Audio Asylum Thread Printer
Get a view of an entire thread on one page
|For Sale Ads|
How do horns and high efficiency effect the sound as compared to speakers without these attributes ?
Paul and others mention dynamics and I would underline that word as well. What is very hard to picture for many is why that is.
If one were to take a sound level meter and measure a reproducible event, one finds that on the "fast" setting and C weighting, the meter usually agrees very well with subjective loudness.
One might conclude that the requirements for a loudspeaker to re-produce "that sound" would be what the SLM indicates (accounting for distance etc) and while this often assumed, reality is very different partly because what "we" experience as loudness is NOT very tightly tied to what it takes to reproduce a given signal. The VU meter (Volume Unit) was conceived to indicate loudness but it deliberately has a time constant and that is becase our hearing also has a time constant, loudness is not only related to how loud something actually is but also how long it has lasted, to be perceived as loud, it mush exist for some amount of time. Very short and loud sounds we do not judge as being loud when compared to that sound sustained for a second or so.
The measurement or recording microphone on the other hand converts acoustic pressure into Voltage signal with a fixed proportionality so that one can examine the Voltage and knowing the mic pressure sensitivity, calculate the instantaneous pressure at any point in the waveform.
What one finds is that many common sounds (and musical instruments) have shockingly high instantaneous SPL's which do not show up on a normal sound level meter. For example, I have a large B&K sound level meter which has a "peak hold" reading which captures the loudest peak (over some short time) and simply throwing a metal tea spoon on the tile floor in the kitchen produced an instantaneous peak spl of nearly 140dB. It was so short and no doubt much of it very high frequencies that to me it only sounded like one of the kids threw a spoon on the floor.
Point is, to reproduce anything such that is has the original dynamics, one must produce all the annoying instantaneous peaks not just the "Average" level related to subjective loudness or an spl meter.
This kind of dynamic limiting is more common than one might think and sadly, the ONLY way to see if a system is suffering from this is to get an oscilloscope and look at the amplifier output. Any instantaneous clipping or limiting in any part of the chain, will be visible if preset on the output terminals. In a more complicated sound system, this is one way to verify the gain structure (gain of each element in the chain) is acceptable.
So (looking only through the window of dynamics) if one has high efficiency speakers with a more powerful power amp, you're more likely to be reproducing the instantaneous peaks compared to a low power amp on low efficiency speakers.
Think about it, if one had a good recording that had 20dB or even 30dB difference between the peak and average level, that also means that the instantaneous peaks are 100 to 1000 times louder than the average level that one would read with an integrating meter, like a sound level meter or our sense of loudness. These recordings sound dynamic, not loud because the loud portions are too short.
Funny thing too, being so short, with instantaneous clipping, one cannot hear it as a flaw and is only detectable when compared to the same music without it, or visible on an oscilloscope.
There are many ways to get higher efficiency and like low efficiency systems, all have engineering difficulties
Further, my old mentor John Meyer (meyersound.com) told me way back in the day, when I was a youngin' and budding audio guy, that a very high quality audio system can be more damaging to our hearing than a lesser quality one, because...
A higher quality, more accurate, system will reproduce the instantaneous peaks which our ears don't have time to re-act to and put the biological "limiting" system into action. THAT was a very valuable piece of information.
This is a confusing area, yes TTS (temporary threshold shift) does take some time to work (particular muscles in ones ear clamp down) however an interesting paradox exists.
A paradox dealing with an aspect of loudspeaker system performance not measured or discussed often.
Lets say you had a stadium and you had to comply with new laws that require sound systems to be intelligible when used for emergency announcements. To pass the intelligibility test they use a language independent test called STIpa.
If one compares an array such as are usually for concert sound or as often installed in a large venue one finds that the myriad of individual sources and individual arrivals in time from each after a single input event produce low intelligibility compared to a system which only has one arrival in time when fed a single event. As the distance from each source to each listener is a variable, a dsp correction for this is only applicable in one location.
While that measure indicates being faithful to the time information is among the things important for intelligibility, a different thing is found with music when switching from the array to a single source, to reach the same satisfying subjective volume, requires about -10dB lower actual average level or more.
There are inherent reasons those arrays sound like they do and not hifi.
It's the difference between "They are here" (the musicians that is) with horn speakers, and "You are there" (in the acoustic space but a bit removed from the musicians) with direct radiator speakers, with thanks to Herb Reichert for stating it like this back in his Sound Practices days. A direct radiator/non-horn speaker is a tuned system with the moving mass of the cone and voice coil adjusted to dominate the performance for acceptably flat performance. With a horn loaded driver the acoustic resistance of the horn dominates the performance, which is a more stable load to drive in the pass band of the horn than the mass dominated direct radiator driver. The result of this is less phase distortion with a properly designed horn compared to a direct radiator design. If you don't believe me check out John Atkinson's phase and electrical impedance measurements of the Sonus Faber Venere 1.5 loudspeaker versus the phase and elec. imp. graph for the JBL Synthesis 1400 Array BG (which has a horn loaded mid and tweeter but a direct radiator woofer). The phase and electrical impedance of the Sonus Faber are all over the place in the critical midrange and well into the high frequencies, and the electrical impedance actually goes way off the graph between 1K Hz and 10 K Hz centered at about 3K Hz! The JBL is drastically flatter between 2K Hz and into the high frequency region. As the JBL has a direct radiator woofer it has quite a bit of variance in the phase and electrical impedance below about 200 Hz where the non horn woofer crosses in.
Because a horn has a higher efficiency advantage over a direct radiator design, horns can produce sound with less Watts and less driver excursion. Lower driver excursion can produce less Doppler distortion where the movement of the diaphragm back and forth with lower frequencies can interfere with higher frequencies being reproduced.
So with horns you get more life like dynamics, but horns may come off as "drier" too, but much of the acoustic space some hear and praise in a good direct radiator design may be due to phase distortion added by the speaker which is really not part of the recording.
> It's the difference between "They are here" (the musicians that is) with horn speakers, and "You are there" (in the acoustic space but a bit removed from the musicians) with direct radiator speakers,
Well yes to a certain extent if comparing with box mounted conventional drivers. However your case doesn't apply if horns are compared with for example electrostatics. Would you care to try that comparison? I'd suggest the "They are here" effect is as much or more so applicable to ESLs.
Truth is that horn drivers are the same mechanically as conventional drivers - they just don't need to generate so much physical movement to produce the sound we hear. That's a good thing but horns have no other significant advantages. I agree they sound more lifelike than most box speakers, but I'd suggest that ESLs are possibly more lifelike still.
I have Avantgarde horns (the ones voted by Stereophile as "Joint Loudspeaker of the Year") and large panel Quads. After many years thinking that the horns knock the socks of other speakers, I'm now realising that this isn't entirely true. For my listening, I'd suggest that the horns offer a more "exciting" listen (great for non-classical music) but the Quads are faster still and "closer to the original sound" and better for classical music. A generalisation maybe, but that's my take on it as owner of quite pricy examples of both types.
There are other advantages. Less cone excursion is a big one, but how about less thermal energy in the voice coil, reduced rear-wave in the enclosure (horns are loud at the 'mouth'), less back-EMF to the amp and much less amp power needed -no more complex circuits.
Plus controlled radiation -fewer problems with wall reflections.
A compression-horn is better than most people think.
Well JP, those are advantages that horns have over conventional box speakers as I acknowledged earlier, but none are valid as advantages over electrostatics. I'd suggest that ESLs take these mechanical advantages a stage further. As I also acknowledge, this doesn't necessarily make ESLs "better" but it probably does make them more accurate.
I respect your support of ESLs - but honestly, they're not more accurate (than horns).
However, along with Magnepan, they held the torch in the dark days of cone speakers, esp. from the 1960s to the 90s. Shocking transparency when they came out.
For one, without a XO, they suffer from Doppler distortion. Some have a XO to the bass but I don't think that saves this problem.
There's low dynamic range -can't be accurate if they're not reproducing live SPLs.
The treble beams.
If used, cone-bass doesn't blend properly with a panel midrange.
They suffer from 'skeletal sound' (have no 'body') as Harry Pearson said.
They rattle and cause self-noise.
They (still) suffer from wall and ceiling reflections, despite being 'figure of 8'.
Unless you have Quads, they don't have point-source imaging.
They're tough to drive.
Their panels damage and don't last forever.
Maybe (some) of these issues were solved. I heard Roger Sander's ESLs are winners.
Were you talking about horns or 'stat panels?
Rooms. Listen to me. Read my profile. Think. Buy "Why You Hear What You Hear" by Eric Heller.
I 'get it' that you're a budding audio guy. Good! But, if you want to be knowledgeable (rather than a 'sheeple'), you have to get your head out of the magazines and read a real book or three. Start with that one, and then, if you're still interested in learning, get Harry Olson's "Music, Physics and Engineering". Trust me, your entire perspective and understanding of sound will change.
One of my points was their 'panels damage', so you're the one not reading.
I read one book each by Eargle and Toole. Then one on mastering audio.
But will any book guide us to the right speaker technology ? How many talk about the virtues of compression-horns ?
Of course on this forum there is polarised support for high efficiency speakers including horns and scepticism of other type of speaker - understandable as most of the posters here (including myself) own these things. Likewise if one goes to the Planar forum, people would generally praise ESLs and criticise horns - again understandably.
However most of what you say about ESLs is simply untrue, at least in relation to the latest Quads. I'm sure other ESLs are equally good or better than Quads.
Like you, I have always discounted Quads as being rather feeble and lacking in bass. However, as a result of a Quad 2905 owner replying to a posting of mine on a completely different subject, I visited his home. I was stunned at the quality of the sound, played at robust levels, in particular the imaging - better than my horns, though both systems need careful speaker placement.
That one audition prompted me into buying a pair. I now have both speakers in my room, each powered by appropriate amps. The horns have 845 monoblocks or Accuphase E-36 Class A and the Quad gollups up the 200 watts from the GamuT. Both are very good systems. Which do I prefer? It's very close - the Quads are undoubtedly better for some music, but the horns are just a tad more "exciting" to listen to on much other music. I'm going to keep the horns, but if I was a classical music fan, I'd keep the Quads.
No polarization -these are very definable problems.
You did the right thing -as panels have all but disappeared from high-end audio. Listeners are using Emerald Physics, ATC, Harbeth, JBL, etc.
Interestingly ATC were the WORST sounding speakers I've ever owned. They were the Active 50s and I bought them to replace KEF Reference 107s. The ATCs were so much "in yer face" you wanted to push them back another 20 feet. I sold them within a few months - the guy who bought them was delighted with them in his rather smaller listening room.
I've found that equipment designed for studio use is often positively unsuitable in the home. The ATCs as I described and the Benchmark AHB2 amplifier which is lifeless, though possibly very accurate. You wanted to turn the volume down and maybe skip to the next track in the hopes of a more exciting listen. I sold that after a few weeks despite it's dead silence - an advantage with horns. It was drearily DULL.
I got the Gamut to replace the Benchmark and even though it's rated as 200 watts, it has gain reduction dip switches and the -14dB setting really gets the horns going. In particular it offers exceptionally detailed bass with the Avantgardes, not heard with any of the SETs or Class A amps I've tried. It's undoubtedly the best amp I've had with the horns.
One of the very few 'high end' amps I have ever liked. Amazingly musical and engaging.
"... only a very few individuals understand as yet that personal salvation is a contradiction in terms."
This flies in the face of reviews, in the past 5 years. "Best speaker they ever heard". There are a number of reviews, BTW..
Even for classical music, people have gone cone or waveguide/horn. Emerald and Harbeth have done best (with former panel owners) but there are more inc. Verity.
Nice info . Other than the wall reflections , what do those advantages actually do for the sound that is heard ?
Well, there might be lower distortion in the motor -as the voice coil stays in the gap. There's also less eddy currents. Due to these things, a speaker doesn't need to be 'damped' so much. Damping keeps the VC cooler, but it might suffocate the signal.
That's just one -but this is not a study of compression horns...
Horns and electrostatics (including planers like the Magnipans) have more in common with each other than they do with direct radiators. R.H. Small commented back in 1969 on the problem of integrating horns or electrostats with direct radiators, and he observed that there would be a 90 degree phase difference between the transfer characteristics of the horns or electrostats verses the mass dominated direct radiators. I've heard Quads a number of times, both re-built originals plus the newer versions at hi-fi shows. They do have horn like dynamics at low to medium volume, but they can't play loud, and they are limited at the bass end too. If you add a direct radiator to supplement the bass you encounter the problem Small talked about. As electrostats are low efficiency and rather difficult to drive (someone once commented that they were like the amp having to drive a capacitor) your amplifier selection becomes limited, and low powered triodes are not going to get the job done. So if your musical tastes allow you to live with these limitations they may be ideal. I've also heard the Avantgardes, and though they sounded good, there is really no contest against the Edgarhorn Titans which are horn loaded all the way down, and this is just my opinion of course. Comparing the Edgarhorns to Quads as to which is "faster", there may be a contest at low levels, but at a loud peak in a classical recording it's really no contest, it takes a horn to catch a horn there. This having been said, I understand your comments as to the Quads compared with the Avantgardes, and I would probably agree if I heard them together at a demonstration.
"They are here" is a mis-application of the phrase. The apparent presence or lack thereof is much more a function of how the recording was made, rather than the type of speaker it's being played on. However, since most horns are more directional than appropriate direct radiators with similar crossover frequencies, there 'is' a difference in the soundfield of the listening room, which may account for the perceived difference regarding the "they are here" aspect. Personally, in a home environment, I prefer direct radiators (or maybe stats if a person has the space for them) because well-designed ones tend to better mimic the radiation patterns of real instruments (this is a generalization, since instrument radiation patterns vary quite lot), thus creating a more natural ambient soundfield.
With regard to high efficiency, this is a sticky wicket. High efficiency has its advantages, particularly for large high-output systems such as for concerts, but, to achieve very high quality sound in a home environment, they also require amplifiers which perform (sound) very good at low output levels. You may already know that many amps have higher distortion in that first 0.5 to 1 watt or so output than they do at higher output levels. Matti Otala and John Curl did landmark research on this back in the early 1970s. So, choice of amplifier becomes much more important.
With regard to horns in general, many are not well-designed or matched with drivers and crossover frequencies which are appropriate. Two of the drawsbacks which result from this are distortion of the air in the horn throat (where the SPL can be VERY high), and acoustic reflections within the horn. These both contribute to what is often termed "horn sound". Some people don't mind it, and actually like it.
Having said that, if a horn sounds fine, then it sounds fine, at least to that listener. I suggest that you look into resources which describe "horn theory" and related physics.
Previously on this forum you told Tom Danley that he should contact a rival who makes array systems and "learn from him". This was a jaw dropping statement (and not in a good way). Or perhaps it was meant to be hilarious. I'm going with the latter just to be safe.
Have you ever heard Tom's Unity horns? If not you could listen to some of the recordings of them Tom has linked to on this forum. Bearing in mind that our ear drums are horn loaded (unless you have had your pina surgically removed, or following an accident), you may want to eliminate that variable by listening with in ear "earbuds". Has the "horn coloration" disappeared? This may establish where the coloration is coming from.
You'll have to give me some context, and tell me what the discussion was at that point in a thread. A URL reference would be even better.
I doubt if I was being unnecessarily harsh on Tom. Probably just a difference of opinion on a topic, but I do tend to not sugar-coat things.
Thank you to all . I think that I understand some it ,probably enough to get me into trouble ! But it does help me to understand what I am hearing in a pair of speakers that I acquired recently .
Another question , if I may...Does anyone manufacture a horn system speaker that would be suitable , size wise , for in-home use that would pass the W.A.F. and be suitable for nearfield listening of @ 6 ft. ? A direct radiator type woofer would be fine . I am intrigued by the mid and tweeter horns though .
What are your WAF limits? Could you get a local craftsman to build your enclosures?
Is something like this a possibility?
The coaxial style is suited to nearfield use. Can have decent WAF (try an image search on "Altec 604"), depending on the construction choices.
I went to Greatplains Audio website and could not find info on a complete
speaker package . I sent them an email request .
I did find some images of complete speakers , on Bing , which may be acceptable . But no info on the builders .
"could not find info on a complete speaker package"
They sell the driver, they sell a crossover, and they suggest a suitable box:
Recommended Enclosure: 9 cu. ft. (254.9 L), with tuning at
45.5 Hz. Recommend two 5.25-inch
(133 mm) diameter by 2.3-inch (58 mm)
long ducts or 2.5"(6.4 cm) x 11" (27.9
cm) horizontal slot for above tuning.
If you go for these, you have to decide exact dimensions and finish for yourself - even if that decision is as simple as copying someone else's build (printing some plans & photos, then giving them to a cabinet maker).
You could of course, change any of this up: buy a similar driver from another manufacturer, or use a crossover of your own devising. The GPA drivers are just one example of a type.
Maybe see if there is an audio club near you, or some other way you can contact people and hear these or something like these.
"I did find some images of complete speakers , on Bing , which may be acceptable"
Usually they are just big rectangles. You can vary the dimensions and get them finished in whatever way appeals & suits your house.
"But no info on the builders"
Since they are just big rectangles, you should be able to commission their construction pretty easily.
Personally, I use 604 clones (from Asian manufacturer P Audio) in sealed cabinets, supplemented with subs. I tinker with other speakers, and always come back to this system. I think coaxials like these are a decent set of compromises:
-simple build (not too big, no complex shapes)
-suitable for close listening position
-they sound big and effortless.
...basically the same merits as the Sansui speakers that got you interested in high efficiency.
Their main demerit, in my experience, is that on some complex / raucous music they don't have as much clarity as good headphones.
Kits and flat packs are also an option, if you're willing to do assembly / finishing. There are some crazy good deals for y'all in the USA.
Standard starting procedure for arranging speakers is to place them into an equalateral triangle with the preferred listening position (with some toe-in for adjustment). So if your listening position is 6 ft. from each of the speakers, then the speakers will be 6 ft. apart. This would imply a small room which would put many horns at a disadvantage compared to electrostatics and box direct radiators. The Altec 19 for instance is a very good speaker, but in a small room there may be too much reflection from the side walls from the mid horn. Some of the Lowther horns may fit your space, the Lowther Medallions for instance are reasonably sized for your space in that they are tall and don't have that big a footprint and they sound very good. The Lowther driver, as in the Medallion, is usually "back loaded", that is to say there is a folded horn on the back of the driver while the output from the front of the driver is direct radiating, and there is no tweeter or mid horn.
Yes, manufacturers of modern horn speaker systems exist (for a mix of vintage and modern with exquisite craftsmanship Volti comes to mind) but please DO be prepared for a sticker shock ;) If you don't mind and are capable of some DIY Wayne Pelham of Pi speakers or Bob Crites who does Klipsh are a good guys and very reasonable. Welcome to horn speakers brotherhood! My decisive moment was acquiring a single diy EV Patrician IV from 50's and since then I'm cursed , ha ha. Good luck to you.
Thanks , but I'm not able to do DIY . I will need turnkey only .
Saw the Crites CS1.5...any thoughts on those ?
Any others ?
The DIYSoundgroup horn systems are pretty much turnkey kits. If you can't DIY you can probably get someone to put them together for you for modest price.
I use a two-way horn/direct radiator system in a small room. It comprises an Altec 414A, Altec 802D, and Altec 32A horn. The 414 and the horn were intended to be used in smaller spaces such as studios and domestic environments. The 802D gives better HF's than the usual 806 driver when used with a properly implemented crossover. This is not something you'll find off the shelf unless you are willing to pony up for a Shindo Petite LaTour.
Post a Message!
This post is made possible by the generous support of people like you and our sponsors: