My thoughts are just one man's opinion. I'm always learning and making incremental improvements as I do.

I'm really interested in your thoughts. At low frequency, a horn can easily (and accurately, I think) be modeled with simple 1P formulas. But at higher frequency, where the diaphragm doesn't move as a piston, where horn features become acoustically large, etc, that's where an empirical approach has always been required. Any ideas on some of those thorny issues? I'm not sure we can make a closed form solution for them, seems like they can't be solved without complex modeing techniques. I guess that really takes us back to the FEA / BEM thread, but these topics are sort of related.