So guys, I'm finally satisfied with the crossover after a few different iterations. It's still 6th order centered around the same area, although just about everything is different about it this time around. I doubt it'd be obvious from the graph so I won't bore you with the details.
One thing I found from listening is that as simulated, they sounded great and neutral already. However on brass and percussion, I found them to just sound that "little bit" more realistic with the tweeter a decibel louder than the previous model. A few reasons this might be the case are:
1) Our ears are probably ever-so-slightly off the tweeter axis, causing some ever-so-slight attenuation
2) It's still pretty damn flat!
3) Baffle step losses may only be about 5db instead of 6db and there's actually some broad baffle diffraction gain below the crossover region that can't be modelled accurately
4) this room has its share lot of HF absorptive surfaces including couches galore
So given that tweeter level adjustment, here's what I end up with:
The reverse null (at the microphone position) looks satisfactory too and shows that the crossover point is almost
exactly 850hz while maintaining inter-driver phase tracking:
We adjusted their placement to be much wider apart than they were earlier (we'll figure out cable management later. Everything here is a work in progress) Also those big batts of rock wool on the front wall are to be covered in cloth):
So, my impressions?
First of all, these speakers do require
some distance for the drivers to properly converge into a single apparent source. (I would say over 2m minimum) Once you've gotten far enough away though, driver integration is perfect. So here I go:
- these speakers throw a
massive soundstage. It's extremely wide, and deep. Perhaps not as
infinitely deep as my open-back Philharmonics, but deep enough that there is very obvious front-to-back layering. It never feels artificially "enlarged" however, and vertically they avoid that awkward "mouth the size of a room" feeling I've heard from other loudspeakers (which I've "possibly" attributed to the effects of odd-order crossovers - this is one of the few speakers of this type that uses
even-order slopes and that was very intentional)
- imaging is both pin-point and diffuse as recordings call for. When you close your eyes, the speakers simply disappear in my opinion. And I haven't even gotten the SEOS-baffle on yet, which in theory will allow the wavefront a better termination.
- yes, the ~14.5 inch paper cone of the TD15M does occasionally "smooth out" a bit of the midrange character ever-so-slightly compared to the probably "flawless" razor-sharp mids of the 8" SB woofer and BG Neo8 planar in my Philharmonics. But make no mistake, it's an extremely resolving and otherwise natural sound... i wouldn't be shocked if in a room full of 100 people, fifty pick the smoother TD15M sound and fifty pick the "quieter" Neo8 sound as sounding more "accurate". In this particular case, I do personally prefer the Neo8 sound. It's also possible that parts of what I'm hearing could be attributed to the tweeter, whose 2nd order harmonic distortion rises below 1khz (which perceptual research shows
is not an actual issue even if it's audible)
- the BA 750 is an amazing tweeter. It required a lot of active response shaping for me to get what I wanted out of it - I'm not sure it'd be worthwhile to use it passively if ruler flat frequency response is the end-goal. I had to boost its response a bit in the crossover region to get my desired transfer function and I imagine anyone letting the driver speak to it would find its response suggesting a crossover point an octave higher. But that would have defeated the purpose of this whole design, between avoiding driver breakup, using the compression driver's low frequency advantage, and getting the desires off-axis response.
So how do the driver's highs sound? Beautiful. Cymbals decay instantly and have a feathery nature to them. Xylophones sound organic and aren't missing a thing. Flutes, Violins, Violas... everything sounds excellent to my ears. Anyone trying to convince you compression drivers are not high fidelity... is just wrong. Now, does the RAAL in my Phils have a better top octave? Yes, I actually believe it does. But do I care?
I'm not sure I do. It's true that sometimes we get caught up overthinking drivers in the frequency extremes. I've stated in the past that even though my sub is ruler flat to below 20hz, I could high pass my system at 100hz and still be really damn happy. It's the same thing here. The top octave on the BA is inoffensive, and everything else sounds clean, sharp, all that good stuff. It's never fatuiging to listen to, and don't even worry for a moment about preconceived notions of horn honk. There were actually some instances today where certain high-pitched voices in something we were watching on TV made me "prepare to cringe" and then I was just left stunned at how I didn't flinch at all. I think an honest argument COULD be made that this driver sounds as good as any out there below 10khz. That said I made sure to notch out a low-amplitude, but very-high Q peak centered at 8700hz (in my crossover design and I'd hope anyone working with this driver keeps that in mind.
- the lack of power compression sounds incredible. I've never thought of my Phils as being difficult to listen to loud -
and I still don't - but these speakers are
easier to listen to on very dynamic source material. The narrow directivity and high efficiency, as well as insignificant driver motion all lead to an amazing effect that has to be heard to be understood.
- What does all of the above mean? These speakers kick hella ass and had me grinning from ear to ear.
