optimized to work as center channel speakers by virtue of their shape, and secondly, and which answers both questions, the more accurate speaker is the more accurate speaker and there should be no real difference between the center and the left and right (in regards to waterfall, impulse response, etc.) The only room for debate is in off-axis frequecy response capabilty which also addresses the phase problems any multidriver system deals with.
Stacked speakers have long been used in sound reinforcement. For example, Bose makes a speaker with 4 drivers in a vertical array. The effect is to 'fan' the sound in a horizontal plane with less sound going toward the ceiling and floor -- great for church PA systems. An MTM (or WTW) will not be as extreme but you'll get some of this effect with sound distributed in a vertical 'fan' with sound falling off to the right and left. This is another way of saying the same thing others have said. This is not ideal. I use three identical two way monitors even though I have a fourth sitting in a closet (at the time I could only buy them in pairs). ACI (www.audioc.com), for example, mentions that two ways are better for the center channel on their web site and sells their Emerald LEs (350ea) and Saphire III LEs (650ea) in singles (though I've never heard these speakers).
Even the MTM design period isn't all that great. Biro Technology has a good article and it appears the designer has found that even a vertical MTM suffers 9-12db dips in frequency response, with what I believe is more ambient energy, assuming your ear level with the tweet. I don't know why there are so many mtm. Granted there's no lobing errors or polar axis tilt, better bass extension and spl, plus higer power handling, those problems can be dealt with in other designs. Needless to say I'm not a big fan any symetric array, whether or not its horizontal or vertical.
MTM's do have several advantages if properly constructed. One would have to do some very serious studying of the drivers, their dispersion patterns, optimum crossover frequencies, the spacing between drivers, placement on the baffle, etc... before building a marketable model. then again, this is true of ANY speaker design.
The reason that the horizontal MTM design is so widely used in the HT market is for the very reason that Ezmeralda and Danner "bag" on them. It has limited horizontal dispersion characteristics. Since you have a speaker on each side of it ( the mains in an ht system ), they are trying to limit side to side dispersion and keep the voices centered. Not only does this improve imaging and localization in this case, it also minimizes lobing and interference with the two other front speakers.
Some of the other benefits are that the designers can use multiple drivers for increased surface area, greater power handling and low frequency extension, lower distortion due to reduced excursion, smaller drivers for improved transient response, etc...
Obviously, another consideration would be cosmetics, as this design is typically less intrusive than a vertical array would be. This type of speaker will work best if mounted on top of the TV with a slight downward angle ( to minimize ceiling bounce ).
I think that Vance tends to look at ( and comment on ) these designs as if they were running them as "stereo" speakers rather than "special application" speakers. I agree that the potential for problems are much greater with a design of this type, especially when mounted horizontally. When properly executed and the layout of the room is taken into consideration, they are hard to beat for this type of situation. After all, most people try to watch a tv / movie while viewing the screen as centered as possible. Unfortunately, most of these designs and "home theaters" are not properly executed, so consumers end up with a lot lower performance than their systems are capable of. Like anything else, you have to buy what will work best for your specific situation. Sean
To quote Vance Dickason on page 189 of The Loudspeaker design cookbook, 6th ed., "The center front channel speaker is commonly positioned horizontally , with the drivers in opposite acoustic polarity to the left/right front speakers. This is done strictly for aesthetic reasons." Note that last sentence. And on the top of page 189 he has graphs showing the effects of the various driver arrangements and their effects on off-axis frequency response: figure 10.11 correllates to figure 10.15 on the facing page of a horizontal MTM. From the first paragraph in the second column on page 189 "The consequence for the horizontally polarized speaker at 30 degree off-axis is a dip in the repsonse at the crossover frequency of nearly 15db." The other three driver arrangements perform far more admirably given the task. And this is also why THX certification is not given to driver arrangemnts of a horizontal mtm (should the manufacturer even bother to apply).
ALL of those variables ( including the "off axis dip" ) would change with different drivers, altering the spacing of those drivers on the baffle, changing the size of the baffle, altering crossover frequencies, etc... Vance even states that the results that he shows for the horizontal mtm array are NOT directly comparable to the vertical mtm array as they are completely different speakers. He's using different drivers in a different box for the two sets of graphs.
While i think that his book is EXTREMELY useful, it is not the "end all" or have all of the answers. It is nothing more than an ongoing collection of research notes and data that he has collected and continues to update. In case you haven't noticed, many, many, many of his graphs do not use the same spacing or plotting characteristics. They are therefore NOT directly comparable in many instances. One would have to extrapolate data from the given charts, some of which are less than clearly marked, for any type of meaningful comparisons under many of the circumstances that he has documented.
Besides all of that, why would anyone that had any knowledge about speakers purposely wire an entire speaker out of phase / reverse acoustic polarity from other speakers in a multiple channel array ??? I can't say that i've inspected dozens upon dozens of center channel mtm's, but i have seen more than a few. None of them were wired with the opposite acoustic polarity to what one would expect. Doing something such as that would cause unstable and very diffuse imaging resulting in speech / vocals that were NOT "anchored" between the L - R mains. This would completely defeat the purpose that a center speaker was designed to achieve. Sean
Nowhere did it say the speaker was wired out of phase, wiring was not even mentioned. It stated "the drivers in opposite acoustic polarity to the left/right front speakers." This would also refer to horizontal tm designs, along with the horizontal mtm's in relation to the left right being vertical driver arrangement. But what is amusing is that many speakers will actually have the tweeter wired out-of phase with the lower frequency driver to maintain a smooth frequency response throughout the crossover region depending on crossover chosen. So dependig on whether or not the center speaker purchased was of the same make as the front l/r, one could actually have thier high-frequencies in the center channel out of phase with the left and right.
I agree with all of the points made. It is not that uncommon in multiple driver designs for one or more of the speakers to be wired out of phase with the others.
I would also like to clarify that i "mistook" the "opposite acoustic polarity" as being wired out of phase. If two different cabinets were wired out of phase in relation to each other, one driver would be pushing outward while the other would be pulling inward. This would result in "opposite acoustic polarity", hence the confusion. It did not strike me at the time of reading this thread that Vance was referring to the polarization of drivers ( vertical vs horizontal ) and dispersion characteristics. Sean
According to the Biro's designer, one does not need a center channel if the mains are less than 2.5 meters apart (personal communication). With my "improperly designed" MTM JMlab's I keep the centers no farther than 7 1/2 feet.
Kslim, you can do fine without a center channel in a small room. If all your audience fits in a couch and in a few pillows on a rug, you're fine. If you get into bigger projection units and TVs you will need the center channel to anchor the dialog.
Now, who says you can't have TWO small two-way speakers as center channels? Why keep one in a closet? One can place one above and one below the TV, or any way that suits you best.
Ever heard of The Penguin Principle?