When a company deviates from the standard design of a dome tweeter above one or more cone midrange/woofers, is it just a way of distinguishing themselves from the crowd, or are there real sonic advantages?
Examples I am thinking of include KEF (Uni-Q), Mirage (omnipolar), Martin-Logan (electrostatic), Dali (ribbon tweeters), etc.

When a company deviates from the standard design of a dome tweeter above one or more cone midrange/woofers, is it just a way of distinguishing themselves from the crowd, or are there real sonic advantages?
Examples I am thinking of include KEF (Uni-Q), Mirage (omnipolar), Martin-Logan (electrostatic), Dali (ribbon tweeters), etc.
At the risk of you trashing me for responding, I think the answer is "both". There are sonic advantages to certain designs, but I also find that a certain speaker company sticks with a particular design as their "signature".

Speakers, by definition, are a compromise. There are advantages to point-source behavior (in an anechoic room) and advantages to controlled dispersion (in a real room). Likewise, both low-order and high-order crossovers have some advantages/disadvantages. Some speakers do better near walls than others. For example, my Snells are designed to be placed near a room's front wall, and can be near a side wall, because of the enclosure design, drivers, and crossovers. Other speakers need to be well out into a room.

Dipoles (like electrostats) are interesting. Because the front/back are out-of-phase, the side radiation is nil, thereby eliminating side reflections, but there's a backwave. Bipoles (Mirage) send sound everywhere, so they may not sound as precise, but they'll fill up a room well. Some speaker designs are more conducive to certain types of music, so it's a personal preference thing.

Another speaker "compromise" is the use of multiple drivers, because it breaks-up the sound into different source locations with different qualities. So it's important to have the drivers integrate well, and that's where some of the common driver placement configurations come from. And as I mentioned, there are different crossover designs. A first-order crossover causes no phase distortion (good), but you end up with two drivers producing the same signal (overlap - bad). High-order crossovers have the opposite qualities.

Then there's cabinet diffraction - which is when the sound from the driver bounces off the cabinet. It's usually bad for high-frequency drivers to do this, as the high-freq is where we get our directional cues from. Solving this problem results in some of the more adventurous designs.

Harold, can you post specific models or pictures?

I am not going to "trash you for responding"! I have nothing against you, per se, we just disagree about CD players. I happen to believe, by contrast, that there are large audible differences between speakers.
Your comment about crossovers is ineresting. I might need to investigate crossoverless single-driver designs as well (Role and NSM come to mind.)
I haven't heard any of the KEFs, but I always found the Uni-Q concept interesting, especially since they also use curved cabinets to resist diffraction.

Harold, can you post specific models or pictures?

Some of the makes/models that look especially interesting include the KEF iQ3, Mirage OMD5, Magnepan MMG, and Duevel Planets. (Would the Magnepans require a beast of an amplifier?)

Well, with the Kef IQ3 you're looking at the tweeter mounted inside the cone in a ported speaker. This is done IIRC, so that the locations of the tweeter & woofer (midwoofer, whatever) is coincident with respect to the three axes. It's part of their shtick if you will.

Well, with the Kef IQ3 you're looking at the tweeter mounted inside the cone in a ported speaker. This is done IIRC, so that the locations of the tweeter & woofer (midwoofer, whatever) is coincident with respect to the three axes. It's part of their shtick if you will.

And does it sound better than the same woofer and tweeter in the same cabinet would sound if mounted separately? That is the real gist of the issue.

Your comment about crossovers is ineresting. I might need to investigate crossoverless single-driver designs as well (Role and NSM come to mind.)

A typical high-order crossover acts like a brick wall, rapidly attenuating out-of-band frequencies from each driver. The pro's are that each driver only handles a specific freq range, and there's very little overlap (summing) of multiple driver outputs. The con's are that the out-of-band freqs are heavily phase-shifted, and also the off-axis response may be distorted compared to the on-axis response.

A first-order crossover is referred to as a linear crossover, because there's no phase shift. The out-of-band freqs are gently attenuated at each driver. The pro's are no phase shift and a uniform off-axis response. The cons are significant overlap of driver outputs (summing), which makes integration of the frequency response difficult. Also, it's a strain for the drivers, which have to cover a much wider freq range. And the woofer will start beaming the high frequencies, plus the tweeter may overheat.

Having no crossover will cause all the driver outputs to be summed, up to the natural limit of each driver. That seems hard to design around, plus each driver will be handling frequencies it's not intended for.

Having no crossover will cause all the driver outputs to be summed, up to the natural limit of each driver. That seems hard to design around, plus each driver will be handling frequencies it's not intended for.

In the designs I am thinking of, there is nothing to "sum" because there is only one driver.
Here is an example:http://www.roleaudio.com/sampan.html

And does it sound better than the same woofer and tweeter in the same cabinet would sound if mounted separately? That is the real gist of the issue.You can't get the same woofer and tweeter separately.

You can't get the same woofer and tweeter separately.

Isn't it basically a separate tweeter mounted where the dustcap on the woofer would otherwise be?:confused:

An important thing I forgot to mention, is that drivers typically behave linearly only within a certain freq range. The crossover is there to keep the out-of-band freqs out of the non-linear range of each driver. Because of this, first-order crossover designs are more expensive, because the drivers are running over a much wider range, and the crossover has to be customized for the non-linearities.

But with a high-order crossover, each driver essentially outputs nothing outside it's ideal linear range, so the crossover can be simpler.

Of course, the no crossover speaker would have no control over the non-linearities of the drivers.

And does it sound better than the same woofer and tweeter in the same cabinet would sound if mounted separately? That is the real gist of the issue.
The advantage is that it's literally coming from one point, not from 2 different points like with a separate woofer/tweeter. But I also see problems: the tweeter output will be reflected off the woofer cone, causing coloration, and also the woofer is vibrating around the tweeter at varying frequencies.

In the designs I am thinking of, there is nothing to "sum" because there is only one driver.
Here is an example:http://www.roleaudio.com/sampan.html
Well, this certainly eliminates crossover issues! To me, this looks like one overpriced Bose speaker. How can a single 3.5" driver produce either any bass or any high frequencies (without severe beaming)? The high freq beaming will mean very limited dispersion, or the opposite of a point source. That's why they invented dome tweeters all those years ago.

And as far as bass, the 3.5" driver can't go below it's own resonant frequency (which must be around 150Hz), so how much can the transmission line do? That freq response spec is either bull, or it's at a very low SPL.

Well, this certainly eliminates crossover issues! To me, this looks like one overpriced Bose speaker. How can a single 3.5" driver produce either any bass or any high frequencies (without severe beaming)? The high freq beaming will mean very limited dispersion, or the opposite of a point source. That's why they invented dome tweeters all those years ago.

And as far as bass, the 3.5" driver can't go below it's own resonant frequency (which must be around 150Hz), so how much can the transmission line do? That freq response spec is either bull, or it's at a very low SPL.

Hmm. I was afraid their claims might be bogus! I have never heard them, since they are only sold online.

Here is another odd design that I have recently run across:http://**********/component/page,shop.product_details/flypage,shop.flypage/product_id,122/category_id,21/manufacturer_id,0/option,com_virtuemart/Itemid,37/

They claim it is "omnipolar", but the tweeter (responsible for most of the directionality!) is mounted conventionally on the front. What gives?:confused:

The advantage is that it's literally coming from one point, not from 2 different points like with a separate woofer/tweeter. But I also see problems: the tweeter output will be reflected off the woofer cone, causing coloration, and also the woofer is vibrating around the tweeter at varying frequencies.Well, the tweeter output is reflected/diffracted regardless of how it is mounted, so the issue is how that reflection is dealt with. The integration used by KEF and, also, by TAD/Pioneer takes this into account in the design of the cone which is supposed to act as a waveguide for the HF.

Geez Harold, they've got a website and forum. Ask 'em!

Here is another odd design that I have recently run across:http://**********/component/page,shop.product_details/flypage,shop.flypage/product_id,122/category_id,21/manufacturer_id,0/option,com_virtuemart/Itemid,37/

They claim it is "omnipolar", but the tweeter (responsible for most of the directionality!) is mounted conventionally on the front. What gives?:confused:
It's a dome tweeter, so it'll have good dispersion, and the small cabinet footprint will minimize diffractions, so this will probably work well in that regard. I'm not so sure about an up-firing woofer though...

Well, the tweeter output is reflected/diffracted regardless of how it is mounted, so the issue is how that reflection is dealt with. The integration used by KEF and, also, by TAD/Pioneer takes this into account in the design of the cone which is supposed to act as a waveguide for the HF.
I guess that's what you pay for with their computer design modeling.

Well, the tweeter output is reflected/diffracted regardless of how it is mounted, so the issue is how that reflection is dealt with. The integration used by KEF and, also, by TAD/Pioneer takes this into account in the design of the cone which is supposed to act as a waveguide for the HF.

How can the cone act like a waveguide? :confused:

How can the cone act like a waveguide? :confused:By its shape. See http://sound.westhost.com/articles/waveguides1.htm for a guide to waveguides. :cool:

Speaking of waveguide designs, please add anything that applies to this thread, if you have the time and inclination.

http://www.avsforum.com/avs-vb/showthread.php?t=1058973
Full range waveguide speakers under 2k a pair

Most dome tweeters are recessed into a curved plate. That is a waveguide, is it not? As for a deep waveguide (like the one in your link), how is that different from a horn?

Most dome tweeters are recessed into a curved plate. That is a waveguide, is it not? As for a deep waveguide (like the one in your link), how is that different from a horn?See the link. Merely putting a radiator in a depression does not constitute adding a waveguide or horn.