If the measurements are right, it looks to me like there are at least three things happening.
First is the midrange doesn't look attenuated to match the other drivers. So I would look at the R4/R5 resistors. Make sure they're installed correctly and not shorted. That's a kind of easy problem to solve because it's just attenuation.
Second, the 300Hz peak. Something is going on there, probably a panel resonance. You might want to damp panels one at a time with a pillow or your hand and see if you can identify the cause. That one could be a measuring artifact too, so move the speakers and microphone and measure again. Best to measure outside. But I suspect that one is a panel resonance.
Third, the lack of treble, and the overall downward sloping trend. I see that a lot with measurement problems. Sometimes I see it with microphones overdriving the input amplifier. One would think the extra harmonics would artificially show excess HF, but I often see that problem cause a downward slope. Or it could be something else in the measurement system. But this could be confirmed by measuring something else and seeing if the downward trend existed there too. If it's truly the speakers, then the tweeter circuit or compression driver is off.
I didn't find any image files in your archive, so I can't see your response curves. I run WTPro and LMS, so the REW data files aren't useful for me. But even if there is no real 300Hz anomaly, there is still something way off. The midrange is much higher than the treble, telling me either the compression driver or the crossover is at fault. If you would like, send the crossovers to me and I'll check them out. We'll get to the bottom of it.
compaddict Messages: 82 Registered: April 2014 Location: Auburn, CA USA
Whole set of new files with corrected SC file.
Both speakers tested 1M and flat as well as some B&W DM 330 cheapies for reference.
Two files renamed for clarity.
REW is the SW used.
Again, we see that there is a huge rolloff up high. It looks like the compression driver tweeter isn't even connected. That's where we need to focus, I think. Concentrate on what may be causing the lack of HF output, whether it be bad crossover, wiring or compression driver.
I think he played them without the extra (or any?) front gaskets.
Could the driver 'bottom' in such a situation? And then affect freq resp?
The midhorn gaskets aren't needed unless output is really high. And by "high" - I mean way past the level anyone would feel comfortable listening.
If pushed that hard, then the driver is pushed past its limits. One can go past xmax with those drivers without damage, but the sound quality plummets. So for hifi applications, an extra gasket isn't needed.
Besides, what we're seeing is a lack of HF content. I think that's where we need to concentrate.
At first, I thought we might have some sort of midhorn panel resonance because initial reports were that voices sounded distorted. I didn't even think the crossover was suspect, because I thought we were dealing with a buzz or something like that.
But after seeing measurements, I think just the opposite. I now think the problem is the compression driver or tweeter circuit. Measurements are showing that there just isn't any tweeter output.