Ah okay, now I think I get what you meant, as you were talking about there being either a time offset or an absolute phase difference between overlapping driver outputs!?
Sometimes I’m a bonehead, I guess, but am I right to surmise, that’ll work only with a cyclical signal having an identical amplitude envelope above and below the 0-line, but not with a irregular musical signal?
Indeed.
We have to bear in mind that an irregular musical signal will include regular cyclical signals from the plucking of strings on a guitar, the hammering of strings on a piano or the resonances from a trumpet or saxophone or the vibration of the skin on a drum. All sinusoidal fundamentals with a range of harmonics!
Some more measurements now that I’ve settled on a final (for now) set of filters.
This is the raw response after driver linearisation:
On reflection, the midrange amplifier could be dropped to medium gain; this would reduce the overall system gain reduction and keep a bit more voltage output on the Motu.
Interestingly, the “Target” curve is Uli Brüggemann’s target curve which if the mid range gain was reduced by ~6 dB would closely approximate the natural response curve of the speaker in the room. (The drivers are linearised nearfield to a quasi-anechoic flat response). There’s much discussion about “target” curves and some who oppose them. Dr. Floyd Toole’s thinking is that they should be more appropriately considered “result” curves rather than target curves as a speaker tuned to a flat anechoic on-axis response will result in in an in-room response with a bass boost and a falling frequency response curve. All of the common target curves follow this more or less. Note that the curves extend to 24 kHz, so the taper in the HF response in the Brüggemann curve isn’t as pronounced as it first appears:
Frequency Response, L&R, Mono and target curve (a 16 Hz subsonic filter was added at the filter generation stage):
Overall, I’m fairly happy, pretty much ±2 dB. The dip in the left at ~100 Hz is due to room asymmetry and attempts to fix it caused more problems than they solved. In the mono, it’s less noticeable. Bass is generally mono below about 200 Hz anyway.
Step response, this is where zero delay plane (ZDP) time alignment and phase linearisation really shines. All drivers are precisely aligned:
There’s a small amount of pre-ringing in the left channel due to the inter-channel phase alignment adjustment, however the worst of it is within 10 ms of the step, so it’s well within the psychoacoustic masking window (20 ms). It’s not audible on any tracks I’ve tested to try to bring it out.
ICCC, or IACC (Interaural Coherence Coefficient) as it’s also known is fairly respectable. The ICCC10 is just over 70%. With a more symmetric room layout (my listening room is the lounge), that would significantly improve.
My room’s RT60 is pretty good, due to the room treatments applied, so further treatment might help the ICCC, but conversely could make the room sound too dead:
Overall, the conversion to a fully active setup with Acourate filters has been a great success. Bass is tighter, the soundstage is much larger and extends left and right beyond the speakers. Depth perception is much greater, both in front of and behind the speakers and the overall sense of realism is just exceptional. I’ve put in more sofa hours in the last few days than I have in the past few months. I just want to sit and listen to music all day every day.
Linear phase speakers are pretty rare - the only commercially available system I’m aware of at present is from Lundahl Sound Systems. I use Karl’s SV2361 volume control before my power amps.
Here’s a link to a paper written on linear phase:
Are these FR measurements time-windowed to simulate quasi-anechoic environment, or diffuse field measurements in the room? What looks unusual to me is the decline in level beginning at 2K towards higher frequencies. While the beginning can be explained with the midrange driver´s expected tendency, it is not clear why the tweeter keeps losing level >4K.
Did you make a subjective listening test comparing the target curve used with a fairly flat response between 500 and 8K?
There is indeed, but you cannot see diffuse field target curves in a room separated from a speaker´s directivity index and the room´s RT60 behavior particularly >2K. Which in your case should not pose significant problems as the RT60 in general is moderate which seemingly efficient midrange absorption (particularly 300-1,100 Hz which is unusual).
What Dr. Toole is calling a decreasing ´result curve´ is indicative of either the speaker´s increasing directivity index towards frequencies higher than 2K (or overdamped treble in the room but that we can rule out in your case). In studio-grade rooms there have always been attempts to counter this behavior by damping midrange and lower midrange more intensely or keeping indirect sound level anyways very low.
For home use this decreasing curve seemingly was accepted by many as inevitable for quite a long time. This has changed with a controversy ongoing for roughly a decade since many technically advanced loudspeaker manufacturers have undertaken remarkable effort to keep the directivity index constant (and high enough) from lower midrange on.
Except from Queen´s `Breakthrough´ intro 
They’re measured at the listening position with FDW applied. The sofa is very close to the rear wall, so there’s absorption on the rear wall and there’s also a combination of diffusion/absorption on the ceiling to reduce the effect of first reflections. Without the treatments, the room sounds horribly live and the RT60 is unacceptably high.
I’ve tried a flatter response in the region between 500 and 8k before and I really don’t like it - it sounds far too “bright” to my ears.
The downward slope from 1 kHz to 20 kHz is ~6 dB and there’s only 4 dB drop between 1 kHz and 10 kHz. It’s more than the Harman/Toole curves, but fairly close to the HATS curve. My listening distance is about 2.7 m from each speaker.
That would be an interesting thing to investigate where it is actually coming from. If you would have steadily increasing D.I. of the speaker, a linearized diffuse field curve might appear bright because it actually is brilliance-rich on axis. But something tells me this is not the case with your setup, and 2.7m of listening distance with 0.3-0.4s of RT60 in the critical range is pretty close to studio standards with the influence of indirect soundfield being anyways low.
That is also interesting. Did the room show any noticeable increase in RT60 towards higher frequencies in typical brillance range before any treatment was applied? The increase in RT60 above 2K is not dramatic after treatment but there might lie an explanation for what you have noticed.
I personally tend to ignore everything above 8K as it can easily be EQed if the decrease in diffuse sound level is not dramatic.
Would rather be concerned about the decrease between 1 and 10K as I do see neither any obvious reason nor necessity for that. RT60 is fairly the same. Maybe tweeter directivity is an explanation as you are using a softdome Scanspeak model. Some of these tend to have a ´decoupled diaphragm center´ thanks to very high inner damping in order to avoid out-of-phase interference. The downside of such a diaphragm is some ring-radiator-like behaviour leading to a steep increase in D.I. between 6 and 12k.
If that theory applies to your setup the brightness you have encountered with flat FR would be visible in an anechoic measurement but no in-room target curve would fit as you have a step in directivity index in the aforementioned frequency band. I would recommend to do the whole EQing of the tweeter region by ear.
I’m posting this here as I don’t want to hijack the whisk(e)y thread.
Having returned by car on Friday from a week’s holiday in Scotland where we saw my Dad at home then made frequent visits to see him at the Beatson Cancer Centre in Glasgow, I found myself back on a flight to Glasgow yesterday morning.
The team there are still trying to put the pieces of a puzzle together, though the feeling is that his course of radiotherapy unearthed “dormant” infection from his original surgery.
Sadly, my dear Dad passed away this afternoon.
It’s been a bittersweet day. A huge loss that leaves an enormous vacuum in our lives, countered by hours of smiles and laughter as we reminisced about our fondest memories and recounted some hilarious stories of our lives with him amongst us.
Somewhat fittingly, here’s the last photo I have of us together at Johnnie Walker Princes Street in Edinburgh at the end of August 2024.
Here’s to you, Dad
My most sincere condolences. 
Hold tight to those precious memories.
Condolences. May he rest in peace…
Condolences @Graeme_Finlayson.
So very sorry to hear Graeme.
What a wonderful photograph, one to treasure. My thoughts are with you.
I am so sorry, Graeme, my sincerest condolences. Take care.
Sincere condolences on your sad loss. A very painful time for you and yours.
Very sad news Graeme and my condolences to all your family
Very sorry Graeme, and what a lovely photo!
Hi Graeme, I am very sorry for your loss.
My condolence to go and all of your family.
P.S. This is a lovely last photo of you and your father.
Hello Graeme, my profound condolences to you and your family. I love the photo of the two of you – a great memory.
I will pour a dram in honour of both of you tonight.
Nathan
Vancouver
Sad News. My condolences to you and your family.
Oh Graeme. I am so sorry. I am raising a wee dram in memory of you Dad. Thoughts of you are with me.
