Nobody said you can't have midrange coming from a large radiating surface. Some of us were pointing out that as the wavelength approaches and exceeds the diameter of the driver, the driver starts beaming and continues from there on up. When you have a range of beaming frequencies, it may sound OK on-axis, but those beaming frequencies go missing in the overall in-room balance. This is expressed as in-room power response.
08-25-14: Dweller
Thanks guys for trying to teach an old dog new tricks.
I'm looking for a way to get a more life-like listening experience.
You're trying to tell me why I can't have it.
A couple of the exceptions cited in this thread have reasons to work. We established that the 6.5" midrange of the Hyperion hps-938 has an active dustcap that funcions as a small-diameter midrange to offset the beaming of the 6.5" section between 2K and the crossover at 3K.
Furthermore, I read up on your Dvoraks with their MTM arrangement with 8" midranges, and it turns out that this is a dipole design. That makes a world of difference, because even though the 8" midranges would be beaming at 1700 Hz, since they're also firing backward and hitting the wall behind, this compensates for the beaming by adding to the power balance in that frequency range. It's also how panel speakers, which tend to have narrow dispersion, have good overall in-room power response, because the backwave hitting the back wall compensates for the narrow dispersion to the front.
As for getting a large radiating surface for the midrange, there are several approaches. First is panel speakers. I have Magneplanar 1.7s. They have approx. 456 sq. in. of bass/midrange radiating surface--radiating both front and back. They are transparent and (with good setup), well-focused. Second, you can get a speaker with dual or multiple midranges such as your Dvoraks had. If you have 8" midranges it helps if they're mounted in an open baffle as your Dvoraks were.
Third, you can get a large diameter full range driver which has compensations for the beaming aspect. I recently visited an audio buddy whose main speakers are Audio Nirvana 15" full-range drivers. These raw drivers are $500-$1000/pair depending on magnet type. His are mounted in large ported enclosures--2'x2'x4'tall, with three big ports to the front. Despite the large diameter, these speakers had excellent power response. I detected no beaming when listening off-axis, and walking around the room the overall tonal balance was excellent with no noticeable suckout.
How'd they do this? The speaker has both a whizzer cone and a copper phase plug. This is exactly what whizzer cones are for--they provide a small diameter cone to keep higher frequencies from beaming. The phase plug also gives a focused surface for these frequencies to bounce off of, futher improving dispersion of the highest frequencies.
He is considering remounting the speakers in an open baffle (like your Dvoraks), however, to open up the sound a bit and eliminate cabinet resonances.
Yet another way to increase radiating surface is with multiple dynamic drivers, with two or more 4-5" midranges. Axiom makes some examples of this including their M100, which has three 6.5" woofers, dual 5.25" midranges plus dual 1" tweeters. They also make an omnidirectional speaker with the same front array plus two more 1" tweeters and two 5.25" midranges firing to the rear. These speakers are engineered by Andrew Welker, who designed those great omnidirectional speakers for Mirage before Klipsch shut them down. Bryston's new line of speakers show a decided dependence on Axiom for their design philosophy, including dual mids and tweeters for lower distortion and greater dynamic range.
Yet another alternative is a tall column speaker with a dozen or so midranges and even more tweeters in a line array.
Finally, the name of the game isn't just radiating area; it's air displacement. A 4.5" driver has a radiating surface of 15 sq. inches. My Mag 1.7s have a radiating surface of 456 sq. inches. Yet, a premium 4.5" midrange might have a maximum excursion of .2", which amounts to around 3 cu. in. of displacement. My big panel, if its excursion is .01" (I'm guessing here, but it's probably in the ballpark), displaces about 5 cu. in. of air even though it's spread over a wider area. So two of those 4.5" midranges would displace about the same amount of air at full excursion.