That is only part of it. Getitng the drivers that close makes them essentialy co-incident making it much easier to do the XO and eliminating many of the “evils” of XOs.
So the answer to the OP’s query is, “As close as you can manage”
EVERY loudspeaker is riddled with compromises, since you have chosen your HF driver your have to chose compromises that fit around its use.
dave
You've now picked an extreme example that applies only to omni speakers (or speakers where you're looking for a specific non-interaction with the baffle edges). I guess, to your point, if you completely restrict the directivity of one of those drivers before you get to that off-axis angle, you're going to get sound from the other one and therefore won't have lobing but only beaming. Still, in most cases I don't see this as being possible, or even necessarily as an overall advantage.
I'm finding tightening acoustic centers on multi-ways takes SQ to a level i haven't heard before.
It's like it has a sound of it's own...so focused...much like electrostatics or planars.
I think there is more to 1/4 WL spacing, than just the trig that calculates how the signal arrives at listening position,.... lobing or not, etc.
I'm hoping to measure for driver-to-driver interactions on-axis, in terms of W/L spacing, to see if any if response variations exist.. ...dead on, centered, on-axis.
Beginning to think they do...
It's like it has a sound of it's own...so focused...much like electrostatics or planars.
I think there is more to 1/4 WL spacing, than just the trig that calculates how the signal arrives at listening position,.... lobing or not, etc.
I'm hoping to measure for driver-to-driver interactions on-axis, in terms of W/L spacing, to see if any if response variations exist.. ...dead on, centered, on-axis.
Beginning to think they do...
Lobing or beaming, can't they be considered (into a cross) in the same broad way?
We could discuss this over any scenario, either ideal or realistic. It could get rather lengthy and go beyond the scope of this thread. In this case I considered an ideal scenario to hone in on the specific point in question. I chose the most obvious example to explain my point. I wasn't thinking omnidirectional specifically.
I suspect those with certain speakers in certain rooms, and who focus primarily on frequency response may find success sooner with closer spacing.
Well said James. Your picture reminds me of this one from Taylor:
https://p10hifi.net/TLS/downloads/taylor-line-array.pdf
dave
This reminds me of the things I saw when I modelled a slot loaded source in this java based particle/wave simulator thing that someone shared a link on my thread but I cannot find it anymore....
They’re both phenomena of sound wave radiation, but they are pretty different for our purposes. Lobing is from the interaction of, and interference with, multiple drivers or physical sound sources, while beaming is from the interference of one sound source with itself.
Hello,
Realize you are having fun with this, but you might be interested in reading info on the 4PI Speaker disign given what speakers you said you had available.
I have built them and they sound very good to myself and everyone who has heard them. FWIW. Guy has lots of info on his site as to why he choose and using the design choices he made.
Pi Speakers - four Pi loudspeaker performance data
and articles at the bottom of this page: Pi Speakers - Technologies
Regards,
Greg
Realize you are having fun with this, but you might be interested in reading info on the 4PI Speaker disign given what speakers you said you had available.
I have built them and they sound very good to myself and everyone who has heard them. FWIW. Guy has lots of info on his site as to why he choose and using the design choices he made.
Pi Speakers - four Pi loudspeaker performance data
and articles at the bottom of this page: Pi Speakers - Technologies
Regards,
Greg
I have been thinking about this quite a bit lately, trying to design a system as close 1/4 wavelength spacing as able. It is quite difficult. Obviously the biggest challenge is mid-tweeter spacing.
A tweeter that can handle a low crossover seems to be the first requirement, then a small mid. Unfortunately, small mids (except maybe the Faital 3Fe22) are very low sensitivity.
This in an interesting applet to play with re. spacing and visualizing the lobes:
Sound Wave Interference Applet
Bill
A tweeter that can handle a low crossover seems to be the first requirement, then a small mid. Unfortunately, small mids (except maybe the Faital 3Fe22) are very low sensitivity.
This in an interesting applet to play with re. spacing and visualizing the lobes:
Sound Wave Interference Applet
Bill
Thanks, Greg. I've actually built the 4pi speakers and really enjoyed them. I eventually gifted them to my brother.
Just wanted to try a different waveguide. Many fellow forum members had postive remarks about the QSC wageguide. Considering I had a pair on hand, along with a pair of JBL 2226H - wanted to see if I could pair the two.
I also have a pair of brand new, still in box, Eminence Delta Pro 12A (i.e., a 12" driver). Perhaps I should try using that instead.
All of our 2-ways have 1/4 wave spacing (or very close). You have to use a midTweter, which the Faital you mention sounds like i thas potential — i haven’t yet EnABLed mine, so it still lags behind the others we have used.
One does pay the price of lower efficiency.
The other option is a coax, factory, or self made, but the higher frequencies involved mean some sort of very carefully calibrated time delay to achieve the last little bit of spacing to reach the 1/4 wavelength spacing. Or use a synergy horn. A 300 Hz should get you close.
dave
An additional piece of my design exercise (in addition to the 1/4 wavelength requirement) was to only use drivers in the range where ka </= 1, where ka= 2pi(f)(a)/c or the frequency where ka=1 is f=c/2pi x r......this ratchets it up further 
Obviously the ka requirement is near impossible for the tweeter without increasing the number of "ways," and is likely not worth the added complexity, so I was thinking of beryllium. Using the Bliesma T34B would allow an XO of 1.3khz, this helping with mid selection.
Drivers I have looked at with their frequency where ka=1
SS 5F 2481Hz
SBA SB65 2166Hz
Accuton C50-8-044 1,985Hz
C51-6-286 CELL 1,978Hz
Faital 3FE22 1,761Hz
SS 10F/4424 1,610Hz
Faital 4Fe22 1,396Hz
Wavecor WF120BD03 1,315Hz
Obviously the ka requirement is near impossible for the tweeter without increasing the number of "ways," and is likely not worth the added complexity, so I was thinking of beryllium. Using the Bliesma T34B would allow an XO of 1.3khz, this helping with mid selection.
Drivers I have looked at with their frequency where ka=1
SS 5F 2481Hz
SBA SB65 2166Hz
Accuton C50-8-044 1,985Hz
C51-6-286 CELL 1,978Hz
Faital 3FE22 1,761Hz
SS 10F/4424 1,610Hz
Faital 4Fe22 1,396Hz
Wavecor WF120BD03 1,315Hz
An additional piece of my design exercise (in addition to the 1/4 wavelength requirement) was to only use drivers in the range where ka </= 1, where ka= 2pi(f)(a)/c or the frequency where ka=1 is f=c/2pi x r......this ratchets it up further
Unfortunately that metric assumes a flat cone and does not take into consideration cone shape or the radiation of a dustcap, so is, in the real world largely not all that useful.
dave
What does this physically give you?An additional piece of my design exercise (in addition to the 1/4 wavelength requirement) was to only use drivers in the range where ka </= 1, where ka= 2pi(f)(a)/c or the frequency where ka=1 is f=c/2pi x r......this ratchets it up further
Obviously the ka requirement is near impossible for the tweeter without increasing the number of "ways," and is likely not worth the added complexity, so I was thinking of beryllium. Using the Bliesma T34B would allow an XO of 1.3khz, this helping with mid selection.
Drivers I have looked at with their frequency where ka=1
SS 5F 2481Hz
SBA SB65 2166Hz
Accuton C50-8-044 1,985Hz
C51-6-286 CELL 1,978Hz
Faital 3FE22 1,761Hz
SS 10F/4424 1,610Hz
Faital 4Fe22 1,396Hz
Wavecor WF120BD03 1,315Hz