Speaker sensitivity vs SQ

Thank you clearthinker for your welcome! I agree with your perception, equipment matching is the essence of a good system. Essentially, any amplifier topology or speakers design can bring us to musical bliss when every element is in perfect synergy. That takes long experimentation, decades of trying combinations out, and ultimately sheer luck on stumbling onto it. After much experimentation we know more, and can hit synergy more often. I do not hold any technology superior to any other, each have their own compromises, and cater for different experience.

In my audio journey my experience is that the major cost element is the time and work I put in it. If you are willing to build the cabinets and crossover, then high efficiency can be built for 6K$ or less. (Sure, a comparable commercial design would be around 35-60K$, but then there’s no need for 10,000 hours invested into R&D.) So, compared to even meager speakers costing today at 10K$, and SOTA around 250K$+, I think that cost is not an issue anymore to high efficiency for a dedicated serious DIYer, who had serious mentors and experience.

To audio2design: I respect where you are coming from, and all of us have different sets of experiences. Thus, I will refrain from holding my experience base superior to yours, and I am open to new experiences. I have heard and seen enough not to pass judgment blindly, and to find treasures at corners where I would never suspect them. That’s the point of this thread, to open our minds to new possibilities. When I wrote about the low level listening with SE, I wanted to convey that it can reproduce music when played at soft level, while complex big amps fall into anemia (crap out) when played at the same soft volume. The SE amp with the H.eff speakers can play distortion free up to 100dB/m levels. While my SE amp has super low, less than 1W output into the 16R H.eff speakers, the output transformers are rated at 25W, and the power supply is build to deliver enough for 300W output. Most SE amps fail at rating the PS and the OPT close to the output level, (and fail even more miserably at supplying a well filtered quiet B+) and thus operate close to transformer saturation, starved PSU and lack of low level detail. When built right, the 1W amp sounds as dynamic and strong and distortion free as an absolutely massive amp would sound. Indeed, you are right, most of what I said depends on implementation. Actually, everything depends on implementation, so I’m glad we can agree. There’s no single immutable point, everything is just a generalization, as basically not a single parameter is kept constant for all equipment installations. I was trying to rack my brains to come up with the difference between H.eff and L.eff that would be a constant. I could think of two aspects, which I found immutable regardless of setup conditions. One is that with H.eff the concept of volume becomes obsolete: adding volume gives the feel of greater dynamic range, but I do not get the impression that the average sound volume is getting significantly louder. It’s the dynamic range that expands. Still, this can depend on the system, and if preamp is not up to spec, you might have different experience. My second observation is more versatile: H.eff allows you access to EVERY record. Labels and pressings one would never think as capable of providing music experience become playable and at shockingly good quality. Not as good as a perfect recording, but they allow you to ENJOY them, and they will sound much better than you ever heard them on any system, while on L.eff they are intolerable.

Getting back to the amp issue: laws of physics dictate that even if we had a perfect way to amplify signals, then the amplification to 1000W introduces x1000 times more entropic distortion than amplification to 1W, and there is no mechanism to get back that loss. As technologies are imperfect, the actual distortions with increasing power are even greater. My experience showed me that tube amps have a much, much greater potential to present microdynamics, harmonic and spatial content. Not the 100+W ones, the small ones. The higher the power the more the intricacies fall apart, and a big fat 200W tube amp just becomes a wannabee solid state amp.

To me, SS amps sound inhuman, with extremely few exceptions. Sure, lots of power and control, but lack of harmonic riches and low level detail. I agree, they sound as perfect PA amps, and that’s king when one is going for the ultimate amplified sound. My comparison is to live, unamplified classical instruments. Tubes do a much better job for me to portray the message that the acoustical instruments portray. 

I understand that systems preferences ultimately boil down to our listening habits, subjective preferences, and how far we came on our audio journey. Thus, such debates will never come to a closure. I hope that everyone find his and her source of joy.

I'm completely agreed with @realworldaudio.
High sensitivity speakers give much more real music reproduction.
The low sensitive speakers, even big and expensive like Wilson Audio, Dynaudio,... have artificial bass reproduction and sound not alive. 
Electrostatic speakers have good microdynamics and sound more alive. But they have issue with macrodynamics and bass.
Regards,
Alex.

also, as an ESL owner, I think about the sonic impact of transformers and edge clamping distortion as well as energy storage in the panel....ain’t no free lunch

Don't forget flexing and subtle mechanical movement of the panel :-).   Good post tomic601.   Lots of hand waving but at the end of the day you are either having accurate, low distortion movement, no matter the efficiency and amp, or you are not.

Realworldaudio,

While I agree with some of what you wrote, ie that high eff speakers will reveal more amp issues, I can't agree with much else. They won't reveal any more of the signal chain than the amp.

I find other statements mainly conjecture based on implementation and nothing about hi/low eff.  Low efficiency speakers are not more dynamic and their bass is not more natural.  If you are running SE tube, good chance you have some frequency anomalies you like and those anomalies can be conducive to low level listening. 

Thanks realworld.
Congratulations on your perceptive first post here.  Keep 'em coming!

I note this thread is well into its 4th page.

You mention equipment mismatching.  How true!
In another recent thread dealing with: What's the most important element in getting good SQ, nobody had mentioned this in 40 posts, so I chimed in.  I say it may be the MOST important single element.  Certainly if you get it wrong.

But I am not in agreement with a generalised preference for H.eff speakers and tube amps.  My Martin Logan CLXs are 87dB so fairly inefficient.  But 400w of Krell Class A (real 1980s Class A, not today's pseudo stuff) does the trick. With +12dB on my pre-amp it's actually very loud at a quarter up.  I have no doubt that some big tube monoblocs would also do a good job. 

My speakers are big but panels.  As you say, to get this kind of SQ and H.eff I would need very big speakers indeed, probably with very high associated cost some of which I would recoup on smaller amplifiers and electricity bills.

Clearthinker, good questions, great thread!You brought up all the eternal questions in audio - high power vs low power and high vs low efficiency, and of course, cost issues. Looking at the average audio product, 1 in 10 sounds very good, while the rest is mediocre to bad. (Due to flaws, or more frequently, due to equipment mismatching. Even for low efficiency speakers, you have to hear about 20 to find one great sounding. The same is true for high efficiency speakers, and given that it's rare to hear even a single one of those, and you need to hear 20 systems with them to have a comparable basis to low efficiency speakers. Most people do not get to that point, and as there's a minuscule chance that one trips onto a great sounding one right at the first try, the automatic assumption is that H.eff speakers are inferior.
The true difference between low and H.eff speakers is that H.eff speakers couple cone movement to the air much better, hence they translate much more of the audio signal to sound. As a result, you hear much more dynamic resolution, and you also hear much more of any defects the system has. It's not just the speakers issues are magnified with high sensitivity, but also the rest of the audio chain is forced to come out of hiding, and their flaws are exposed. That's why slid state amps sound generally quite bad or at least disappointing with them. Indeed, most tube amps have issues as well, but a good tube amp can be built in a reasonable budget that will sound good, while I have yet to hear a solid state amp that sounds even remotely acceptable once efficiency goes beyond 100dB/Wm.
The issues with H.eff were not just affordability but availability as well. The retailer can stock half a dozen to a dozen plus speakers in the same space that a single pair of H.eff speakers require. When a store owner can stock 5 pairs of speakers total, he's not going to stay in business. Also, while nowdays the price range for H.eff drivers went down, in the 50s-60s when the low efficiency speakers changed the scene the driver costs for H.eff drivers were much much higher than low eff drivers, and the relative cost of even a simple audio system rivaled that of a motor bike or a car, so the only choice really was the low efficiency version. Plus, who wants to give up half the living room space to a stereo system when you can get one that can be shoved onto the shelves? Also, you cannot sell the weak amps and sources with H.eff speakers as any deficiency will be glaring at you, so that's another key factor that limits H.eff and favors low eff speakers.
Class D for the low end is a very enticing solution, and is a fantastic compromise if you want H.eff midrange and top end. Yet, the stark difference in sensitivity between the two parts of the spectrum will come and bite our donkeys (or, asses). Your ears will come to realize after a while that there is a huge disconnect, and you will loose interest in the sound. To me, one of the biggest advantages of high efficiency is the efficient bass. It sounds first as if there was lower bass extension, because you are not getting that "pressurized" feeling that your head is about to explode with the sound pressure. However, you will notice that the sound is not a porcelain muppet freakshow anymore with angry goblins kicking your seat, but you are hearing much more natural presentation - base will sound breathing and alive, and when you turn the volume down the soundstage does not collapse.Amplifier power: for 100dB/Wm efficiency you need 60 milliwatts for a VERY LOUD volume, and you still have tremendous headroom even when using a 500milliW (half a W) amplifier. The quiet passages will play at microwatt levels - that is a few millionth of a watt! So, if you have an amp that excels at 6000W that;s not necessarily going to be a virtue here, as does it also excel at a millionth of a watt?
I have a youtube channel dedicated to audiophile education, (Real World Audio), and most of my videos are about these subjects, distilling my 20+ years of experience building & designing speakers and amplifiers.

No free lunches indeed. There are those who'd like to sell the idea that big means much more expensive, but to them boutique, often vintage drivers with field coil or alnico magnets in addition to luxury, hand-build finishes is the only validating entry into the high efficiency arena. Their (affiliated) business, obviously, but NOT representative of what's possible with high efficiency and large size at more reasonable prices that can as well produce great sound.

The real non-free lunch here is size (and the product segment typically associated with that), as we're seeing considerable expense being shelled out already for a speaker package not much larger than a meager shoebox. Sound is physics, meaning displacement and efficiency is irreplaceable as that which accommodates truer dynamic envelope, low distortion and ease - core parameters into what is perceived authentic as a live event, yet they're wildly neglected for above reasons; not price, but size predominantly and the product narrative that follows here. Last but not least: live sound mayn't be that big of a draw after all..

a systems engineering approach should always compare the output with the input, including impulse response. Sure a hyper well executed high efficiency driver using best available materials, engineering, machining, etc are expensive. I think any reputable end speaker company designing own drivers, even semi custom w Scanspeak, are similar $$$$$. Some of those take into consideration that output is not a measure of quality.

I also believe, no designer should ignore ear brain. Many just sell distortion we like. Flavorizers.....

sure, the holy grail is low mass, breakup far from the passband, pistonic, with super high strength motors and high duty cycle....nothing has changed....

also, as an ESL owner, I think about the sonic impact of transformers and edge clamping distortion as well as energy storage in the panel....ain’t no free lunch

glad the conversation has returned to civility....

And still a formidable contender in the small box category is the BBC LS3/5A! I have a pair of Rogers 11 ohm's. 

Quad 57's still sound more like real music than most of today's overpriced offerings! I own two pairs! 

Large size costs more in all ways more to build, ship, store more materials used more finish applied. It's just the way it is. Size is one of the most costly parts of horns.

There are a lot of examples when amplifier has mixed SE and PP stages.
For example some SET amplifiers have first stage SRPP and PP amplifiers have SE first stage before phase spliter.
Another example - SE preamp and PP amplifiers.
Do all these amplithis have a prominent 5th harmonic in addition to a 2nd and a 3rd ?

In P-P amps what is concerning is what circuits are in the feedback loop. They will affect which harmonics have more prominence. IOW the topology affects the distortion signature. IMO a single-ended preamp can work alright with a balanced amp, but I'm not sure that is what you are asking as P-P amps cover quite a range of circuits!

Great conversation. I think some of you would really be interested in JTR speakers. They were mentioned earlier in the thread. JTR does pro audio as well as some high sensitivity home speakers and subwoofers. I am looking forward to hearing them because of the great reviews I have read.

Hi @atmasphere ,

There are a lot of examples when amplifier has mixed SE and PP stages.
For example some SET amplifiers have first stage SRPP and PP amplifiers have SE first stage before phase spliter.
Another example - SE preamp and PP amplifiers.
Do all these amplithis have a prominent 5th harmonic in addition to a 2nd and a 3rd ?

Regards,
Alex.

My reply is intended to illustrate why & how some high performance speakers are purposely designed knowing that they will have to be low in efficiency because of the laws of physics and engineering. Mini-monitors, like the sealed BBC LS3/5a's and the rear vented KEF LS-50's which have fairly good bass, superb mids & highs but low overall efficiency (82db to 84db)  have obviously earned a prominent place in the market. While higher efficiency spkrs are the ideal it it's impossible to achieve both extended bass & high efficiency is small cabinets. Therefore, if you want a mini monitor to achieve bass the efficiency will have to be limited to about 82db to 85db. The smaller the speaker box, the lower the efficiency must be to avoid sounding thin. The laws of physics & engineering just can't be ignored.  But there are tricks to make it possible for small woofers to perform well in both the bass and midrange in small enclosures. The 1st is the use of small magnets which raises a woofer's Total Q factor(Qts) which partly determines the bass output consistent with Thiele-Small parameters. The 2nd trick is to increase the compliance of the woofer surround which lowers the woofer's resonance frequency. A 3rd way is to select high mass cones effectively raising the Qts and lowering the free air resonance frequency. A 4th method is to use crossovers that simultaneously tame the mid frequencies of the woofer so that they don't drown out the woofer's bass output and at the same time use woofer crossover inductors (coils) with a high enough measured DC resistance to raise the woofer's effective Qts. A 5th method in sealed boxes that accentuates bass but lowers efficiency is to judiciosly overstuff the cabinets with absorbent material eg fiberglas or polyester pillow stuffing which makes the woofer behave as if it's in a much larger box. The stuffing amount can be adjusted to change the Qtc (the box's Qt) ideally between .707 and 1.0. A final thought...complex crossovers can be very helpful in compensating for imperfect woofers, tweeters, & mids but they lower overall efficiency and often tamp down dynamics. Obviously there's lots to consider.

My reply is intended to illustrate why & how some high performance speakers are purposely designed knowing that they will have to be low in efficiency because of the laws of physics and engineering. Mini-monitors, like the sealed BBC LS3/5a's and the rear vented KEF LS-50's which have fairly good bass, superb mids & highs but low overall efficiency (82db to 84db)  have obviously earned a prominent place in the market. While higher efficiency spkrs are the ideal it it's impossible to achieve both extended bass & high efficiency is small cabinets. Therefore, if you want a mini monitor to achieve bass the efficiency will have to be limited to about 82db to 85db. The smaller the speaker box, the lower the efficiency must be to avoid sounding thin. The laws of physics & engineering just can't be ignored.  But there are tricks to make it possible for small woofers to perform well in both the bass and midrange in small enclosures. The 1st is the use of small magnets which raises a woofer's Total Q factor(QTS) which partly determines the bass output consistent with Theile-Small equations. The 2nd trick is to increase the compliance of the woofer surround which lowers the woofer's resonance frequency. A 3rd way is to select high mass cones effectively raising the QTS and lowering the free air resonance frequency. A 4th method is to use crossovers that simultaneously tame the mid frequencies of the woofer so that they don't drown out the woofer's bass output and at the same time use woofer crossover inductors (coils) with a high enough measured DC resistance to raise the woofer's effective QTS. A 5th method in sealed boxes that accentuates bass but lowers efficiency is to judiciosly overstuff the cabinets with absorbent material eg fiberglas or polyester pillow stuffing which makes the woofer behave as if it's in a much larger box. The stuffing amount can be adjusted to change the QTC (the box's QT) ideally between ,707 and 1.0. A final thought...complex crossovers can be very helpful in compensating for imperfect woofers, tweeters, & mids but they lower overall efficiency and often tamp down dynamics. Obviously there's lots to consider.

@mijostyn --

Just looked over at your profile, and you have an impressive set-up from what I can assess (haven’t heard it, obviously, but I’m sure it’s sonically extremely capable). You wrote "Sound Labs" being your main ESL’s, but the profile reads Acoustat’s - which is it? A friend of mine have a pair of older Acoustat’s, smaller than yours (can’t remember the model number), and they’re very coherent sounding, transparent and informative. Having a single transducer cover the entire range from 100-20kHz is an impressive, and not least an important feat when done well; being ESL’s I can only imagine that it is. Moreover being a line source (as such) from floor to ceiling adds to their traits.

Personally and over time I’ve never been fully convinced or taken the type of presentation from the line sources I’ve heard, but as executed in your case, with subs augmentation and the ESL’s relieved below 100Hz to boot, I may feel differently about it. You don’t shy away from cone area with the subs either, and with lots of power at hand from your QSC’s bodes well for headroom and clean bass. Have you tried spreading out your subs DBA-style? Not meant as an advocation of mine, but just curious what you’ve come to find here, in case.

I can certainly relate to where you’re getting at with your set-up imagining a large, enveloping and coherent sound field that’s fairly uninhibited LF-wise as well. It’s not at all unlike what I’m after, though quite obviously very differently executed.

@johnk --

Community made the best horn designs even today we haven’t equaled what they offered during the 1970s. Their multi cells are by far the most advanced on earth the leviathan and the radials all not bettered by modern designs. I have been beta testing for a few companies they know I have many horn types to use the horns they made to go with there new designs are just based on older designs but are smaller. So while it is possible to design better horns today we don’t because of the size and costs to build them.

It’s the recurring roadblock for sound reproduction in a home environment: size. JMLC horns (with the proper calculated horn profiles, that aren’t mostly used) are great, I find, and don’t skimp on size. The upcoming bigger EV horns of mine (HP9040) mayn’t dance with the best, but I’d wager they bring about advantages (compared to smaller, more modern horns) precisely because of their size and controlling directivity as low as they do. Getting a closer coverage pattern match over the cross-over range arguably is one of the "macro parameters," and being successful seeking out and attaining others as well is what it’s about, basically - to me, at least. Size restrictions keep one from hitting many of said parameters, but more than cost considerations it appears size itself (and the change in design narrative) is the primary obstacle.

Most speakers of low sensitivity have more elaborate crossovers to even out the frequency response.  It is just an engineering tradeoff, but since more power is needed, anyway, better drivers, also of low sensitivity can be used.  When I switched from a 200 wpc Phase Linear amp pushing stacked Advents, I was in heaven, at least until I heard a Threshold amp and Audire amp drive really good speakers,

Community made the best horn designs even today we haven't equaled what they offered during the 1970s. Their multi cells are by far the most advanced on earth the leviathan and the radials all not bettered by modern designs. I have been beta testing for a few companies they know I have many horn types to use the horns they made to go with there new designs are just based on older designs but are smaller. So while it is possible to design better horns today we don't because of the size and costs to build them.

They are very interesting speakers and IMHO very nice looking. It is a world away from what I deal with.

My approach to subwoofers is totally different to Phusis. His are 7 dB more sensitive but much larger 20 cf to my 1.9 cf ! I have to use four of them to create a line array and each one has two 12" drivers in it. They require power and a computer to control them but they will go flat down to 18 Hz. The enclosures being small and cylindrical are very stiff and the cabinet resonance is way above the operating range and very well damped I can't even see it running them full range. They break up at 2000 Hz. Actually, I should say "it" not "they" as there is only the mule built. The idea was to minimize enclosure vibration and resonance using woodworker friendly materials. The woofers are in phase and located at opposite ends of a 28" cylinder. Their reactive forces cancel out keeping the affair from shaking. The walls vary between 1 7/16 to almost 2". I have not weighed it yet but it is darn heavy. Definitely a two person lift. Now I have to build the other three and finish them all in polyester satin black. They should have no problem keeping up with the Sound Labs.
Psyched to say the least. ESLs are so simple relative to other designs. There really is not much to them. Working around their limitations is pretty easy. You just need the right amp and you have to take the low bass away from them. They will do it but it so compromises them.

@atmasphere --

The T-1 has a 250Hz horn, using a 4" compression driver that is field coil powered. The diaphragm is made of beryllium and employs a Kapton surround, which is how it goes so low without breakups. I don’t know the spec on the T-1 but I do know that the 3" diaphragm in the T-3 has its first breakup at 35KHz. The breakups are a pretty big deal; I got to hear what the regular TAD diaphragm did compared to the CAL unit and it was not subtle- much smoother!

The combination of larger VC diameters, beryllium diaphragms and being field coil powered sounds like a recipe for sonic excellence. That’s next level rocket stage for sure..

@phusis The T-1 has a 250Hz horn, using a 4" compression driver that is field coil powered. The diaphragm is made of beryllium and employs a Kapton surround, which is how it goes so low without breakups. I don't know the spec on the T-1 but I do know that the 3" diaphragm in the T-3 has its first breakup at 35KHz. The breakups are a pretty big deal; I got to hear what the regular TAD diaphragm did compared to the CAL unit and it was not subtle- much smoother!

@mijostyn --

.. my set up is fully active but you could never use SET amps in my system. I really require big amps. I was talking in the context of very efficient speakers. SET amps are very romantic and if that is what you like then by all means but a class A SS amp particularly a Pass or Curl design is going to be more accurate, have more control over the woofer and still have a little of the romance. I would also rather buy records than $1000 tubes.

SET’s isn’t the way I’m heading either, but some of the best of them I’ve heard (300b-based by David Wright of UK) didn’t sound "romantic" per se; just natural, as I described earlier. The Ongaku (211 tubes) is warmer sounding to my ears (as is most Kondo), but I’m thinking whether this character has to do with the silver transformers mainly (instead of cobber).

@johnk --

Costs for high-eff drivers are greater, Size high- eff designs are larger can cost more to build ship take up more room at dealers and distributors.

Relative to driver diameter and weight I don’t see high eff. pro units being notably more expensive vs. hi-fi dittos, although of course they’re usually larger and take more power which then reflects their price. My main issue is taking this out of proportion with high eff. boutique drivers from the likes of TAD ($2,000 per 15" unit as mentioned by Ralph), older WE’s and Vitavox (both crazily priced), as well as (to a lesser degree) Great Plains Audio/older Altec’s or other. Brands like B&C, EV (older units in particular), RCF, 18sound, BMS and others are more fairly priced, and still great high eff. pro drivers. Some if not most of you guys commenting on this make your own speakers or are affiliated with people who are, and the drivers being used as examples here mayn’t be more widely representative.

And I will also add that buyers don’t want changes. And they have been told for decades that all horns are problematic and they bought the line that small is better. Face it would you buy a toaster that was better than others but was larger and looked not like a toaster? We are simple tools and want to blend in and conform with others not to stand out.

Well put - completely agree. I’m sure most audiophiles would find my using pro cinema speakers and 20 cf. horn variant subs to be if not laughable, then oddly.. different. Your horn set-ups are much larger still. Be that as it may I’m not going back to smaller, low eff. all-direct radiating boxes.

@atmasphere --

Thanks for your very interesting elaborations on SET’s and distortions types.

As I understand it, my speakers are currently about $33,000/pair. Mr. Mears is correct. The coupling was optimized on a computer and the result is very smooth and seamless. People often comment on hearing T3s (and T1s) that the speakers sound more like ESLs in that they are so fast and seamless. IOW no ’horn artifact’ at all.

I must have thought of the T1.5’s/T1’s re: price. $33k/pair or thereabouts for the T.3’s isn’t cheap by any stretch of the imagination, but compared to many other offerings in high-end and the JBL Everest’s mentioned earlier I’d say, from what I’ve read by you and others, that they are very fairly priced in light of the drivers used (Alnico and field coil magnets), overall finish/use of hardwood and being handmade.

@mijostyn --

Phusis, those are quite some speakers you have. What are you using for subs? They must be unusual to keep up with your main speakers.

Thanks, I think ;) The horn mounted on top of the EV main cinema speakers is the HP940 and crossed at just below 800Hz. Within a few weeks they’ll be replaced with the (much) bigger HP9040 horn using the existing (and still fitting) DH1A comp. driver. The EV bass cabs are meant for use with both of these horns, the entire speaker called either TS940D or TS9040D depending on the horn mounted to them. The intention using the bigger horn is to control directivity all the way down to the cross-over point (as low as 500Hz, if desired), and thus make for a somewhat better coverage pattern transition between the twin 15" woofers and the horn. The smaller horn I use now has a mode between 2-3kHz, whereas the bigger one moves it down to ~500Hz. Choosing a cross-over in the 650-700Hz region with the same steep slopes should avoid any outright mode issues. We’ll see.

The two subs are tapped horns (called "MicroWrecker") fitted with a 15" B&C woofer each (15TBX100). Tuned at 23 to 24Hz, 97dB sensitive, and taking up 20 cf. per horn. They are unusual, yes, not what one typically sees in hi-fi, if rarely at all. They sound different to direct radiating twin 18" ported subs (their effective air radiation equivalent) being more smooth, enveloping and effortless sounding. They fit the EV mains great.

@atmasphere --

Like any other tech, horns have benefited dramatically from the aid of computer optimization. I owned Altecs, Klipsh and EV horn systems and there’s no way I would go back to them; as you say they were very colored (and lacked bandwidth, particularly in the bass). The Classic Audio stuff is a different beast altogether- neutrality is a good descriptor. The speakers produced by Audiokinesis using waveguides had a similar neutrality- you found yourself involved with the music rather than the speakers.

The David Gunness developed HP-series horns from Electro-Voice (based on Don Keele’s Constant Directivity design’s from the 70’s) aren’t of the earliest "squeamish" types, and supposedly the bigger variants HP640 and 9040, as well as the earlier HR-series, are very well liked by audiophiles-into-horns. I’m sure the Classic Audio horns are better being newer, computer aided developments as well as their more inert hard-wood material use, albeit smaller; this is where physics and sheer size comes into place for directivity control down to the x-over point, and that’s not trivial. Compromises, compromises..

Some of the older Horn loaded speakers like Altecs and JBLs that I had a lot of experience with when I was young were all very colored. Back then K horns were the best but they did not image well and though not as colored as the other horn speakers you still knew you were listening to horns.

Like any other tech, horns have benefited dramatically from the aid of computer optimization. I owned Altecs, Klipsh and EV horn systems and there's no way I would go back to them; as you say they were very colored (and lacked bandwidth, particularly in the bass). The Classic Audio stuff is a different beast altogether- neutrality is a good descriptor. The speakers produced by Audiokinesis using waveguides had a similar neutrality- you found yourself involved with the music rather than the speakers.

@atmasphere wrote: 

"Harmonic orders above the 2nd decrease rather slowly as compared to a circuit that has a ’cubic non-linearity’ (produces the 3rd as the primary distortion component). An amplifier that has this quality has its distortion decreasing much faster as the order of the harmonic is increased! This is important since the ear uses higher ordered harmonics to sense sound pressure (and assigns the quality of harshness and brightness to them). IOW, an amplifier with a cubic non-linearity will sound more detailed (because distortion masks low level detail) and **smoother** because the higher ordered harmonics are at a lower level. 

"In terms of circuit design an amplifier with this characteristic must be fully differential and balanced from input to output. In this way even orders are cancelled with each stage in the amp (instead of being compounded), leaving the 3rd as the primary distortion component, at about 1/10th what you would get with a single-ended circuit, assuming that neither employs any feedback." 

VERY INTERESTING!! 

I think this explains part of the difference I hear between your amps and good single-ended triode amps, in particular:  

" An amplifier that has this quality has its distortion decreasing much faster as the order of the harmonic is increased!"  And, 

"...even orders are cancelled with each stage in the amp (instead of being compounded), leaving the 3rd as the primary distortion component, at about 1/10th what you would get with a single-ended circuit...". 

Thank you for putting this down in writing.  I copied it to a file so I can find it again.  

Duke 

My opinion of speakers is very jaded now. I've been living with large ESLs for so long that most other speakers sound.....small, even if they go loud.
I find horn loaded designs very interesting but I have yet to hear one that I would consider. I have not heard Ralph's speakers, I'm sure they are excellent. I have heard a few at shows that were not impressive, but what ever is at a show. Some of the older Horn loaded speakers like Altecs and JBLs that I had a lot of experience with when I was young were all very colored. Back then K horns were the best but they did not image well and though not as colored as the other horn speakers you still knew you were listening to horns. Frankly, I liked the Cornwall better than the K horn.
I owned Heresy's for 10 years and for a teenage rocker with a Dynakit Stereo 70 they were great until I got Large Advents and a Phase Linear 700. The Advents were more natural and imaged very nicely. 
The Heresy's were the last and only horn speakers I ever owned. The problem is there are painfully few situations where you can hear them in a controlled situation. Some of them are very large, very expensive and few dealers are going to keep them on the floor. 
Those of us here have gone down our own path and become entrenched in one technology or another and always seem to think ours is best.
There is no single best. There are thousands of bests. Since we can't hear all of these systems it is fun to talk about them and hear about what other people are doing. In many instance you might learn something that you can apply to your own system.

@mijostyn  I don't know about Curl's designs so much, but Nelson does make fully differential amplifiers so he is able to take advantage of the lower distortion afforded, and yet have a nice 3rd to help mask the higher orders.

Atma-Sphere that sound great but is way above my pay grade. But, I think you just explained why I like Pass and Curl amps so much?
Phusis, those are quite some speakers you have. What are you using for subs? They must be unusual to keep up with your main speakers.

What are the T3’s sold for - close to $100k? Re: mentioned "better interface" it reminds of what Simon Mears told me of the importance of the coupling between the compression driver exit and the throat of the horn it’s mounted to; transition, transition, transition - as he put it.

@phusis As I understand it, my speakers are currently about $33,000/pair. Mr. Mears is correct. The coupling was optimized on a computer and the result is very smooth and seamless. People often comment on hearing T3s (and T1s) that the speakers sound more like ESLs in that they are so fast and seamless. IOW no ’horn artifact’ at all.


Regarding the SETs, IMO/IME their main advantage is that as the power is reduced, the distortion decreases linearly to unmeasurable. This is important because (to trot out an old expression) it really is all about that first watt. But just so you know, this character is not unique to SETs, although it is rare in push-pull amplifiers. But you can imagine since I’m writing this that I know of a few amps which share this important characteristic. But one **disadvantage** of SETs is that their primary distortion product is the 2nd harmonic, as well as low power and troubles making bandwidth due to the output transformer.


Now we all know that the 2nd harmonic is innocuous in that the ear is insensitive to it (and because the ear converts all forms of distortion into tonality, it assigns the quality of ’warmth’ and ’bloom’ to this one). However, the ear assigns this same quality to the 3rd harmonic as well, but there is an important distinction. Circuits that have a 2nd harmonic as their primary distortion product mathematically have what is known as a ’quadratic non-linearity’. Its not so important to know the math, but if you feel like working it out what you will see is that harmonic orders above the 2nd decrease rather slowly as compared to a circuit that has a ’cubic non-linearity’ (produces the 3rd as the primary distortion component). An amplifier that has this quality has its distortion decreasing much faster as the order of the harmonic is increased! This is important since the ear uses higher ordered harmonics to sense sound pressure (and assigns the quality of harshness and brightness to them). IOW, an amplifier with a cubic non-linearity will sound more detailed (because distortion masks low level detail) and **smoother** because the higher ordered harmonics are at a lower level.

In terms of circuit design an amplifier with this characteristic must be fully differential and balanced from input to output. In this way even orders are cancelled with each stage in the amp (instead of being compounded), leaving the 3rd as the primary distortion component, at about 1/10th what you would get with a single-ended circuit, assuming that neither employs any feedback.

Now if you mix single-ended and push-pull, you wind up with a prominent 5th harmonic in addition to a 2nd and a 3rd (put another way an amp like this has **both** cubic and quadratic non-linearities). This is why many people prefer SETs, but those same people find that if they hear an amp with similar concepts (triode, class A, zero feedback) executed fully differential, that it has all the desirable properties of SETs without a downside. BTW this difference is easy to hear (its not subtle) and of course its also easy to measure.

Costs for high-eff drivers are greater, Size high- eff designs are larger can cost more to build ship take up more room at dealers and distributors. And I will also add that buyers don't want changes. And they have been told for decades that all horns are problematic and they bought the line that small is better. Face it would you buy a toaster that was better than others but was larger and looked not like a toaster? We are simple tools and want to blend in and conform with others not to stand out.

@phusis, yes, my set up is fully active but you could never use SET amps in my system. I really require big amps. I was talking in the context of very efficient speakers. SET amps are very romantic and if that is what you like then by all means but a class A SS amp particularly a Pass or Curl design is going to be more accurate, have more control over the woofer and still have a little of the romance. I would also rather buy records than $1000 tubes.

@atmasphere --

I have. The Classic Audio Loudspeakers are **easily** in the same league. If I had to compare, the CAL is a bit smoother, owing to a better interface between the throat and horn, resulting in far less artifact. The field coil compression driver is also a higher performance bit of kit.

I would expect no less of the T3’s, or CAL speakers overall. The DD67000’s are great speakers, and somewhat better than their smaller K2 S9900 sibling than I initially thought, but there’s a shred of "splashy-ness" to their midrange than when not there lends a smoother imprinting, as you point to with the CAL’s. What are the T3’s sold for - close to $100k? Re: mentioned "better interface" it reminds of what Simon Mears told me of the importance of the coupling between the compression driver exit and the throat of the horn it’s mounted to; transition, transition, transition - as he put it.

@mijostyn --

Phusis, I would bet there are a host of SS amps that would handily out perform any SET amp. Any small class A Pass amp for sure. JC1’s definitely and most probably all of Atma-Sphere’s amps. Like trying to race a 911 in a VW Bug.

"Outperform;" per your ears or more from a theoretical perspective? And in what context of speakers? SET’s, great ones at that, are dependent on very high efficiency speakers (and not too heavy an overall load) to perform their best, at least from my chair. Paired properly as such I find the combo is magical, or just rather natural and uninhibited sounding. I suspect most haven’t really auditioned them in the context I refer to, and if they have the speakers here (typically all-horns) may been so far from their usual sonic menu that an assessment would veer closer to their imprinting, and not the amp(s). Myself I certainly can’t complain using a variety of SS-based amps in my set-up, but I wager a substantial aspect of this rests in it being a fully active configuration.

Phusis, I would bet there are a host of SS amps that would handily out perform any SET amp. Any small class A Pass amp for sure. JC1's definitely and most probably all of Atma-Sphere's amps. Like trying to race a 911 in a VW Bug.

They are also a lot nicer looking!!

How would you compare them to the likes of JBL Everest’s (DD67000, or one of the other two variants), if you heard them?

I have. The Classic Audio Loudspeakers are **easily** in the same league. If I had to compare, the CAL is a bit smoother, owing to a better interface between the throat and horn, resulting in far less artifact. The field coil compression driver is also a higher performance bit of kit.

@atmasphere --

They are TAD 1602s. Pricey, and a bit different from the EV. I forgot to mention they have Alnico magnets.FWIW the midrange driver employs a field coil and a beryllium diaphragm with a Kapton surround. They are Classic Audio Loudspeakers model T-3. I had the cabinets custom-built by CAL to be a bit taller than stock so they are flat to 20Hz.

Never heard Classic Audio Loudspeakers, for that I imagine I'd have to go abroad to the US. From what I can tell they must be excellent speakers. How would you compare them to the likes of JBL Everest's (DD67000, or one of the other two variants), if you heard them?

@clearthinker --

If I understand you correctly, you make a new point after 119 posts to this thread:
A large majority of audiophile speaker designs are low efficiency because a large majority of audiophiles don’t like some of the sound characteristics of high efficiency speaker designs. Or, as you say, they think they don’t.

It’s certainly a type of sound quite a few audiophiles don’t warm to particularly, likely because of another type of sound they’re typically exposed to or, yes, they dismiss this segment of speakers out of hand/sans 1st hand experience. Or, they may simply not like the ones they heard, for whatever reason, which is perfectly fair.

Habitual use is important to stress here, I find. It’s a bit akin to the sonic difference between passive and active speakers perhaps; using the same speakers in one and the other configuration I find the former is generally the more euphonic sounding, softer, less clear, less resolved and less transiently "snappy"/more smeared. To some passive here is the more "musical" and warm variant and thus more pleasing, whereas to others active is the musically more honest, resolved, transparent and less bottlenecked presentation and therefore what they prefer. Below is an excerpts from a JBL K2 S9800 review that highlights a universal characteristic of horn-loaded, large-woofer speakers and their type of presentation:

Horn speakers are not to everyone’s taste, and while a number of visitors agreed that this speaker seems remarkably free from the vices normally attributed to the breed, some listeners might find it a little too ruthlessly revealing, preferring something rather more laid-back and restrained. While the K2 does have some aggressive tendencies when worked hard, taking no prisoners among poor-quality software, sources, or amps, its ability to "suck in" the listener and create involvement in even unfamiliar material is unparalleled in my experience.

One aspect of performance that’s somewhat different from the norm lies in the way this speaker interacts -- or rather doesn’t interact -- with the listening room. Most speakers have relatively wide dispersion in every direction, so that although the sound you hear is dominated by the direct sound from speaker to listener, it is richly augmented by room-reflected sound. However, both the K2’s large bass/mid driver and its mid/treble horns seem to have comparatively narrow dispersion through the midrange and treble, so there’s less room sound than is usually the case.

In that respect the K2 shows certain similarities to dipole panel loudspeakers such as the Quad ESL-988, which have a figure-eight-shaped distribution. This means that you hear more of the direct sound coming straight to you from the loudspeakers, and proportionally less sound reflected from the walls, floor, and ceiling than you would using a speaker with a much wider dispersion characteristic, such as the B&W Nautilus models.

Whereas the Nautilus 800 will tend to fill the room more, creating a strong impression that the musicians are actually sitting right there, the K2 and Quad provide more of an "open window" onto the recording sessions, revealing more of what the recording engineer intended, but less of the illusion that musicians are actually playing in front of you. This is neither praise nor criticism, as there’s neither right nor wrong here, but it is a relevant observation that has a significant impact upon the character of the listening experience.

This quality undoubtedly contributes to the very precise imaging, alongside this bulky speaker’s surprisingly good transparency. It’s not quite a match for the best dipole panels here, but is rather better than most box loudspeakers in this regard.

https://www.soundstageultra.com/equipment/jbl_k2_s9800.htm

If that’s correct, I wonder how much the sound characteristics of the puny 10 watt often SET amps that are commonly used to drive high efficiency designs have got to do with it. Like some of the posters here, I certainly don’t like them. Would they sound different (better?) driven with a high current amp with a big power supply, even if the wick has to be turned a long way down? They would be under better control, particularly in the bass where most of the problem lies.

No, through high quality high efficiency speakers, not least very high eff. (i.e.: from ~100dB’s on up) all-horn speakers I find great quality SET’s to bring out the best in them. There’s a combination here of aliveness, vibrancy, uninhibited yet naturally warm presence and a lit-from-within sense of presentation that’s quite unique and utterly beguiling. To take full advantage of SET’s you need very high eff. speakers so only to use as little as possible of the few watts available, and stay in the very low range of distortion. By comparison some Solid State amps (less so SIT’s) with the same very high eff. speakers sounded somewhat grey-ish, a bit mechanical, less folded-out and just downright flat.

Myself I previously used a pair of very high eff. all-horn speakers, but for some reason I ended up not giving them the SET they deserved and instead veered more towards active with SS amps (30 watts pure Class A to the horns and Class D variants in the half kW range further low) and horn hybrid speakers that I use now. Again the passive vs. active sonic marker applies, though literally here; active with the horn hybrids I’m using now (~100 to 110dB’s sensitivity) infuses a vitality and ignited-ness, even with SS amps, that to me is akin to the sonic imprinting of SET’s + all-horns, the difference though being an added sense with the active config. of evermore ease, resolution and (a sense of) unlimited power delivery.

Alex
We shall agree to differ and remain friends.
Doesn't happen much here I know.
But I'm hoping for some changes and to expose the trolls and the bigoted and closed minded for what they are.  Know who they are but don't name names.
Different opinions honestly and humbly held are the lifeblood.

My first thread has been quite a success with 120+ contributions.
Thanks to all.

These are field coil-fitted drivers, right? In that presumable case their price, though way above what needs to serve a fitting purpose here, makes a little more sense.

@phusis 

They are TAD 1602s. Pricey, and a bit different from the EV. I forgot to mention they have Alnico magnets.FWIW the midrange driver employs a field coil and a beryllium diaphragm with a Kapton surround.  They are Classic Audio Loudspeakers model T-3. I had the cabinets custom-built by CAL to be a bit taller than stock so they are flat to 20Hz. 

@mijostyn ,

Maybe I'm missing something.
I have a resistor attached on chassis of my phonostage. The chassis is on open rack and good ventilated and it is much bigger than any voice coil. The power dissipated on this resistor just 8 Watt. And this power 8W make a big chassis temperature about 20C hotter compared the room temperature. It is not bad, but the voice coil is inside the little box and much smaller. 

Regards,
Alex.

Hi @clearthinker ,

The low sensitive speakers sound boring because they eat microdynamics that exist at live performance. They are simply much less musical.
The audio business (producers, sellers, reviewers) cheat you and sell you cheaper stuff as something good and fancy. 
I'm sorry. I don't want to be offencive. You asked what do people think about this topic. It is just my honest opinion.

Regards,
Alex.

Alex, who said you need 100 watts to drive a voice coil at normal volume? First of all music is dynamic and made up of a palette of frequencies. You don't listen to a sine wave. Even with less efficient speakers RMS wattage is going to be pretty low at normal listening levels maybe 10 watts at most while still hitting peaks of 100 watts. A woofer that is well ventilated is it's own fan. That is the way it is designed. The drivers I use can take 600 watts continuously indefinitely. I doubt they have ever seen 100 watts RMS 

@ctsooner wrote: 

"it's simply not true to make a statement that a higher sensitivity speaker is more costly to make..."  

Based on sixteen years of manufacturing fairly high efficiency speakers, I disagree.  

The enclosure is usually the most expensive component and high efficiency calls for large, typically labor-intensive enclosures.  As enclosure size goes up so does the enclosure cost, even moreso if the larger enclosure is also more complex.  And enclosure size has two hidden costs:  Shipping cost and opportunity cost.  The latter is due to the sheer amount of space the speaker takes up in a showroom.   

Many other costs also are typically (though not always) higher with high efficiency, but enclosure cost is always higher for the larger enclosure assuming an equivalent enclosure build technique, and dramatically so for dramatic size differences, and even moreso as the enclosure complexity increases.  Sort of like the cost of building a house goes up as the size and number of rooms (complexity) go up. 

Duke  

Thanks Alex.
In terms of my original question, I am rather afraid you are saying high efficiency speakers are rare because they're rare.  That won't get us far.

And on your second point. Or are systems electronics commonly used to drive them often flawed?  I have always felt that in many cases it is the flaws that make them less 'boring'.

Hi @clearthinker ,

The problem is because high efficiency speakers today are rare a very few audiophile had opportunity to listen good implemented system with high efficiency speakers.
The other factor is the average high efficiency speakers are more transparent and shows more flaws of the system electronics.

I remember, for me it was shock when I at first listened a good high efficiency speakers system back at 2002. I was like a medieval knight who saw an army armed with aircraft and tanks.
Since that day I heard many systems, but any low efficient system didn’t impress me. In the best case they sound just OK. They also sound more boring in therm of music.
If any thing impressed me - it always was high efficient system!

Regards,
Alex.

@atmasphere --

...  A higher efficiency driver that can make 98dB with 1 watt and is 8 ohms is a different beast altogether. This describes the 15" drivers in my speakers at home and they cost $2000.00 each.

These are field coil-fitted drivers, right? In that presumable case their price, though way above what needs to serve a fitting purpose here, makes a little more sense. 

You can find 15" drivers that cost $200.00. I'd be very interested to learn of a 15" with the same 22Hz free air resonance and 150 watt power handling that cost $200, or even $600.00. Have at it! But I think you'll find that such simply does not exist.

The Electro-Voice DL15W woofer/mids of my pro cinema speakers are 97dB sensitive/8 ohms, have a 21Hz free air resonance, and take 400 watts long term. And there are two of them in each speaker (+ subs). Being they're high-passed below ~85Hz further adds to their power handling. I'm unaware of their price being they're discontinued, but my guess is around $300 retail per driver at the time.

Thank you phusis.
If I understand you correctly, you make a new point after 119 posts to this thread:
A large majority of audiophile speaker designs are low efficiency because a large majority of audiophiles don't like some of the sound characteristics of high efficiency speaker designs.  Or, as you say, they think they don't.

If that's correct, I wonder how much the sound characteristics of the puny 10 watt often SET amps that are commonly used to drive high efficiency designs have got to do with it.  Like some of the posters here, I certainly don't like them.  Would they sound different (better?) driven with a high current amp with a big power supply, even if the wick has to be turned a long way down?  They would be under better control, particularly in the bass where most of the problem lies.

@mijostyn ,

How can you cool the driver voice coil if you need 100 watts to listen to it at normal volume?
Even if you put a dedicated fan it wouldn't help!
In contrast if you need just 5 watt, you don't have this problem even with the most straightforward driver coil design.

Regards,
Alex.

 

@clearthinker --

I'd agree with poster @ctsooner here. Pro manufacturers, the more or less sole supplier for this segment, have developed and build their high efficiency drivers for decades now, and relative to their size, material use and R&D are actually very fairly priced compared to "hi-fi" drivers in general. Boutique high eff. drivers with AlNiCo magnets or field coils housed in luxuriously finished cabinets and all will always be (much) more steeply priced, though are hardly representative in this context. Fortunately they needn't be that pricy to show the merits of high efficiency designs. 

Largely it comes down to size, sound type and association, I believe; high efficiency speakers are closely related to if not directly derived from the pro sector (certainly driver-wise), a sector audiophilia isn't too comfortable with, and coupled with their sonic imprinting as typically more direct, dense, dynamic and present (also due to their dispersive nature) may not appeal to audiophiles and their more widespread exposure to a "softer," thinner, more laid-back and "reverberative" sound (some even feel the dynamics capabilities of high eff. speakers to be "exaggerated"). I'm sure many regard high eff. designs as more "brute" and lacking subtlety, not least when linked to the pro/studio environment, but it's often very far from the truth (did I mention conjecture?). As if sheer volume wasn't an issue already high eff. into the lower octaves takes some very serious size - there's no noodling around that requirement either - and when most have jumped ship with the high eff. main speakers long ago it's even more rare seeing subs in this category of speakers. 

What's most popular in hi-fi is also a popular narrative, and that's an inertia not easily brought to hold.