"Can't Do 3D Like Other Technologies". Fair Criticism of Horn Speakers?

[jdza]

Hello

Those look familiar. 2432 and a 2404?? Do you EQ them flat or just leave them be with no CD compensation??

Rob

They are 2425 on 2344 and 2404. There is no eq. Both are simply piggybacked of the main drivers but pulled down in level to such a degree that they are not audible from the listening position nor have a measurable effect on the in-room frequency response. The experiment started off as an attempt to alleviate a mild irritation that I only became aware of after more than 10 years of listening. Trebles had a very mild tendency to "hang around" the tweeters, blooming in neither space nor energy. Not in lateral distribution but in height and depth and critically "bloom". I think this "bloom" maybe what Duke is referring to in a little more scientific way than I can? Borrowing from "conventional" speakers and a bit of Hiraga, I added 2404. The effects were not subtle at all. 2344 was an almost "what if" idea and started off as an equally crude experiment that ended as "that's perfect, don't mess with it" unscience.

Regarding horns playing forward; IMO there are 3 factors: 1. Hornish sort of speakers are big and most lack bass. So they are shoved against back walls. Like any speaker against a wall, the image will be flattened and forward. 2. The way most hornish speakers are presented will be with an up-tilted upper mid and lower treble response. This tends to flatten soundstage and push images in that soundstage forward. 3. Carolus drilled this into me: In any multi-driver horn speaker, it is critical that all drivers above bass must be physically aligned. I can add that in my experience with TAD/Kinoshita type semi-horn speakers where this is not doable some sort of electrical delay is critical to get sound staging and imaging correct. TAD (and Kinoshita) achieves this by using an asymmetric 2nd order highpass but much higher order lowpass crossover in order to delay lf output.

I am sure all these effects can easily be compensated for in the digital domain but I am not into digital or digitising analogue signals so cannot comment on DSP except saying that I have never heard it sound pleasing to my ears. That does not mean it cannot be good.

As always just my personal opinion based purely on my limited experience.

 

[jdza]

Jdza, if you don't mind me asking, WHAT IN THE WORLD made you decide to try the additional tweeters, and then the additional upper-mids as well, firing INTO your bass horn flares? Shooting sound off in other directions is highly counter-intuitive for a horn system, though obviously I think it's a good idea. I think you are the first person I've come across who is doing this on their own.

The subjective improvement probably seems out of proportion to what one might expect from merely adding a little bit more reverberant energy which arrives after several milliseconds of time delay.

How did the idea come to you in the first place?

Jdza, if you don't mind me asking, WHAT IN THE WORLD made you decide to try the additional tweeters, and then the additional upper-mids as well, firing INTO your bass horn flares? Shooting sound off in other directions is highly counter-intuitive for a horn system, though obviously I think it's a good idea. I think you are the first person I've come across who is doing this on their own.

The subjective improvement probably seems out of proportion to what one might expect from merely adding a little bit more reverberant energy which arrives after several milliseconds of time delay.

How did the idea come to you in the first place?

Sheer ignorance and absolute stupidity? With this sort of thing, one is basically sitting in the speaker. A bit like that old cartoon I suppose. As such, I long ago discovered that conventional rules do not always apply. As a long time Maggie Tympani fan I always missed that "room behind the speakers thing", so missing in horns. I suppose it was just an attempt to get that back. Somehow it brought other benefits with (so-far) no negatives.

Interesting, I see in one of Living Voice's commercial installations a rear-firing 2405 too.

 

[Duke LeJeune]

Sheer ignorance and absolute stupidity? With this sort of thing, one is basically sitting in the speaker. A bit like that old cartoon I suppose. As such, I long ago discovered that conventional rules do not always apply. As a long time Maggie Tympani fan I always missed that "room behind the speakers thing", so missing in horns. I suppose it was just an attempt to get that back. Somehow it brought other benefits with (so-far) no negatives.

I totally get that. My background was in SoundLabs, which are conceptually similar to your big Maggies. Then I heard some good horns, and wanted to combine the attributes. So about thirteen years ago I made a bipolar hybrid horn speaker - woofer & horn on both front and back. What I'm doing nowadays is arguably a more small-room-friendly evolution thereof. Credit to Jim Romeyn for his discovery that firing the additional energy upwards works well in most applications.

Interesting, I see in one of Living Voice's commercial installations a rear-firing 2405 too.

Living Voice... wow... now that's validation for both of us!!

The original Sonus Faber speaker, and the more recent Aida, both include what looks like a rear-firing mini-monitor, so apparently it's going down at least into the midrange region.

SoundField Audio uses rear-firing drivers whose outputs are digitally delayed, which is arguably a very intelligent use of DSP, though I prefer to stay all-passive and tube-friendly.

The last time I saw Terry Cain at an audio show, he had started using a rear-firing tweeter in his back-loaded horns. He grinned from ear to ear when I looked at the back of his speaker and my eyes popped out like in a cartoon.

Danny Ritchie and Clayton Shaw have both used high efficiency dipolar coaxials with the cover removed from the backside of the compression driver, so that it also radiates to the rear.

Siegfried Linkwitz, Peter Snell, von Schweikert, Revel, and Boenicke use or have used rear-firing tweeters to good effect, and I'm sure there are many others.

I've helped several Original Quad owners add up-firing tweeters behind their Quads, as the Quads have an absorptive mat behind the tweeter panel so their reverberant fields are actually lacking in top-end energy. Imo the up-firing tweeter-behind-the-panel works better than adding a forward-firing supertweeter because the '57's already have decent on-axis highs, but are lacking in off-axis highs.

 

[jdza]

Another of my non-scientific experiments that just stayed. This is the low mid driver ( 200Hz-1k) To be honest I think it had no effect at all but maybe just maybe it made tonal balance a little more, dare I say it-natural? I must confess that I am astonished at exactly how little output there is from the rear of an unloaded compression driver diaphragm.

Thousands of hours and millions of dollars were not spent in developing the rear acoustically transparent, non-resonant, audio enhancing bug screen. It is a borrowed and subsequently mutilated kitchen device.

 

[Duke LeJeune]

Thousands of hours and millions of dollars were not spent in developing the rear acoustically transparent, non-resonant, audio enhancing bug screen. It is a borrowed and subsequently mutilated kitchen device.

But... but... surely it's a mil-spec, hospital-grade, laser-cut, cryo-treated, precision-mutilated kitchen device... ??

 

[jdza]

But... but... surely it's a mil-spec, hospital-grade, laser-cut, cryo-treated, precision-mutilated kitchen device... ??

I am not sure. I cannot ask my wife as she'll then realise were here cherished stainless steel sieves disappeared to all those years ago. I can tell you that the black securing devices were specifically engineered to keep a heart-lung machine's disposable bits together and applied with a special tool to secure them to the exact tension. Any resemblance to a cable tie and a 50c cable tie gun is purely coincidental.

 

[DaveC]

I've noticed some major differences in presentation between horns. JMLC vs Iwata for example, aren't really in the same universe.

IME, a wideband JMLC application will provide the best 3D, enveloping soundstage, which is why I use it. If the horn covers enough bandwidth it can also be listened to relatively nearfield, which takes the room out of the picture to a far greater degree vs most types of speakers and allows you to hear the spatial cues in the recording more clearly. The result is more of a "you are there" presentation vs most systems and, for better or worse, more of a window into the mastering done on the recording as well. While other kinds of horns have their advantages, for example CD tends to sound more even out of the LP, they don't provide as much overall clarity in the LP, and do not project a 3D soundstage as well as JMLC. For me, I don't care that much how the system sounds while walking around the room, and I understand why in many systems this is important, but in this kind of system you are hearing a far greater amount of direct vs reflected sound at the LP. Also, if you can maintain a close to ideal polar response over a great majority of bandwidth and there is no crossover in the frequency range where hearing is most sensitive, IME this is enough, the prioritization of perfect polar response outside of this area is not nearly as important and you have to compromise other things to achieve it.

I have experimented with rear firing energy by putting the midrange driver into an open-back cabinet and using reticulated foam to control the level of the rear-firing sound, I've tried up-firing tweeters and in all cases prefer to completely damp the backwave and not use an up-firing tweeter, for me it takes away from the more direct sound and only finds it adds more of a "room sound" vs hearing the sound of the space in the recording. I find the augmentation done in this way potentially pleasant though, and I understand why many like it, but it's not for me, not in the context of a nearfield JMLC horn speaker anyways, which I understand is not very common.

 

[bonzo75]

I completely agree with JMLC

 

[Duke LeJeune]

I have experimented with rear firing energy by putting the midrange driver into an open-back cabinet and using reticulated foam to control the level of the rear-firing sound, I've tried up-firing tweeters and in all cases prefer to completely damp the backwave and not use an up-firing tweeter, for me it takes away from the more direct sound and only finds it adds more of a "room sound" vs hearing the sound of the space in the recording.

There is a threshold above which too much rear-firing energy starts to degrades clarity. It varies a bit with the specifics, but as a ballpark starting point you might try -12 dB relative to the front-firing drivers (measured at the listening position), and that's with the suggestions in the next two paragraphs in play.

The spectral balance of the rear-firing energy matters, and reticulated foam is likely to selectively attenuate the shorter wavelengths. If using an up-firing tweeter its on-axis response doesn't really matter; it's the power response (total output) that does.

Finally, imo you want to minimize the energy arriving before ballpark 10 milliseconds time delay. This is one of the big advantages of your exceptionally well-behaved radiation pattern, so you don't want your rear-firing energy spoiling that due to short reflection paths and/or radiation patterns which are too wide. David Griesinger has this to say:

"Transients are not corrupted by reflections if the room is large enough - and 10ms of reflections free time is enough."

 

[Audiophile Bill]

I've noticed some major differences in presentation between horns. JMLC vs Iwata for example, aren't really in the same universe.

IME, a wideband JMLC application will provide the best 3D, enveloping soundstage, which is why I use it. If the horn covers enough bandwidth it can also be listened to relatively nearfield, which takes the room out of the picture to a far greater degree vs most types of speakers and allows you to hear the spatial cues in the recording more clearly. The result is more of a "you are there" presentation vs most systems and, for better or worse, more of a window into the mastering done on the recording as well. While other kinds of horns have their advantages, for example CD tends to sound more even out of the LP, they don't provide as much overall clarity in the LP, and do not project a 3D soundstage as well as JMLC. For me, I don't care that much how the system sounds while walking around the room, and I understand why in many systems this is important, but in this kind of system you are hearing a far greater amount of direct vs reflected sound at the LP. Also, if you can maintain a close to ideal polar response over a great majority of bandwidth and there is no crossover in the frequency range where hearing is most sensitive, IME this is enough, the prioritization of perfect polar response outside of this area is not nearly as important and you have to compromise other things to achieve it.

I have experimented with rear firing energy by putting the midrange driver into an open-back cabinet and using reticulated foam to control the level of the rear-firing sound, I've tried up-firing tweeters and in all cases prefer to completely damp the backwave and not use an up-firing tweeter, for me it takes away from the more direct sound and only finds it adds more of a "room sound" vs hearing the sound of the space in the recording. I find the augmentation done in this way potentially pleasant though, and I understand why many like it, but it's not for me, not in the context of a nearfield JMLC horn speaker anyways, which I understand is not very common.

Dave,

What drivers are you using in your jmlc right now? have you tried wide banders in jmlc or only CD?

best.

 

[Audiophile Bill]

There is a threshold above which too much rear-firing energy starts to degrades clarity. It varies a bit with the specifics, but as a ballpark starting point you might try -12 dB relative to the front-firing drivers (measured at the listening position), and that's with the suggestions in the next two paragraphs in play.

The spectral balance of the rear-firing energy matters, and reticulated foam is likely to selectively attenuate the shorter wavelengths. If using an up-firing tweeter its on-axis response doesn't really matter; it's the power response (total output) that does.

Finally, imo you want to minimize the energy arriving before ballpark 10 milliseconds time delay. This is one of the big advantages of your exceptionally well-behaved radiation pattern, so you don't want your rear-firing energy spoiling that due to short reflection paths. David Griesinger has this to say:

"Transients are not corrupted by reflections if the room is large enough - and 10ms of reflections free time is enough."

Duke,

What frequency ranges are you recommending for this -12dB rear projected field? I am sure my system back layers so well right now due to the bass being slot loaded open baffle and due to the toe towards the listener meaning rear inner cones (from front) firing diagonally into rear stage.

best.

 

[Duke LeJeune]

What frequency ranges are you recommending for this -12dB rear projected field?

It depends.

Make your best estimate about what your speaker's off-axis response is like, and then you want your rear-firing drivers' power response to "zig" where your front-firing drivers' off-axis response "zags".

(Just for the record, there is one aspect of what I do that I'm keeping as a "trade secret". But it's a refinement, not an essential element.)

 

[Audiophile Bill]

It depends.

Make your best estimate about what your speaker's off-axis response is like, and then you want your rear-firing drivers' power response to "zig" where your front-firing drivers' off-axis response "zags".

(Just for the record, there is one aspect of what I do that I'm keeping as a "trade secret". But it's a refinement, not an essential element.)

Thanks Duke.

 

[Robh3606]

There is a threshold above which too much rear-firing energy starts to degrades clarity. It varies a bit with the specifics, but as a ballpark starting point you might try -12 dB relative to the front-firing drivers (measured at the listening position), and that's with the suggestions in the next two paragraphs in play.

Hello Duke

One of the last DIY speakers I was experimenting with I was using an original Heil driver. I am used to horns and although it sounded good I missed the clarity. I ended up using a bat of fiberglass to attenuate the rear wave and got the clarity back without loosing the spaciousness. Balancing act for sure but fun!

Rob

 

[DaveC]

Dave,

What drivers are you using in your jmlc right now? have you tried wide banders in jmlc or only CD?

best.

Hi,

I have a custom 4.5" wideband driver with a super light paper cone, it has a rising response that is corrected by the horn, and it also rolls off the low end under 400 Hz. I add a 2nd order high pass so I get a ~24 dB/oct high pass combined.

I'm not interested in CDs until I tackle a bass horn to go with it. IMO an overlooked aspect of speaker design is having the drivers' sound characteristics match. For bass I'm currently using a 15" woofer with a paper cone, and it matches the midrange's sound very well. My speaker is basically a single driver with as many compromises mitigated as possible, so I need to make it seamless across the frequency range. Substituting a different woofer can ruin it, even if they are similar. It HAS to be this particular woofer to achieve a seamless result. Of course they are expensive, custom made and take weeks to have built, if you're lucky.

As far as JMLC, I believe it preserves the fine detail of the recording better which results in a more defined, layered, 3-D soundstage.

 

[DaveC]

There is a threshold above which too much rear-firing energy starts to degrades clarity. It varies a bit with the specifics, but as a ballpark starting point you might try -12 dB relative to the front-firing drivers (measured at the listening position), and that's with the suggestions in the next two paragraphs in play.

The spectral balance of the rear-firing energy matters, and reticulated foam is likely to selectively attenuate the shorter wavelengths. If using an up-firing tweeter its on-axis response doesn't really matter; it's the power response (total output) that does.

Finally, imo you want to minimize the energy arriving before ballpark 10 milliseconds time delay. This is one of the big advantages of your exceptionally well-behaved radiation pattern, so you don't want your rear-firing energy spoiling that due to short reflection paths and/or radiation patterns which are too wide. David Griesinger has this to say:

"Transients are not corrupted by reflections if the room is large enough - and 10ms of reflections free time is enough."

-12 dB seems about right, several inches of foam were required to get the balance right... What I did was rudimentary compared to what you have done though, I think what you've done works much better. If folks like having the backward radiating energy, I think you've done the best job providing that in any speaker.

I agree as far as delay times too, and if they are long enough they can add a sense of spaciousness. I think this can have a positive effect in many situations as extending decay can result in perception of enhanced clarity and realism. OTOH, if you have a system and recording that are capable of enough resolution then it's no longer as beneficial, as our psychoacoustic expectations will have been met with only the direct sound. This is my goal, to try to extract as much fine detail as possible to the point adding decay no longer results in a psychoacoustic increase in the perception of clarity, realism or spaciousness. This is where I diverge from many of the mainstream designers that believe preference dictates having those reflections present and audible, and thus making sure the off-axis frequency response is as close as possible to on-axis, so these reflections match the direct sound. I try to avoid all of that by eliminating 1st reflections and listening fairly nearfield to a speaker with narrow dispersion. This also works in a normal room, where the wide dispersion w/reflections is best achieved in a more controlled environment.

Bass is also key, the system needs to be able to reproduce the peaks and nulls of the recording venue in a convincing manner. In a large venue these are much lower in frequency vs a relatively small listening room, so the room shouldn't contribute it's own peaks and nulls, at least not to the point it's distracting, and it needs to be capable of very low frequency reproduction to mimic a much larger space. I believe this is why subs can increase the sense of 3-D envelopment, you derive a sense of space from the low frequency behavior of that space.

 

[Audiophile Bill]

Hi,

I have a custom 4.5" wideband driver with a super light paper cone, it has a rising response that is corrected by the horn, and it also rolls off the low end under 400 Hz. I add a 2nd order high pass so I get a ~24 dB/oct high pass combined.

I'm not interested in CDs until I tackle a bass horn to go with it. IMO an overlooked aspect of speaker design is having the drivers' sound characteristics match. For bass I'm currently using a 15" woofer with a paper cone, and it matches the midrange's sound very well. My speaker is basically a single driver with as many compromises mitigated as possible, so I need to make it seamless across the frequency range. Substituting a different woofer can ruin it, even if they are similar. It HAS to be this particular woofer to achieve a seamless result. Of course they are expensive, custom made and take weeks to have built, if you're lucky.

As far as JMLC, I believe it preserves the fine detail of the recording better which results in a more defined, layered, 3-D soundstage.

nice - thanks Dave.

I do totally agree with your point about matching driver characteristics. In particular the choice of diaphragm materials in CD versus the woofer material.
Personally I am not a fan of the man made woofer type materials that you find now - I prefer the doped paper woofers.

 

[bonzo75]

As far as JMLC, I believe it preserves the fine detail of the recording better which results in a more defined, layered, 3-D soundstage.

Very well put

 

[Audiophile Bill]

-12 dB seems about right, several inches of foam were required to get the balance right... What I did was rudimentary compared to what you have done though, I think what you've done works much better. If folks like having the backward radiating energy, I think you've done the best job providing that in any speaker.

I agree as far as delay times too, and if they are long enough they can add a sense of spaciousness. I think this can have a positive effect in many situations as extending decay can result in perception of enhanced clarity and realism. OTOH, if you have a system and recording that are capable of enough resolution then it's no longer as beneficial, as our psychoacoustic expectations will have been met with only the direct sound. This is my goal, to try to extract as much fine detail as possible to the point adding decay no longer results in a psychoacoustic increase in the perception of clarity, realism or spaciousness. This is where I diverge from many of the mainstream designers that believe preference dictates having those reflections present and audible, and thus making sure the off-axis frequency response is as close as possible to on-axis, so these reflections match the direct sound. I try to avoid all of that by eliminating 1st reflections and listening fairly nearfield to a speaker with narrow dispersion. This also works in a normal room, where the wide dispersion w/reflections is best achieved in a more controlled environment.

Bass is also key, the system needs to be able to reproduce the peaks and nulls of the recording venue in a convincing manner. In a large venue these are much lower in frequency vs a relatively small listening room, so the room shouldn't contribute it's own peaks and nulls, at least not to the point it's distracting, and it needs to be capable of very low frequency reproduction to mimic a much larger space. I believe this is why subs can increase the sense of 3-D envelopment, you derive a sense of space from the low frequency behavior of that space.

What are your thoughts on the horns using compression driver mids / treble with reflex cone bass then, Dave?

 

[Audiophile Bill]

Hi,

I have a custom 4.5" wideband driver with a super light paper cone, it has a rising response that is corrected by the horn, and it also rolls off the low end under 400 Hz. I add a 2nd order high pass so I get a ~24 dB/oct high pass combined.

I'm not interested in CDs until I tackle a bass horn to go with it. IMO an overlooked aspect of speaker design is having the drivers' sound characteristics match. For bass I'm currently using a 15" woofer with a paper cone, and it matches the midrange's sound very well. My speaker is basically a single driver with as many compromises mitigated as possible, so I need to make it seamless across the frequency range. Substituting a different woofer can ruin it, even if they are similar. It HAS to be this particular woofer to achieve a seamless result. Of course they are expensive, custom made and take weeks to have built, if you're lucky.

As far as JMLC, I believe it preserves the fine detail of the recording better which results in a more defined, layered, 3-D soundstage.

In terms of jmlc profile, I completely get why folks would use it as an upper mid range. The main drawback (imho) is that massive increase in mouth circumferential size for same fc as say a tractrix thus you get to the point where it isn’t practical to use say for upper or mid bass. In a multi-way system, you would have massive issues with height unless you used different horn geometries for these duties - or put it another way, a multichannel 5 way style Cessaro but all in jmlc would be like 10 feet high. Creating some arc in the structure to address is imho a bad idea too (to account for this) because you are shifting time alignment in doing so. No free lunch.