There is a LOT more to loudspeaker design than this but..
1. Don't know, would have to measure both..
2. No.
3. Yes.
Better (more limited) questions though!
Hi Scott, all
Model the two drivers you’re talking about with the same physical spacing. Unless they are within ¼ wavelength or less, they do not add coherently into one new source, they radiate as two separate sources and produce an interference pattern (lobes and nulls). Now play with your crossover “enough’ to cause the .5 dB variation on axis you’re talking about. Now look at how the polar pattern changes as you cause that effect.
You will notice that when you cause that .5 dB on axis variation, you also produce a phase shift AND you alter the polar pattern.
You are not simply causing a half dB change in amplitude, you changing where the sound goes at that frequency, your changing the time as well so how do you know which is causing what you hear?
It is unlikely your ears are that much more sensitive than what audiologists can detect with an amplitude only change, but you are describing a situation where you are changing more than just amplitude.
In large scale sound, that interference pattern is why even a slight wind can cause an audible effect; it is moving the interference pattern multiple sources produce.
Siegfried Linkwitz found a couple DIY versions of a Synergy horn and included them in his page, one at 50 feet and another at 400 feet, it is the constant directivity and lack of interference pattern that allows them to sound essentially the same.
http://www.linkwitzlab.com/Constant_directivity_louds.htm
No, the K horn and lascalla are both a pair of point sources, they upper and lower sections radiate independently and do not sum coherently into one source because they are acoustically too far apart and radiate from two different points in time and space.
Best,
Tom
Model the two drivers you’re talking about with the same physical spacing. Unless they are within ¼ wavelength or less, they do not add coherently into one new source, they radiate as two separate sources and produce an interference pattern (lobes and nulls). Now play with your crossover “enough’ to cause the .5 dB variation on axis you’re talking about. Now look at how the polar pattern changes as you cause that effect.
You will notice that when you cause that .5 dB on axis variation, you also produce a phase shift AND you alter the polar pattern.
You are not simply causing a half dB change in amplitude, you changing where the sound goes at that frequency, your changing the time as well so how do you know which is causing what you hear?
It is unlikely your ears are that much more sensitive than what audiologists can detect with an amplitude only change, but you are describing a situation where you are changing more than just amplitude.
In large scale sound, that interference pattern is why even a slight wind can cause an audible effect; it is moving the interference pattern multiple sources produce.
Siegfried Linkwitz found a couple DIY versions of a Synergy horn and included them in his page, one at 50 feet and another at 400 feet, it is the constant directivity and lack of interference pattern that allows them to sound essentially the same.
http://www.linkwitzlab.com/Constant_directivity_louds.htm
No, the K horn and lascalla are both a pair of point sources, they upper and lower sections radiate independently and do not sum coherently into one source because they are acoustically too far apart and radiate from two different points in time and space.
Best,
Tom
with my somewhat flexy walls, the Klipschorn is unlistenable and has to be shut off within a minute of attempted listening - even harpsichord sounds "thuddy" - a Karlson will sound better in bass to midbass quality plus do well with vocals, cello and bowed bass viola. I've owned the Speakerlab K-ripoff 40 years ago then two sets of real Klipschorn. They are elegant look-wise. La Scala has a 2nd order (outdoors) rolloff below 110Hz.
Hi Tom,
I've only heard Unity's.. as for the rest, please provide highly detailed and relevant measurements.
1. In room (small room) directivity patterns at freq.s lower than 400 Hz at a meter or more, and
2. Broad-band anechoic directivity plots with no more than .25 db "stepping" and based on axial measurements no more than 5 degrees in "distance" within at least a 120 degree "window".
I think without that data it will be little more than a series of communications that may be entirely without merit - and yes, perhaps principally or even exclusively by me (..and I'm not a person that has an aversion to being wrong).
I'm assuming that the fact it does very reasonable square waves answers a lot of other questions I could pose. We can assume I believe that crossover integration is done correctly, What I am curious about is the possibility of errant resonance in the mid range from firing cone midranges into compression holes. PWK changed his slot size in the Khorn at least once. I would imagine this would be a magnitude more critical for mid hole/slot. Also, are there mid range drivers efficient enough to only require one or two instead of several and still keep up with the typical compression tweeter driver?
That would a pointless exercise by itself. You would at least have to measure a direct radiator in the exact same room/position/etc. for comparison. It would also be an examination of the one part of the Synergy horns that is not remotely new, so I guess I don't see the point at all.
Last edited:
How do I know? As an absolute - I don't. I can accept that.
Note however that a crossover encompasses more than the summed high and low-pass sections. Eq'ing-out deviations well above or below the crossover region are typically a fundamental part of the process. I still feel fairly confident in my prior statement, but it would be best to hear from others.
This is not for a point of comparison with another design, and it is not pointless.
If you have any similar measurement available - please provide a link.
Graci!
..the .5db was just a reference that was made and disputed.
It's nothing more than a starting-point for grosser deviations, but without some base-line even 3db or more peaks or dips will be contested.Hi Scott
Unity’s are at least 10 years old, the first ones are about 15 years old and while they did radiate as a single driver, none of them were as phase coherent as the Synergy horns. Time was the last thing which kept them from being what Dick Heyser described as a true single broad band source.
1 the loudspeakers directivity is not affected by the room but what you measure at the LP is governed by both, the polar map I posted has a low end cutoff of 30Hz, not 400hz.
2 I am not clear what you ask for, can you please supply measurements of any loudspeaker done the way you describe as a frame of reference?? Also hopefully include the time or acoustic phase response.
Similarly, without hearing them, you will be stuck with the philosophical view, I don’t think you will be able to grasp the importance of the things I mentioned like coherent addition of multiple sources into a single source, the reduction of reflected / late sound directivity produces and especially the time aspect.
Hi Peteloni
Quite a few years ago, I posted some square waves for fun here.
I am sure they could be found by searching.
The design has gone through a lot of revisions as well, the purpose of the holes is twofold.
First, they along with the trapped air volume under the mid and lf cone drivers form an acoustic low pass filter which is set to be just above the electrical low pass in the crossover.
This greatly reduces the harmonic distortion all drivers produce or rather it stops that from getting into the horn body and radiating.
The mid range (where harmonics fall in the range ones ears are most sensitive) are the area with the most headroom and so very low distortion. Second, by keeping the holes in the corners, they have a very small impact on the horns response so that the hf driver response looks very much like the response with no holes.
In order to drive the horn coherently like one source and benefit from horn loading, smaller drivers worked better than a larger driver. Many aspects of this entire system are tied directly to the wavelengths involved for each range in fact that is the key to making them work.
Best,
Tom
Unity’s are at least 10 years old, the first ones are about 15 years old and while they did radiate as a single driver, none of them were as phase coherent as the Synergy horns. Time was the last thing which kept them from being what Dick Heyser described as a true single broad band source.
1 the loudspeakers directivity is not affected by the room but what you measure at the LP is governed by both, the polar map I posted has a low end cutoff of 30Hz, not 400hz.
2 I am not clear what you ask for, can you please supply measurements of any loudspeaker done the way you describe as a frame of reference?? Also hopefully include the time or acoustic phase response.
Similarly, without hearing them, you will be stuck with the philosophical view, I don’t think you will be able to grasp the importance of the things I mentioned like coherent addition of multiple sources into a single source, the reduction of reflected / late sound directivity produces and especially the time aspect.
Hi Peteloni
Quite a few years ago, I posted some square waves for fun here.
I am sure they could be found by searching.
The design has gone through a lot of revisions as well, the purpose of the holes is twofold.
First, they along with the trapped air volume under the mid and lf cone drivers form an acoustic low pass filter which is set to be just above the electrical low pass in the crossover.
This greatly reduces the harmonic distortion all drivers produce or rather it stops that from getting into the horn body and radiating.
The mid range (where harmonics fall in the range ones ears are most sensitive) are the area with the most headroom and so very low distortion. Second, by keeping the holes in the corners, they have a very small impact on the horns response so that the hf driver response looks very much like the response with no holes.
In order to drive the horn coherently like one source and benefit from horn loading, smaller drivers worked better than a larger driver. Many aspects of this entire system are tied directly to the wavelengths involved for each range in fact that is the key to making them work.
Best,
Tom
Hi Tom,
I've only heard Unity's.. as for the rest, please provide highly detailed and relevant measurements.
1. In room (small room) directivity patterns at freq.s lower than 400 Hz at a meter or more, and
2. Broad-band anechoic directivity plots with no more than .25 db "stepping" and based on axial measurements no more than 5 degrees in "distance" within at least a 120 degree "window".
I think without that data it will be little more than a series of communications that may be entirely without merit - and yes, perhaps principally or even exclusively by me (..and I'm not a person that has an aversion to being wrong).
Scott, those some pretty ridiculous demands.
1. There is little directivity below 400 Hz in any speaker, let alone the Synergies. They probably do have some directivity below 400 Hz. Tom already provided a graph of that. Why would it change in a small room? What purpose would such a graph serve?
2. How many measurements have you seen that meet those demands? What is the basis for these requirements?
Even if we had these, there would be nothing more that could be gleaned from these measurements that we can't already hypothesize based on what we have.
So this:
Source: Red Spade Audio
Might not be such a great idea? No corners there... I always wondered if a round version of the synergy horn would work. Or even a synergy with different horn profile. Things like the folded back mouth radius on a LeCleac'h horn etc.
Last edited:
I'd say Scott is making some very reasonable assumptions from the data provided. Unfortunately, they mostly miss the mark because he hasn't heard them. That's the trouble with speakers, isn't it?
The Syngergy horns do not sound forward (quite the opposite), they don't sound unbalanced (at least not most of the line), and they do sound better than the vast majority of direct radiator speakers. In an A/B test, they make the colorations of direct radiators very obvious by comparison. Are they perfect? No. Do they sound more like real life than the data might indicate? Yes. Will they play insanely loud without obvious distress? Hell yes.
OT: As for the felt in horn and HOMs, well, Western Electric was doing that in the 1930s, if not earlier. No fancy name for it there, either.
The Syngergy horns do not sound forward (quite the opposite), they don't sound unbalanced (at least not most of the line), and they do sound better than the vast majority of direct radiator speakers. In an A/B test, they make the colorations of direct radiators very obvious by comparison. Are they perfect? No. Do they sound more like real life than the data might indicate? Yes. Will they play insanely loud without obvious distress? Hell yes.
OT: As for the felt in horn and HOMs, well, Western Electric was doing that in the 1930s, if not earlier. No fancy name for it there, either.
No doubt where it started, W.E. did almost everything first anyway. Except for maybe Tom's idea, that one may in fact be new.
Not a demand, a request - and I think a kindly worded one at that (..as it was intended to be).
An unusual requests, yes.
As Tom points out - the loudspeaker has it's own directivity pattern, but that's not the same as its pattern in-room. (..which will of course change from room to room and position to position).
The key point here however is: at what point does that transition (freq.) reasonably occur for this design under the small-room (listening room) context? i.e. At what point does directivity reasonably become moot for this design in a listening room context under our sort of user application (stereo hifi)?
THAT's (part of) what I want to know, and to the best of my knowledge has not ever been shown.
It's an unusual request precisely because unusual claims are being made.
Principally that:
Increased directivity is *better* because it reduces room effects. (..which, as the theory goes, apparently has a completely positive subjective effect.)
Note that I'm not arguing against this at lower freq.s, rather:
I simply want to be reasonably well informed as to how low that higher directivity pattern can go.
I'm also wondering if there is an approach that can improve upon this in some manner - even if the result is an exceedingly small improvement.
Is any of that unreasonable? Can it be reliably obtained some other way?
Can you formulate a reasonably accurate hypothesis, (within a very modest freq. range), of where this point will occur? 500 Hz? 450 Hz? 400 Hz? 350 Hz?
Actually it wouldn't matter.
I'd still want to see an actual result (..but there certainly is no expectation of that).
Ok I'm going to drill down to what might be the real isuue. I think we all know that a conical horn is going to be if not the most efficient, then at least the least colored. Done, that one is out of the way. I think we all know that a compression tweeter at the apex of a horn sounds great. Done that one is out of the way. I think we all know that below a couple of hundred hz bass is going to do whatever the hell it wants to do in a given room and mostly all we can try to do is aim it, or put it in a corner. The truth is we are all along for the ride. done, Done, put it away. we all know that the closer you can get yourt drivers together the better. Done put that away,.
Now for the real crux of the matter. What are those cone driven mid-ranges really doing under compression and firing into the side of the horn? does the bend make them ragged? is there peakiness? is there cancellation? Gentlemen it is the mid range firing into the side of that horn that is actually going to determine what this thing sounds like, because all else can be dealt with in the crossover. Show us a curve of the mid and horn *only* in the spectra that they alone are producing, not in the crossover areas we will largely know what is *really* going on in this speaker. It aint the tweeter HF area and it damn sure is not the woofers LF area. The sound of this speaker will largely be the sound of whatever convolutions the midrange/cone/slots/compression firing into that horn at what basically right angles. Yes we know the math, yes we know about wavelengths. Show us that midrange driver curve alone and you have shown us the speaker. If that curve is not a mess then this speaker, as I highly suspect sounds great.
Now for the real crux of the matter. What are those cone driven mid-ranges really doing under compression and firing into the side of the horn? does the bend make them ragged? is there peakiness? is there cancellation? Gentlemen it is the mid range firing into the side of that horn that is actually going to determine what this thing sounds like, because all else can be dealt with in the crossover. Show us a curve of the mid and horn *only* in the spectra that they alone are producing, not in the crossover areas we will largely know what is *really* going on in this speaker. It aint the tweeter HF area and it damn sure is not the woofers LF area. The sound of this speaker will largely be the sound of whatever convolutions the midrange/cone/slots/compression firing into that horn at what basically right angles. Yes we know the math, yes we know about wavelengths. Show us that midrange driver curve alone and you have shown us the speaker. If that curve is not a mess then this speaker, as I highly suspect sounds great.
Last edited:
..Unfortunately, they mostly miss the mark because he hasn't heard them. That's the trouble with speakers, isn't it?
The Syngergy horns do not sound forward (quite the opposite)..
Yup, only having heard the Unity and other midbass horn designs.
Not forward. That's interesting.
At what point does the soundstage (and the images within it) start relative to the speakers?Hi Scott
Pano is right, you have reason to be suspect, this is something different.
Loudspeakers have their own directivity pattern BUT that doesn’t change in a room, the loudspeaker interacts with the room so what you get at the LP is direct and reflected sound.
The more directivity the speaker has, the greater the distance what you hear is dominated by the direct sound (also called near field in hifi land).
In commercial sound, (other signal chain and loudspeaker aspects aside) it is the direct sound which is related to your ability to understand words or produce a good STIpa measurement, it is reflected sound which inhibits that. In a stereo, it is the direct sound that carries the “information” about the stereo image, it is the reflect sound which is not part of the signal which while possibly pleasing in some cases, harms the actual recorded image.
Think of it like a flashlight, the room doesn’t affect the radiation pattern but what you see is from the direct and reflected light. To project an image, you want the room to be dark and the projection beam projecting as little as possible off the screen. To project an image one needs light that started as a point when it passes through the film, not a surface or pattern.
There is actually a lot that can be gleaned from the CLF data file.
You can get the viewer and data file for the SH-50 here;
CLF data | Danley Sound Labs, Inc.
The speaker was measured over a sphere and a number of parameters derived. For example, conventional polar plots You ask about directivity at 400Hz, you can read it over a wide frequency range if you want but reading the polar plot view at 400Hz, the level 90 degrees off axis is about -15 dB relative to on axis, to the rear the level is about -15dB.
The speaker was measured by an independent company and not retouched etc.
This was taken every 5 degrees over an entire sphere and a number of parameters derived.
This is probably not a format used in hifi, in fact it is a format that one can use to look at the 3D balloon data used to make the EASE model used in large system design.
One thing this also provides is an indicator of the linear loudness range. In this case a music shaped noise is fed to the speaker and the level slowly raised in 3dB steps every 5 min. AT the point the response shape deviates 3dB at any point, that is the max input Voltage in the CLF file. Yes you can go louder, this is below the electrical capacity but the response will be starting to suffer.
Like I explained, can they be made flatter? Sure but for this market, high power crossover parts all cost real money so in commercial sound they aren’t as flat as is possible.
Personally I would love to make a hifi version hifi has been my hobby since I was a kid, I had been trying to sell the idea of making a kit available through a fellow enthusiast in the biz but so far unsuccessfully.. I have a pair of SM-60’s with a more elaborate crossover that is + - 1dB with a slight tilt for living rooms, in a nice cabinet would fly but the company is busy elsewhere right now.
I design the stuff but I don’t run the company and may partner has done an amazing job doing it his way so I am not going to argue.
What I can do is try and make our products as close to signal faithful as I can given the constraints. AT the moment this is enough to win most side by side auditions, even with larger systems that cost several to many times more. This is why so many stadiums are installing them this year.
So far as what can be measured, one can’t measure stereo image, that is subjective but what you can measure is the MTF for a loudspeaker.
An MTF is a measure of resolution; see the link below as it applies to optics;
Modulation Transfer Function - what is it and why does it matter? - photo.net
For a loudspeaker, one can measure the depth of modulation also, that depth is the difference between on and off producing a tone, turing on and off at an increasing rate.
A series of these MTF’s allow one to calculate a prediction of language independent speech intelligibility called STIpa.
STIPA Pro | Studio Six Digital
If you take your own MTF’s with say ARTA, one sees that just as in intelligibility, as you add reflected sound, the depth of modulation / articulation is reduced. MANY things limit the MTF but among them is the direct to reflected sound level difference captured by the MTF.
Now it is certainly possible to have a very poor articulation but also have very enjoyable sound, if you hear a choir in a large church that is an extreme example.
What I am talking about is what governs how much of the signal information reaches your ears as intended.
I hope this fills in some blanks
Best,
Tom
Pano is right, you have reason to be suspect, this is something different.
Loudspeakers have their own directivity pattern BUT that doesn’t change in a room, the loudspeaker interacts with the room so what you get at the LP is direct and reflected sound.
The more directivity the speaker has, the greater the distance what you hear is dominated by the direct sound (also called near field in hifi land).
In commercial sound, (other signal chain and loudspeaker aspects aside) it is the direct sound which is related to your ability to understand words or produce a good STIpa measurement, it is reflected sound which inhibits that. In a stereo, it is the direct sound that carries the “information” about the stereo image, it is the reflect sound which is not part of the signal which while possibly pleasing in some cases, harms the actual recorded image.
Think of it like a flashlight, the room doesn’t affect the radiation pattern but what you see is from the direct and reflected light. To project an image, you want the room to be dark and the projection beam projecting as little as possible off the screen. To project an image one needs light that started as a point when it passes through the film, not a surface or pattern.
There is actually a lot that can be gleaned from the CLF data file.
You can get the viewer and data file for the SH-50 here;
CLF data | Danley Sound Labs, Inc.
The speaker was measured over a sphere and a number of parameters derived. For example, conventional polar plots You ask about directivity at 400Hz, you can read it over a wide frequency range if you want but reading the polar plot view at 400Hz, the level 90 degrees off axis is about -15 dB relative to on axis, to the rear the level is about -15dB.
The speaker was measured by an independent company and not retouched etc.
This was taken every 5 degrees over an entire sphere and a number of parameters derived.
This is probably not a format used in hifi, in fact it is a format that one can use to look at the 3D balloon data used to make the EASE model used in large system design.
One thing this also provides is an indicator of the linear loudness range. In this case a music shaped noise is fed to the speaker and the level slowly raised in 3dB steps every 5 min. AT the point the response shape deviates 3dB at any point, that is the max input Voltage in the CLF file. Yes you can go louder, this is below the electrical capacity but the response will be starting to suffer.
Like I explained, can they be made flatter? Sure but for this market, high power crossover parts all cost real money so in commercial sound they aren’t as flat as is possible.
Personally I would love to make a hifi version hifi has been my hobby since I was a kid, I had been trying to sell the idea of making a kit available through a fellow enthusiast in the biz but so far unsuccessfully.. I have a pair of SM-60’s with a more elaborate crossover that is + - 1dB with a slight tilt for living rooms, in a nice cabinet would fly but the company is busy elsewhere right now.
I design the stuff but I don’t run the company and may partner has done an amazing job doing it his way so I am not going to argue.
What I can do is try and make our products as close to signal faithful as I can given the constraints. AT the moment this is enough to win most side by side auditions, even with larger systems that cost several to many times more. This is why so many stadiums are installing them this year.
So far as what can be measured, one can’t measure stereo image, that is subjective but what you can measure is the MTF for a loudspeaker.
An MTF is a measure of resolution; see the link below as it applies to optics;
Modulation Transfer Function - what is it and why does it matter? - photo.net
For a loudspeaker, one can measure the depth of modulation also, that depth is the difference between on and off producing a tone, turing on and off at an increasing rate.
A series of these MTF’s allow one to calculate a prediction of language independent speech intelligibility called STIpa.
STIPA Pro | Studio Six Digital
If you take your own MTF’s with say ARTA, one sees that just as in intelligibility, as you add reflected sound, the depth of modulation / articulation is reduced. MANY things limit the MTF but among them is the direct to reflected sound level difference captured by the MTF.
Now it is certainly possible to have a very poor articulation but also have very enjoyable sound, if you hear a choir in a large church that is an extreme example.
What I am talking about is what governs how much of the signal information reaches your ears as intended.
I hope this fills in some blanks
Best,
Tom
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.