That is another nice review James. Obviously they have all the attributes to be good sounding speakers. Although I have to say the bass response is more limited than I would expect for a speakers of that size. The 2.1 KHz crossover is nicely managed but there is discontinuity visible as is to be expected from the timing issue of filters of that order.
But it really does highlight the confounding factors of center channel speakers.
I don't think it can be emphasized enough that a center speaker really does need to have power handling comparable to the right and left main speakers. However it most situations it is just not practical to have it as large as many towers. Usually the center is in an adverse position, near a rear boundary and close to a large reflecting TV screen. Yet it has to have superb voice clarity for dialog. As I have said here many times, center speakers present by far the biggest challenge to speaker designers of all other speakers. The challenges are many fold greater.
Lobing issues of the standard and ubiquitous horizontal MTM designs really make them non starters.
I have found boundary issues are a serious problem, which makes some frequency tailoring below the 400 Hz range mandatory in my view. Reinforcement of the rear boundary is a highly significant issue from my investigations, especially below 100 Hz.
The next issue is time delay. The standard crossover topologies employed in the speech discrimination band with spaced drivers do definitely impair speech intelligibility.
So as I have said before a full range speaker works well, but with the drawback of limiting power output.
A well done coaxial design is a very good solution. I use two SEAS coaxial drivers, which gives a speaker with 400 watt power handling capacity or close. It is also a low Qts driver (0.24) which gives it a good leg up in the speech department. Two drivers allows for tailoring of proximity effects. So that has proved a really good solution. With TL loading F3 is specked and measured at 44 Hz. Size remains a problem though. It is however smaller than the speaker under review.
The other solution is the three way center with a vertical mid/tweeter arrangement. As you point out this requires that the spacing of the either side woofers be placed within a wavelength of crossover frequency. This really does require a mid that can cover the whole of the speech discrimination band. This means the driver needs to be physically small and powerful. That leaves precious few candidates, but they are out there. The Morel unit I selected is small and powerful. It also allows for the selection of a 3/4" dome rather then a 1" dome, which does help with HF dispersion.
My last observation is the I really think the usual common QB4 ported box arrangement is not ideal for center speakers. A sealed alignment I think is preferable. This by itself tends to ameliorate the boundary issues, by having a falling LF response from a higher frequency but at a gentler slope.
The real test of a center speaker is whether it has excellent speech intelligibility without having to have its level increased which destroys program balance. In addition to intelligible the speech must sound like natural human speech.
This relatively small center is capable of very high spl. without distress. Crossover points 400 Hz and 4 KHz. So it fulfills my criteria for center speaker design.
This really is the area were a lot more work is required. The development of active speakers with DSP to better handle boundary effects and above all, allow for perfect time alignment, I believe would be a huge step forward and solve a lot of outstanding problems.
