What are the pros & cons of high-efficiency horns?

[Robh3606]

Did you define your target SPL for music or HT? Can we know the value?

Hello Micro

It's based on THX reference level but used for music as well. I normally do listen at around 85dB so for me to meet THX I would need 105 db system and 115 db LFE. My system sensitivity is based around JBL E-145's which are about 98db 1 watt 1 meter. So with a pair and accounting for listening distance I can power them with 100 watts and meet my goal with about 10db of headroom to spare.

Rob

 

[morricab]

I see you want to keep the debate purely subjective, mainly on preference and your particular subjective finds of a very small numbers of cases. OK, an interesting view, but then IMHO you can not claim of physics evidence in your opinions.

My reference to formats was only to show that many people prefer an objectively compressed format to a non compressed format, not to debate recordings. But why do you say that "they took extra care in the LP production to minimize distortions" in the Valkyries" conducted by Solti? If this was just the main objective shouldn't a HiRez version sound better than the LP?

I have listened to so many exceptions to my own mental high-end rules and prejudices that I always try to be anecdotal and reporting facts as I perceive them. But when I have objective data I trust I use it as a complement of my opinions.

And yes, several Dynaudio kits sounded much better than the 1.8. I was a good friend of the local distributor in the 90's and we even built a pair in Corian, copying the Wilson's, but at that time we were not aware of his technique of controlling Corian resonance with lead!

As I have not made any measurements to date on driver compression, I use the experiential evidence I have, which is largely consistent with the research I have read on the matter. So, it is not clear to my why I cannot claim physics is likely the culprit (isn't it always?). And as you said it is IYHO.

And I had to point out that your reference to formats is often beside the point. The point is that often there is significantly wider dynamic range recorded on analog than on digital and it has nothing to do with the format and everything to do with the recording engineering. The materials with the box set on "The Valkyries " is explicit that they took extra care to minize the distortions that vinyl can experience...so it has nothing to do with the master tapes. It is possible that the high rez version sounds just as wonderful...I pesonally haven't heard it but I would argue that I haven't heard any high rez sounding as good as a good slab of vinyl in a top vinyl playback system.

If you have so many exceptions to your own rules, doesn't this just really call into question your ability to sort this stuff out? I have found that I am being drawn inexorably one direction by what I hear live vs. recorded. I have heard all the biggest and best SS and big box speakers and, no, to what I hear they are not as dynamic, not as present and not as "live" sounding. The best of the breed are encroaching on horn like sensitivity (like the big Goebel Divin series...Wilson seems to have gone back the other direction) but still don't have it (it looks like the Goebel is using pro-drivers at least) quite. Yes, it is devilishly difficult to get horns and other very high sensitivity speakers to be smooth and coherent...something I struggled with for a very long time and thus my love for huge planar speakers (like you I believe), which at least to a certain level did some aspects of dynamics (particularly micro) and were very coherent.

 

[christensenleif@msn.com]

I have a custom fullrange hornsystem and one of the many positive qualities, is that I don´t need to play loud to hear a fullbodied balanced sound, but I CAN without ripping my ears off
very low distortion is kind of spooky, because I don´t realise how loud I play until I start to get warmed up in my ears...
TAD 4003 and 2002 are true works of art and industrial quality products, unlike a lot of the underground boutique bling drivers out there
best
Leif

 

[Duke LeJeune]

Another reference I almost forgot about. Which basically shows the temperature rise may not be a bad as expected using program material some of us would actually listen too.

https://www.stereophile.com/content...r-voice-coil-temperatures-making-measurements

The methodology of that Stereophile test was imo flawed, in that it was not optimized for detecting the instantaneous voice coil heating which occurs when a high-wattage peak hits a voice coil.

They sampled the voice coil temperature at fixed intervals and tracked a moving average of the samples. They should have specifically sampled the temperature at the instant of the peaks and compared those instantaneous temperatures to the average. Sampling at a fixed interval allowed time for cooling and therefore did not catch thermal modulation effects, which can compress the peaks and thereby rob the music of dynamic contrast.

I am friends with a recording engineer who has been measuring and documenting thermal modulation (compression of peaks) in loudspeakers that come through his studio for decades. He has asked me not to give out specifics because he plans to write a paper on the subject. Briefly he finds a few decibels of compression on instantaneous peaks with speakers that have poor thermal modulation behavior, and in general he finds a correlation between high efficiency and freedom from thermal modulation, which is not surprising.

I had a conversation with Floyd Toole once about thermal modulation, the short-time-constant cousin of thermal compression. He said that thermal modulation is a real effect which he has often observed but it has not been adequately studied, and was unlikely to be adequately studied as the funding for that sort of thing would have to come from the prosound side, where currently there is insufficient interest.

I've also had conversations with Earl Geddes on the subject, and not surprisingly he also believes that thermal modulation is a limitation present in most home hifi speakers.

Anyway the significantly higher efficiency of horn systems, combined with the significantly bigger (i.e. higher thermal mass) voice coils of large drivers, implies that they will have correspondingly less thermal modulation at normal home-audio SPLs than moderate-efficiency conventional speakers.

Another possible contributor to the improved dynamics observed in horn systems is their narrower radiation pattern. The correspondingly weaker reverberant field results in a relatively lower in-room "noise floor" being present as new sounds emerge from the drivers. I'm not necessarily arguing for or against a well-energized reverberant field here; like most things in audio there are tradeoffs involved.

On the subject of radiation patterns and tradeoffs, note that there is a tradeoff relationship between on-axis SPL and pattern width. A narrow-pattern horn will have a higher sound pressure level on-axis, but it is not necessarily putting out more acoustic energy. By way of analogy, a variable-nozzle garden hose delivers higher on-axis pressure when the pattern is narrow, but the same amount of water comes out whether the pattern is wide or narrow.

 

[microstrip]

The methodology of that Stereophile test was imo flawed, in that it was not optimized for detecting the instantaneous voice coil heating which occurs when a high-wattage peak hits a voice coil.

They sampled the voice coil temperature at fixed intervals and tracked a moving average of the samples. They should have specifically sampled the temperature at the instant of the peaks and compared those instantaneous temperatures to the average. Sampling at a fixed interval allowed time for cooling and therefore did not catch thermal modulation effects, which can compress the peaks and thereby rob the music of dynamic contrast.

I am friends with a recording engineer who has been measuring and documenting thermal modulation (compression of peaks) in loudspeakers that come through his studio for decades. He has asked me not to give out specifics because he plans to write a paper on the subject. Briefly he finds a few decibels of compression on instantaneous peaks with speakers that have poor thermal modulation behavior, and in general he finds a correlation between high efficiency and freedom from thermal modulation, which is not surprising.

I had a conversation with Floyd Toole once about thermal modulation, the short-time-constant cousin of thermal compression. He said that thermal modulation is a real effect which he has often observed but it has not been adequately studied, and was unlikely to be adequately studied as the funding for that sort of thing would have to come from the prosound side, where currently there is insufficient interest.

I've also had conversations with Earl Geddes on the subject, and not surprisingly he also believes that thermal modulation is a limitation present in most home hifi speakers.

Anyway the significantly higher efficiency of horn systems, combined with the significantly bigger (i.e. higher thermal mass) voice coils of large drivers, implies that they will have correspondingly less thermal modulation at normal home-audio SPLs than moderate-efficiency conventional speakers.

Another possible contributor to the improved dynamics observed in horn systems is their narrower radiation pattern. The correspondingly weaker reverberant field results in a relatively lower in-room "noise floor" being present as new sounds emerge from the drivers. I'm not necessarily arguing for or against a well-energized reverberant field here; like most things in audio there are tradeoffs involved.

On the subject of radiation patterns and tradeoffs, note that there is a tradeoff relationship between on-axis SPL and pattern width. A narrow-pattern horn will have a higher sound pressure level on-axis, but it is not necessarily putting out more acoustic energy. By way of analogy, a variable-nozzle garden hose delivers higher on-axis pressure when the pattern is narrow, but the same amount of water comes out whether the pattern is wide or narrow.

Yes, the Stereophile study is not the proper way to do it .

But I can't understand all this talk about the difficulty of measuring transient thermal compression with audio bandwidth signals - a simple null test carried in a bridge using three high power constantan resistors and the speaker under test will immediately measure it. It is a common laboratory practice, even students do such experiments in laboratory classes, although for more interesting subjects ... Please do not tell me it is a mysterious and hard to study problem - manufacturers have been measuring it since long, currently thanks to digital technology anyone wanting to study it could do it at home using a quality USB soundcard and free software, using either signals or real music. All that is needed is time and real motivation. People can consider that digital is an inferior sound format, but ADCs and DACs do not lie if used in their bandwidth in measurements ... The truth is that no one cares to do it, it does not seem to be a real interesting problem . And yes, many poor speaker units suffer from thermal compression, but the problem was solved in good ones.

Again the main problem for such study would be the psycho-acoustics needed for interpretation of the data. What is the threshold of audibility of such effects? Is it the same for mono and stereo? Is it a real issue at common listening levels? IMHO the same type of conflicts that we have in analog versus digital or tubes versus solid state would be transposed to such studies.

Radiation patterns strongly affect stereo sound reproduction - but then we are addressing people preferences, not absolute sound quality.

 

[Robh3606]

"'The methodology of that Stereophile test was imo flawed, in that it was not optimized for detecting the instantaneous voice coil heating which occurs when a high-wattage peak hits a voice coil. "

Hello Duke

I don't see that as an easy measurement. The effect would be asymmetrical as the heating would be almost instantaneous with some delay and a great deal more delay as cooling is through air convection which obviously would take longer. I can see where larger coils with smaller air gaps would be a distinct advantage cooling down the VC.

Has anyone ever compared 2 identical drivers one with a conventional air gap the other loaded with ferrofluid. The oil like medium of the fluid would be a much better medium to cool the voice coil. I could see the use of ferrofluid being a distinct advantage to help mitigate the effect. The only compression drivers that I know for sure that use ferro are the JBL 2431/435Al and the 2435/435Be.

I also have to wonder just how audible it is short term?? Transients are by definition short lived and if there is several dB of short term compression how would you hear it?? Looking at this from the point that if your peak is 108dB vs 105dB how would you know without some reference to compare too?? Would you be able to hear the difference at those levels?? Would it just start sounding compressed like when your amps run out of juice??

What am I missing here??

Rob

 

[Folsom]

I think it's reasonable to ask how much dynamics are desired. "As much as possible" I don't think actually leads to a pleasurable listening session necessarily. Powerful and perceptible volume scaling are more entertaining than extreme crest factor parts of the music.

However you can obviously fall so far short that the music is very bland.

 

[Duke LeJeune]

But I can't understand all this talk about the difficulty of measuring transient thermal compression with audio bandwidth signals - a simple null test carried in a bridge using three high power constant resistors and the speaker under test will immediately measure it ... The truth is that no one cares to do it, it does not seem to be a real interesting problem .

I'm more inclined to believe experts who say it's a real thing than to believe a flawed test that said it's not, but on the other hand it's not an important enough issue to me that I'm going to try to measure it myself.

Radiation patterns strongly affect stereo sound reproduction - but then we are addressing people preferences, not absolute sound quality.

Radiation patterns are at the heart of all interaction with the listening room, so yes they matter a lot. I think some aspects are quite relevant to the topic of absolute sound quality, and some are matters of personal preference.

The effect would be asymmetrical as the heating would be almost instantaneous with some delay and a great deal more delay as cooling is through air convection which obviously would take longer.

The way it was explained to me, the heating actually IS instantaneous. A 100 watt transient is like your voice coil becomes the tip of a 100 watt soldering iron for an instant. The subsequent heat transfer to the air and to the surrounding motor structure takes more time of course.

Transients are by definition short lived and if there is several dB of short term compression how would you hear it?? Looking at this from the point that if your peak is 108dB vs 105dB how would you know without some reference to compare too?? Would you be able to hear the difference at those levels??

My guess is that whatever feeling was intended by that peak would be more strongly conveyed at 108 dB than at 105 dB, even if you didn't obviously "hear the difference" in the levels. In my experience a good horn system excels at conveying emotion, and I think preservation of the peaks plays a big role in that.

 

[bonzo75]

I have a custom fullrange hornsystem and one of the many positive qualities, is that I don´t need to play loud to hear a fullbodied balanced sound, but I CAN without ripping my ears off
very low distortion is kind of spooky, because I don´t realise how loud I play until I start to get warmed up in my ears...
TAD 4003 and 2002 are true works of art and industrial quality products, unlike a lot of the underground boutique bling drivers out there
best
Leif

This was a copy of the pioneer tad exclusive with a 4003 and a 1601c. Quite good in fact TAD highs are excellent quite electrostatic like in nature, very transparent, nuanced and decaying. He was driving it with a DIY pass so planning to try a couple of SETs there at some time to see how it works

 

[Folsom]

For real, when does your DIY adventure start?

 

[christensenleif@msn.com]

the DIY Pass really won´t cut it reg decay, timbre and harmonics, but is a cheap and quite rude way to get sound out of outstanding drivers
IMNSHO

and btw yes I´ve built several 5 and 5 Turbo as all out as is possible, and also have extensive listening experience with many of the other models...

 

[jdza]

I totally disagree with both First Watt not being good enough for a system with TAD drivers and with the silliness of using SETs on a pair of reflex ported TAD/Kinoshita speakers. In my main system I rolled through numerous SETs before settling on multiamping with FW on 2 of the 4 bands.

In my second system one look at the measurements twin reflex loaded TAD 15s with compression drivers will tell you that full range SETS with that complex crossover is big no no. The passive TAD crossover is a complex 6th order quasi balanced Linkwitz Reilly with Zobel compensation. The quasi balanced circuit causes a very nasty measurable resonance outside of the woofers' bandwidth that may or may not be audible but is sure likely to upset amplifiers. The treble crossover is second order with both series and parallel notch filters that may also not be pleasant for SETS. Besides the thought of trying to control the massive 15 inchers with puny SETs is totally scary.



This is what that SET will be seeing
fullsizeoutput_7de by jdza, on Flickr
With one or two of these behind it
Screenshot 2019-09-11 at 09.42.48 by jdza, on Flickr







But hell, what do I know anyway.

 

[christensenleif@msn.com]

I never said SET was right/enough for this speaker
I know of the complexity of the TAD filter with notches and the works
I just said DIY Pass wasn´t capable of maxing this speakers potential......"good enough" is relative....
best
Leif

 

[bonzo75]

I have not heard the SETs on it yet. His is a two way not a three way. I had no idea it was such a complex crossover.

SETs do great on 15 inch drivers, even 18 inch ones, depending on the driver. They can fail even on tiny ones.

 

[morricab]

I totally disagree with both First Watt not being good enough for a system with TAD drivers and with the silliness of using SETs on a pair of reflex ported TAD/Kinoshita speakers. In my main system I rolled through numerous SETs before settling on multiamping with FW on 2 of the 4 bands.

In my second system one look at the measurements twin reflex loaded TAD 15s with compression drivers will tell you that full range SETS with that complex crossover is big no no. The passive TAD crossover is a complex 6th order quasi balanced Linkwitz Reilly with Zobel compensation. The quasi balanced circuit causes a very nasty measurable resonance outside of the woofers' bandwidth that may or may not be audible but is sure likely to upset amplifiers. The treble crossover is second order with both series and parallel notch filters that may also not be pleasant for SETS. Besides the thought of trying to control the massive 15 inchers with puny SETs is totally scary.



This is what that SET will be seeing
fullsizeoutput_7de by jdza, on Flickr
With one or two of these behind it
Screenshot 2019-09-11 at 09.42.48 by jdza, on Flickr







But hell, what do I know anyway.

Size of the driver is not an inherent factor as to whether a SET can drive it or not. Sensitivity and impedance are factors. As to the crossover, this looks like typical over-engineering in order to make the speaker measure as good as possible. Probably it would sound a lot better overall with a simpler crossover that also made it more friendly to the kinds of amps that can get the most out of high sensitivity drivers.

 

[christensenleif@msn.com]

TAD also implemented the hole in the top of the TH-4003 horn, which is a Helmholzresonator to curb a broad peak around 3-4K as far as I recall

 

[Folsom]

I think it's more fair to say the Firstwatt is a poor match.

 

[bonzo75]

Also reading Bruce Edgar.s comments he did a lot of experiments and found the 4003 to sound good only in the radial horn unlike other TAD drivers

 

[christensenleif@msn.com]

don´t agree
just need to adress the slight peak
me and a friend tested compression drivers and horn combos for 3 years and the 4003 correctly implemented sounds best in all horns we tested compared to all other drivers we tested
spectacularly low distortion too, wich is already easily hearable during the distortion measurement itself...
in 2404 the 4003 is meant to go all the way up in a CD like horn which gives it a sort of forgiving character due to the hf roloff, which in itself is ok
I use it from 600-6K in JMLC 270 with TAD TD-2002 in JMLC2500 on top
best
Leif

 

[bonzo75]

All old theater horns were massive drivers to address an audience of 1000 at flea watts