Thanks for the review. Altogether a good, well-reasoned review of what the product does and what my objectives were.
The only commentary I really have is with respect to your interpretation of some of the measurements, and the anomalies seen in the response curves.
Pretty much every anomaly that shows up is not due to resonances, but due to diffraction related artifacts.
It’s astonishing when you look at driver response, and make measurements going from very close in, all the way out to 2m, and see the effects of diffraction start to intrude and mess up the smoothness of the response!
Even the dip at 2.8kHz is due to diffraction imposed on the tweeter, and not the crossover point. The 8kHz dip is due to tweeter/cone interference, and the 14kHz dip is due to the cone/baffle interface.
These are not easy to get rid of and are a consequence of the concentric process. There are some designs that have mostly eliminated some of them, but that process can introduce other problems.
As you show, the most severe dips even out off-axis. The 2.8kHz evens out in the vertical direction. As a result, I always design to balance the overall power response with the axial response rather than optimize for one parameter only.
Through experience, I have found the best compromise. I could easily, with DSP, equalize any response curve I want to look very flat. This never sounds the best solution.
I also never listen to any speaker directly on-axis. This is always the point of maximum symmetry and worst-case frequency response due to diffraction, even with regular non-concentric speakers.
Measures that can be taken to reduce diffraction, such as very heavily radiused cabinet edges (2” radii for example) almost never look good and also considerably add to the cost.
The secret to speaker design, especially low cost or very high value, is applying the best compromise between necessity and cost.
Thanks once again for a great review.
All the best
Andrew