Originally Posted by LTD02
"I'd say that looks pretty damn good. The reason his on-axis measurements looks the way it does (climbs slightly from upper bass to the crossover region and then tapers off slowly) is due to him optimizing the power response and not the on-axis response."
i agree but the horn is too small for the driver. he should have a nominal 15" horn in there. one of the seos might improve things, drop the crossover point, hold pattern control, and lower the crossover region.
I disagree that it needs a nominal 15" wide horn. The H290 doesn't have significant roundovers and should hold pattern low enough for his design...just barely.
Although I agree the SEOS is a better horn/waveguide (disclaimer: I contributed to its design) the difference is relatively small when compared to the difference between a well-designed crossover and poorly designed one.
Frankly, IMO the biggest reason people complain about horns (or any speaker for that matter) is poorly designed power response. Parham understands this and it shows in his measurements. I'd guess that most people who slap together DSP based crossovers think they've eliminated the tough part by getting rid of the need for modeling and passive components. The problem is that the real "magic" is in understanding what the target response should look like. Flat on-axis is not what you want. It is dependent upon the whole system. Some people think this is voicing by ear but in fact it is measurable.
In a nutshell, I see no flaws in the 3pi/4pi designs. Of course, there are always potential refinements (better components, roundovers, etc) and different approaches (dipole vs horn, 2-way vs 3-way, etc) but a well designed speaker is worth its weight and all other differences are on the fringe.