5 way horn speaker system project - tapped, bass, mids and tweeter passive active

I think there is one last hurra left in the 5 way (mostly front loaded) horn system.

The idea would be top down.

-Raal lazy ribbon tweeters (13000Hz up)
-Vitavox S2s on Le Cléac'h horns (800Hz to 13000)
-JBL 2482s on Tractrix T200 horns (350Hz to 800)
-LF provided by the SH50s (sub 100Hz somewhere to 350)
-Tapped horns (20Hz to sub 100)

similar to this but with the Tractrix T200 doing mids..

IMG_20240321_213014.jpg


Before I start moving lots of gear and making stands to support the horns in the best positions (centres as close as possible)..

One question for @Cask05.
Would the strategy of no, or the odd 1st order XO here and there, PEQ to trim off unwanted freqs, and with time delays to optimize phase, be an idea to try?

It's the only thing I've not tried with those horns.. kinda curious as to how it might sound.

I could rest it in peace then😂


-
 
One question for @Cask05.
Would the strategy of no, or the odd 1st order XO here and there, PEQ to trim off unwanted freqs, and with time delays to optimize phase, be an idea to try?
What is occurring in the "zeroth order" (PEQ-only) crossovers is that they are eliminating the all-pass phase growth of passive electrical filter networks or canned IIR crossover filters.

In loudspeakers not having "unity summation apertures" (i.e., non-MEHs), what you're doing with the PEQ-only zeroth order crossovers is avoiding some of the problems of not having unity acoustic summation across the multiple drivers in the loudspeaker. In multiway loudspeakers mounted on flat baffles having multiple apertures, the output simply sounds better--more cohesive, among the other effects. Some of these subjective effects are enumerated here:
  • Wider Soundstage
  • Greater Perceived Depth of Soundstage Image
  • Increased Perception of Bass Extension
  • Tighter Perceived Bass
  • Elimination of Harshness of Sound in the Upper Registers
  • Greater Perceived Realism and Naturalness
  • Disappearance of Loudspeaker Localization In-Room (i.e., can no longer localize the exact location of a loudspeaker with eyes shut)
  • Greater Perception of Music Resolution
So all of these effects are related to phase distortion in loudspeakers--or the absence of the associated defects in loudspeaker performance if the all-pass phase shifts are eliminated). That's what you can get with using zeroth-order crossovers--even in multiple aperture multiway loudspeakers (assuming time alignment of the multiple drivers drivers is achieved on-axis).

While I do not really condone multiple-aperture loudspeakers nowadays (especially those whose driver centerlines are separated by more than 1/4 wavelength at crossover points), simply eliminating the all-pass phase growth problems on these type of loudspeakers will more or less achieve some or all of the enumerated improvements listed above, to the degree that unity summation of driver output is achieved or not achieved by the loudspeaker's spread-out driver centerlines, the composite loudspeaker directivity control reaches down to the room's Schroeder frequency, and the listening room itself is treated for early reflections around the loudspeakers and listeners' ears.

Chris
 
  • Like
Reactions: 1 users
Thanks Chris,
Excellent explanation and confirmation.

This will be a project for next winter I think.

It will be interesting to hear, but I really struggled with the sound of the Le Cléac'h atop the SH50, that distance driver centre to centre is obvious once you've heard otherwise.
No way to get less than 1/4 wavelength distances there.
Esp sitting fairly close in smaller hifi room situation.

I'm still immersing myself in the Synergy sound. Really liking it.
 
  • Like
Reactions: 1 user
What is occurring in the "zeroth order" (PEQ-only) crossovers is that they are eliminating the all-pass phase growth of passive electrical filter networks or canned IIR crossover filters.

Hey Chris, got a suggestion/request...

My understanding is you have or had, a SH-50.
If correct, I figure you no doubt made transfer function measurements of the passive xover outputs.

Could you post those?...the electrical outputs' transfers? ..(not the drivers' acoustic outputs.)
Maybe you have, and sorry if i missed/can't find them.

And then also post your "zeroth-order" filter set that replicates those electrical transfer functions?
Simple comparison, and very instructional for all i think.
 
Could you post those?...the electrical outputs' transfers? ..(not the drivers' acoustic outputs.)
Since this subject actually has little or nothing to do with the subject of this thread, this will be the only response on this subject I'll make here (to avoid added thread dilution):

The data I measured were acoustic measurements (using REW) of the combined output, and the individual measurements of the three "ways"--with and without passive networks in-line. That occurred in the summer and fall of 2018--after I was obliged to pay $3250 USD for a new Danley SH-50 (still in my possession). Unfortunately, in May 2020, those data were lost in an external HD failure. In effect, I paid for the full "MEH 601" course tuition and donated significant time and energy to the endeavor, absorbing what I found during extensive testing. I think I achieved more than a passing grade for that work.

Since then, I've had little interest in doing that exercise again to simply replace that information. What I need to know is still in my head. I don't need to repeat that learning experience for posterity. I learned what I had intended to learn and applied those lessons to the K-402-MEH and also all other DSP settings in my setup, and in helping others dial-in their loudspeakers (remotely) with minimal all-pass phase growth.

I also had to "unlearn" a lot that I had inadvertently picked up along the way that wasn't really useful to getting better sound quality. That's probably the most important part of that exercise that was learned. I recommend that exercise highly for those that thirst for that knowledge.

Here's why I haven't posted the results of those measurements: because it just doesn't feel right to do so without others experiencing anywhere near a full understanding of what is actually there--and most importantly, why it's there. I learned a great deal...but it also cost me to get that information.

Since then, I found and introduced what I call a "zeroth order" crossover filter approach to achieve the same type of phase response as the SH-50--but for loudspeakers not of the MEH type, but rather of the multiple aperture/multiway type (but still having full range directivity). The subjective results were astounding. I believe that what I learned should be in new chapters rewriting existing "DIY loudspeaker" books, and I've been writing about that subject since then.

I've seen the effects of putting out a little information "free of charge" and experiencing getting bitten (quite viciously) for the favor. I don't have any interest in doing it again.

Chris
 
  • Like
Reactions: 1 user
Chris, thanks for the reply and explanations.
Sorry to hear of the data loss....I've had a few trips down that bad lane too....

I see you took the 3 ways' acoustic measurements both raw and with the passive xover in place. Nice.

I think what's missing, especially given all the discussion on MEH crossovers, are transfer functions of the passives' 3 outputs.
First, it would shine great light on what filters the passive xover is using.

And seems to me, it would be the definitive way for you to demonstrate your "zeroeth-order" technique can indeed replicate what the passive is doing.
It would be a measurement based..... no-if-ands-or-buts ..... proof (or possibly not) of the technique's viability. (albeit just for the SH-50)

So, as relatively easy as it would be to take transfers of the 3-ways electrical outputs, and work on reconstructing them with DSP.......
I think it would be immensely appreciated by all. And perhaps create conclusive evidence for your recommendations.

I've been thinking the same thing about my complementary linear-phase xover technique.... that perhaps more than continually writing about it......
I should simply post a thread that starts with raw measurements, and then shows the steps that are taken and the measurements of those steps, to get to a full processed reponse.

Anyway, best, Mark
 
  • Like
Reactions: 2 users