I take back relatively 'easy-peasy' with the SH-50's low-order xover .......Lol
Yikes! what a piece of gear. Makes me wanna run away from passives even faster (if such were possible haha)
I've seen that the SH-50 can be ordered active, and with presets provided via their processors, or DSP amps.
I haven't looked in a while, but this thread has me wanting to revisit what the active presets are. I'm going to pull out the Danley ASC-48 processor I have from the days I was trying to see if getting involved with live sound was a good idea. Update firmware, software, and preset library...plus I see it now has Q-Sys integration which I'm already thinking of ways to use.
In addition, if I see a decent deal on a used SH-50, I think I'm going to grab it....to dissect. Wanted to do this forever.
I want to take transfer functions of the crossover outputs, and make them the targets for equivalent FIR files. I'd like to compare....both the digital replication of the passive xover, and then also my own standard high-order lin phase xover method.
Question for you...when taking a transfer function of a passive xover, with the goal I have in mind of digitally replicating it, do you take the transfer with the drivers connect to xover? I would think so, but not sure here....out of my depth.
Many more state side than over here.
If you buy a SH50 digital version 2nd hand, I suppose the XO 'rights' transfers with them?
That would be interesting to examine!
Not sure what you mean by 'digital version' ?
If you mean do you get rights to the settings in the DSP processor to freely give or share....well, I don't care what the law says, but that is not anything I'd ever do.
I think you need to own the speaker to have rights to the XO and presets, and the rights stay with the speaker. (just my 2c on what feels ethical to me...)
The SH-50 version I want is the regular passive xover version....again to dissect for the purpose of learning.
But same thing would/will hold with however I can copy the passive digitally....if I can build a good set of replicating FIR files, I would not view that they are mine to give out. Be happy to discuss how the xovers work generally...but that's about the line I think.
Besides, I don't want the passive xo or a digital replica or it....I want to try my own hand on the speaker...THAT i could give out
Ok, with the risk of being totally wrong, here is my two cents.
Sometimes these discussions get really confusing and tecnically detailed so I get lost. What is the goal, what are we aiming for from a theoretical viewpoint? And how do we get there in practice, no mystics and guessing.
I am NOT an educated engineer, but have been reading about synergy horns for a long time now and am very slowly building my own. I have this idea of what a REAL synergy-horn is, given it is big enough to be "fullrange":
It is as close as you can get to a "perfect" fullrange-pointsource-loudspeaker-driver that reproduces the musik as true to the original signal as possible. Minimum-phase as a perfect dirac-pulse in the audible spectrum. Reproducing natural sound! Like a perfect fullrange-driver without a cross-over!
This is what I think is the goal Tom Danley follows, if I have understood him right. Like when he mentions Richard Hayser and his research.
And here is what I want to point at with this:
I do not think it is possible to make a "real" synergy-horn with the "synergy-magic" with higher order filters, because you introduce excess phase per definition!
I don't follow all this third order or fourth order acoustic filter slopes-talk. A filter is a filter regardles if acoustically or electrically or combined and introduces phase-rotation, or what? Am I totally wrong here? No offense intended!
I may be mistaken, but I don't think we can do better than a minimum-phase-device for soundreproduction!? That is the goal, and as I understand things, Chris approaches it in his way in the IIR domain and Mark in the FIR domain.
Well, I hope this comes through as it is intended, as a call to relate soundreproduction to physics/theory. I am on deep water here!
And thanks to all for participating in these threads. I learn a lot.
Regards
Steffen
Sometimes these discussions get really confusing and tecnically detailed so I get lost. What is the goal, what are we aiming for from a theoretical viewpoint? And how do we get there in practice, no mystics and guessing.
I am NOT an educated engineer, but have been reading about synergy horns for a long time now and am very slowly building my own. I have this idea of what a REAL synergy-horn is, given it is big enough to be "fullrange":
It is as close as you can get to a "perfect" fullrange-pointsource-loudspeaker-driver that reproduces the musik as true to the original signal as possible. Minimum-phase as a perfect dirac-pulse in the audible spectrum. Reproducing natural sound! Like a perfect fullrange-driver without a cross-over!
This is what I think is the goal Tom Danley follows, if I have understood him right. Like when he mentions Richard Hayser and his research.
And here is what I want to point at with this:
I do not think it is possible to make a "real" synergy-horn with the "synergy-magic" with higher order filters, because you introduce excess phase per definition!
I don't follow all this third order or fourth order acoustic filter slopes-talk. A filter is a filter regardles if acoustically or electrically or combined and introduces phase-rotation, or what? Am I totally wrong here? No offense intended!
I may be mistaken, but I don't think we can do better than a minimum-phase-device for soundreproduction!? That is the goal, and as I understand things, Chris approaches it in his way in the IIR domain and Mark in the FIR domain.
Well, I hope this comes through as it is intended, as a call to relate soundreproduction to physics/theory. I am on deep water here!
And thanks to all for participating in these threads. I learn a lot.
Regards
Steffen
Last edited:
Hi Steffen, that is 100% correct....that high-order crossovers introduce excess phase rotations..... for conventional IIR crossovers.
That is why speaker designers have traditionally strived for lower order xovers.
However, complementary linear phase xovers do not introduce any phase rotations, at any order.
They introduce excess fixed time latency, like time of flight, instead.
Btw, a quick look at the latest set Danley presets for their DNA SC-48 processor, shows five Synergy models that when biamped use linear phase hp/lp.
The SH and SM models using lin phase, range in horn patterns from narrow to wide. Sh-50 doesn't.
Hi Speedysteve,
Maybe post your raw measurements of bass, mid, tweet so that people can see what you are starting with. I know for sure my 4" celestion mids on a 50x50 horn the same size as the SH50 had approx 6th order acoustic slopes at both ends (raw measurement no EQ). Not even close to 10dB/octave .
Maybe post your raw measurements of bass, mid, tweet so that people can see what you are starting with. I know for sure my 4" celestion mids on a 50x50 horn the same size as the SH50 had approx 6th order acoustic slopes at both ends (raw measurement no EQ). Not even close to 10dB/octave .
As Tom Danley wrote in 2012 regarding the SH-50:
https://www.whatsbestforum.com/threads/question-about-danley-sound-labs-50s-or-60s.5738/
"with a passive crossover, there is a practical limit to how many of the response features you can address, each usually requires 3 passive parts, all sized for stability up to high power operation."
The SH-50 passive crossover, as complicated as it is, obviously does not come very close to the goal of a flat phase and frequency response.
Reading the SH-50 manual, found the SH-96HO (4 – 15” woofers, 6 – 4” mid frequency drivers and 1 x 1.4” high frequency driver) can be bi-amped between the woofers and the passive mid-high section:
"The high/mid section is high passed at 400 Hz at 24 dB Butterworth."
"In the BIAMP setting, there is no crossover applied to the four 15 inch drivers!
You are strongly advised to use a crossover set to high pass of 40 Hz at 24 dB Butterworth and 600 Hz at 12 dB Butterworth. "
Tom also writes: "The complication here is that when you measure what DSP units do after specifying filter X and eq Y, they are nearly all somewhat different and so specifying the filter type, frequency and slope etc are actually the least reliable way of programming one, measuring the output the most.."
Yes. The crossover requires driver termination.
Read across the driver terminals and simultaneously across the input to the crossover.
The comparison (difference) of the two measurements is the transfer function of the crossover itself.
The transfer function of the crossover does no include the time constant of the physical offset, which may partially compensate for the acoustic and electrical filters delays.
Tom also wrote: "In the development of the Synergy horns, there were many small steps before they measured / acted like they only had a single driver in them..The remaining correction I could not execute with passive components and the physical / acoustic arrangement of the drivers is one more step in this direction. I have an external DSP correction using one of our processors on my passive SH-50’s which does this last part..."
The SH-50, designed for commercial installations, is around 20 years old.
Tom's relatively recent four way HRE1 designed for home theaters uses four amp channels and FIR DSP.
Art
This gets into psychoacoustics, but I think a huge part of the reason that the SH-50 sounds so unique, is that the drivers are located on various points on the Z axis, but the output is in phase.
Three anecdotes:
* I rented a pair of SH-50s when I lived in a condo that's the size of a postage stamp in San Diego. Our daughter, who was in elementary school, asked me if the speakers were plugged in. When they were playing just two feet from her chair. I challenge you to show me a speaker where someone listening two feet away can't even tell if they're plugged in. That's just an absolute bonkers level of "speaker disappearing." If you put up a faux wall and put a set of SH-50s behind the faux wall, I think that people would have a tough time determining if the SH50s are ten feet away or fifteen feet away. Their location is almost completely nebulous, and the only thing that really gives them away is if you can SEE them.
* if you check out "crossovers: a step further" by LeCleach, he makes a solid case that you can have higher order filters that have phase response that's about 80% as good as a first order filter but only if the drivers are at different locations on the z axis.
* I'm pretty sure that I'm one of about 5-10 people who've heard a Lambda Unity Horn back-to-back with a Danley SH-50, and though the two speakers look similar, they sound completely different. I think a great deal of this is the crossover.
Nick McKinney (RIP) chimed in on the crossover, fifteen years ago, here: https://www.diyaudio.com/community/threads/another-unity-horn.117537/post-1711878
Hi Patrick B,
I second the idea that if there is one thing that makes a unity/synergy sound truly unique, it's the way they can disappear.
When I've compared different syn builds, and had 3 or 4 of them lined along the wall, i can get confused which one is playing...even when only one is playing !
I love it...
But that said, all the hullaballoo about the crossover being the magic is just nonsense to me.
The magic cake is the acoustic design.
The icing is xovers and delays and EQs that take advantage of the acoustic design.
And I think, hell i know, there's more than one way to skin that cat.....
I second the idea that if there is one thing that makes a unity/synergy sound truly unique, it's the way they can disappear.
When I've compared different syn builds, and had 3 or 4 of them lined along the wall, i can get confused which one is playing...even when only one is playing !
I love it...
But that said, all the hullaballoo about the crossover being the magic is just nonsense to me.
The magic cake is the acoustic design.
The icing is xovers and delays and EQs that take advantage of the acoustic design.
And I think, hell i know, there's more than one way to skin that cat.....
Yes, will do.
As you say I've seen much steeper acoustic slopes when measuring with horn cutoff and driver roll off too.
True of all the horns I've used over the years.
When I've got the second one made and something to listen to, I will seriously look at FIR with my Najda DSP. I know little about the subject other than scraps I've read.
Will keep my mind active if nothing else😀
It would be good to see the response of the prototype before recommending changes.
The volume of the angled area in your prototype area lowers the acoustical band pass frequency compared to straight holes.
A cone filler to reduce the enclosed volume would raise the upper cutoff frequency.
Does the driver gasket provide enough clearance for the mid driver's excursion at full power?
The volume of the angled area in your prototype area lowers the acoustical band pass frequency compared to straight holes.
A cone filler to reduce the enclosed volume would raise the upper cutoff frequency.
Does the driver gasket provide enough clearance for the mid driver's excursion at full power?
Last edited:
Ah so that's the effect they have! Thanks!
What I did to mitigate driver excursion issues, rather than machine away thickness in the mounting baffle was to cut a ring into the sound deadening mat that's glued on, to provide around 1.5mm extra room.
I didn't take a pic of that though.
I'd seen pretty deep machined rings in other builds.
However, even at the highest listening levels I'll ever see, and music that energises them most, the 5" drivers are hardly moving. A very gentle finger feel test proved this.
The 12" move more, but its pretty minimal - no where near the mounting face. I had perhaps falsely assumed that the basket / gasket was designed provided enough clearance.
Not had an issue mounting the Eminence Kappa Pro 15LFIIs in my Tapped horns, or Kappa 15As, & the Kappa 12As in various straight horn incarnations I've made over the years.
These don't not see PA volumes in their sheltered life with me - perhaps that is it..
Didn't have a chance to measure last night as laptop was being used by wife.. but I did play around more with toe- in.
Inspired by Mr Danley having written this many moon ago;
"In the home, that array ability isn’t needed but can be exploited in a different way.
A physical boundary is an acoustic mirror, something like a second source.
As a result, one can place an SH-50 or SM-60 against the wall ( which with the cabinet wall against the room wall, puts them toed in) and NOT have ANY reflections.
I did this in my old room which was long and narrow. Putting the speakers on the side walls this way eliminated the side wall reflections which destroy the stereo image and made the image subjectively larger / wider, a miracle!.
In commercial sound, often these speakers are mounted on a ceiling pointed down using this same boundary effect. The only “negative” (which isn’t) is that the lf and lower mid response is raised a few dB but can easily EQ fixing both magnitude and phase."
I too have a long, fairly narrow room, but with added complication of loft ceiling slopes - the above
did improve things.
Having a long room I've found is typically a good thing, and the MEH design is just what you need if the room is also narrow, as you've pointed out from Tom's quote, above. I encourage others to try his suggestion out (placing the horn mouth against the wall to eliminate side wall early reflections). If you need to add a little absorption to control midrange early reflections from irregularities in the side wall, I've found that 2' x 2' x 1" acoustic absorption tiles just at the horn mouth does this job very well to suppress any excess early midrange reflections to regain the stereo phantom center image.
I believe that the real culprit in small rooms (i.e., too small rooms--acoustically) is cross-channel reflections from the back of the room reflecting back towards the opposite side ear--which really interferes with the human hearing system's perception of depth and width (envelopment). Having the opposite side reflections delayed a bit more seems to be the key to eliminating this problem (i.e., 20-30 ms delays are the magic point usually for the human hearing system where the mind hears a separate reflection instead of a modification of the "direct arrivals"--at a delay where the human hearing system's precedence effect begins to break down). This translates to a minimum room depth+width in order to hear a good soundstage and listener envelopment sound field in a way that the human hearing system can accept as "real".
Chris
Unfortunately, I don't know what to tell you about the canted ceiling/walls. In the other instances where this has occurred, the ultimate solution was to completely cover the canted portions of the ceiling/walls with absorption material within a metre or so of the loudspeakers in order to suppress the excess reflections from them.
Apparently, the human hearing system is okay with rectangular walls, but canted walls/ceiling physically have to be outside of the 20-30 ms precedence effect delay (total path length from loudspeaker to reflection/s to the ears) or listener distortions occur. The first requirement, of course, is to have full-range directivity control out of your loudspeakers (which you have with the MEH), but this can be not enough to completely mitigate the early reflections from these surfaces.
Chris
Apparently, the human hearing system is okay with rectangular walls, but canted walls/ceiling physically have to be outside of the 20-30 ms precedence effect delay (total path length from loudspeaker to reflection/s to the ears) or listener distortions occur. The first requirement, of course, is to have full-range directivity control out of your loudspeakers (which you have with the MEH), but this can be not enough to completely mitigate the early reflections from these surfaces.
Chris
Generally I've not found this room too difficult.
I acoustically treated it some time ago, and it helped the Tractrix and Le Cléac'h horn sound.
I used viewing reflections of the horns in a mirror held along the ceilings to determine the less obvious places.
Its worked well.
I acoustically treated it some time ago, and it helped the Tractrix and Le Cléac'h horn sound.
I used viewing reflections of the horns in a mirror held along the ceilings to determine the less obvious places.
Its worked well.
The Celestion TF0510MR appears to have a gasket deep enough for it's very limited excursion, the spec sheet does not even include any information for it's coil or gap height or Xmax. It's response drops off at ~12dB per octave below its Fs of 482 Hz.
Many builds are using small mid drivers with enough excursion that rings to provide adequate clearance are required.
The Kappa 12A's Xmax is only 3.25 mm (Xlim 11.5mm), the gasket looks to be around 7mm past the surround, so more than +6dB power over Xmax would be required to start slapping the baffle.
If you have matched the SH-50 dimensions, the Fb (box tuning) is around 90Hz, maximum cone excursion above Fb is ~110Hz.
At one meter, one cabinet at 110Hz reaching Xmax is ~120dB SPL , and over 126dB at 7mm excursion.
At 110dB, excursion at 110Hz may be ~ 1mm.
The 12" driver in the DSL SH-50 has over double the Xmax of the Kappa 12A.
So yes, "sheltered life" requirements are vastly different than the PA world
Art
So, some measurements.
Did with PEQ (no X/Os apart from 1st order on HF).
Then without PEQ
LF, Red with, green without
Is classic 1st order acoustic / work PEQ cutoff?
MF
Steep 2 or even 4th order?
HF
Steeper than 1st order I'm guessing.
All 3 with polarity reversed on MF
I tried all manner of X/Os on MF just to see.
None helped get it better then the last all 3 plot.
Played around with the LF gain to get more, and reduced HF a little, for classic fall off slope, but forgot to take a plot of that(dooh!).
Feels like I'm closer..
Will have 2nd Synergy horn ready tomorrow (I hope) for proper listen 😀
Did with PEQ (no X/Os apart from 1st order on HF).
Then without PEQ
LF, Red with, green without
Is classic 1st order acoustic / work PEQ cutoff?
MF
Steep 2 or even 4th order?
HF
Steeper than 1st order I'm guessing.
All 3 with polarity reversed on MF
I tried all manner of X/Os on MF just to see.
None helped get it better then the last all 3 plot.
Played around with the LF gain to get more, and reduced HF a little, for classic fall off slope, but forgot to take a plot of that(dooh!).
Feels like I'm closer..
Will have 2nd Synergy horn ready tomorrow (I hope) for proper listen 😀