I owned 4430 for about a decade. They were given to me by John Eargle for the work that we did in bringing about the (temporary) union between Ford and JBL. He hand picked my pair and had them measured. At the time they were good speakers, but I came to like them not so much over time.
It was as a replacement for the problems with the 4430 that I designed the Summa and its ancestors. To me, the Summa was dramatically better than the 4430's. Today, I could not tolerate the harsh sound of the 4430.
And just to be clear again, one can EQ a bad horn at a single point, but diffraction and resonances cannot be fixed globally. No acoustic problem can be "fixed" with EQ, DSP or otherwise. It's a three dimensional problem and one dimensional solutions are incomplete.
I can imagine that the 4430 is the best speaker some people have ever heard or owned and they would likely say they’re the best ever. This would be based on their experience.
As a note I gave my 4430’s to my daughter, who loves them. I still have my 4435’s in my office and I love them.
Dr. Geddes has said in some of his posts that he (some times or often?) listens in the 105dB SPL range if I remember right and the 4430/4435’s sound harsh to me approaching that level as well. There has also been discussion that some types of distortion increase in audibility as overall level increases.
The small format driver and 2344 horn combination just doesn't get loud nicely and is something I have just come to expect. Something I did not expect came like this. I bought a Bryston DAC recently and decided to put it in the signal chain of the office system for first test since this is the system I listen to the most.
There are several subtile differences in the sound and one really big one, the 4435’s play considerably louder without harshness than ever before. I hope I can somehow measure what I am hearing, objectively.
I don’t want to derail this into 2344 thread. Do I still like mine? Yup. Do my 4365’s outperform my old 4430’s? In every way? Yup.
I hope this thread turns into the making and measuring of actual horns!
Barry.
As a note I gave my 4430’s to my daughter, who loves them. I still have my 4435’s in my office and I love them.
Dr. Geddes has said in some of his posts that he (some times or often?) listens in the 105dB SPL range if I remember right and the 4430/4435’s sound harsh to me approaching that level as well. There has also been discussion that some types of distortion increase in audibility as overall level increases.
The small format driver and 2344 horn combination just doesn't get loud nicely and is something I have just come to expect. Something I did not expect came like this. I bought a Bryston DAC recently and decided to put it in the signal chain of the office system for first test since this is the system I listen to the most.
There are several subtile differences in the sound and one really big one, the 4435’s play considerably louder without harshness than ever before. I hope I can somehow measure what I am hearing, objectively.
I don’t want to derail this into 2344 thread. Do I still like mine? Yup. Do my 4365’s outperform my old 4430’s? In every way? Yup.
I hope this thread turns into the making and measuring of actual horns!
Barry.
Just to be clear, 105 dB would not be the norm, but it certainly is possible - at times! - on peaks with "C" weighting. My compression driver is no bigger or better than CD on the 4430's, its pretty much the same. But on my system there is never any harshness. I, of course, wondered why this is. This is why B&C and I did the studies that we did to find out the reasons, which are now perfectly clear to me. I won't belabor this thread with a repeat of all that work other than to say that it is all available - as close as the link below.
For those looking for the Earl and B&C test papers you can find them
here Compression Driver Subjective Test (if Earl doesn't mind I posted the direct link)
here Compression Driver Subjective Test (if Earl doesn't mind I posted the direct link)
Posting that is fine, but there is another paper that follows that one. The first one (above) shows that nonlinear distortion in the compression driver is not likely to be what we hear at high SPL. The second paper http://www.gedlee.com/Papers/AES06Gedlee_ll.pdf describes what is likely to be the cause. In my experience the conclusions drawn from these two papers turns out to be exactly correct. We hear the horn, not the driver and minimizing the diffraction by using a well designed waveguide will eliminate the problem.
Sry for not answering yet.
The magnet is always the same in strength, ofcourse. But the EBP (Fs/Qes) says nothing at all about the reproduction above the bass, the Q factor describes the bandwidth/height of the resonance. So the speaker behaves ofcourse a lot different when the system is in resonance (fs) than at other frequencies. Above, the Q does not have much (if at all) influence, that means calculating with the Q are not applicable there. Don't believe me? NP, You'll find a lot of excellent midrange drivers with a Qts above 1.
The suspension has to control the movement. If it's too weak the cone starts to tumble or sag which causes the coil to scrape on the pole plate or -core, don't forget, you'll have to add a LOT of weight, for 30Hz it's 96g, that's more than the whole Mms originally is! And by worsening the Mass/Motor ratio, you'll lose spl - a lot! The Celestion got a (from the parameters calculated) spl of 96,4dB. If you add weight until you reach, let's say just 30Hz, the EBP drops vastly, from 81 to 54 and you'll reduce the spl to meager 91,1dB. Well, it's pretty useless for a 15" then, isn't it?
The magnet is always the same in strength, ofcourse. But the EBP (Fs/Qes) says nothing at all about the reproduction above the bass, the Q factor describes the bandwidth/height of the resonance. So the speaker behaves ofcourse a lot different when the system is in resonance (fs) than at other frequencies. Above, the Q does not have much (if at all) influence, that means calculating with the Q are not applicable there. Don't believe me? NP, You'll find a lot of excellent midrange drivers with a Qts above 1.
The suspension has to control the movement. If it's too weak the cone starts to tumble or sag which causes the coil to scrape on the pole plate or -core, don't forget, you'll have to add a LOT of weight, for 30Hz it's 96g, that's more than the whole Mms originally is! And by worsening the Mass/Motor ratio, you'll lose spl - a lot! The Celestion got a (from the parameters calculated) spl of 96,4dB. If you add weight until you reach, let's say just 30Hz, the EBP drops vastly, from 81 to 54 and you'll reduce the spl to meager 91,1dB. Well, it's pretty useless for a 15" then, isn't it?
I don't think that answers my point. The "Q" of a resonance is only defined at resonance (for any resonance), it's not a function of frequency. But the same BL that yields the resonance Qt is the same BL that gives a driver its pass-band efficiency. So, in essence, it means that they are the same thing, just in different domains.
No, that's wrong. The Q defines the hight and the width ('bandwidth') of the resonance, quite well observable at the fs impedance peak (symmetrical on an ideal speaker) but also on the frequency response or, to be more exact, it's how high the spl at the fs is compared to the average level. That means, it actually is a function of frequency. You can easily try it out with any sealed enclosure simulator. At Qt=1 the spl will always be at exact 0dB at the Fs in that enclosure, no matter what driver you use. For other Qt's you can calculate the level the same way.
The upper bandwidth limit/roll-off of a speaker is not mainly determined by the Qt unless it's in a bandpass enclosure (double/single ventilated bandpass, compression chamber with or without a horn (or the other way around) etc. or mechanical filtering like downfire).
The upper bandwidth limit/roll-off of a speaker is not mainly determined by the Qt unless it's in a bandpass enclosure (double/single ventilated bandpass, compression chamber with or without a horn (or the other way around) etc. or mechanical filtering like downfire).
All this I've been finding to be oh so true.
I've been do a lot of measuring....and as anyone will quickly see ....when magnitude and phase are smooth and flat, impulse approaches ideal, and vice-versa....
I mean, after all.....they are the same measurement...
As far as directivity ....
I use boatloads of taps to imbed minimum phase EQs and linear phase xovers into FIR files, using rePhase and firDesigner, going in a bank of miniDSP openDRCs.
It's easy to make near perfect magnitude and phase response at any mic location chosen.
BUT only for that one location ! Move the mic and beautiful traces start wiggling..
What's clear is that the more constant the directivity, the better the mag and phase traces hold their smooth flat beauty as the mic is moved around.
It's become so easy to make beautiful traces at one location, they seem superfluous.....
......a set of on and off axis traces that hold together is what counts IMO.
And the most important thing, the only thing that matters really, is how great the speaker sounds with a nice set of traces.
PS...and FWIW.....I measure and tune outdoors, to then bring speakers indoors and see what placement works and what acoustic solutions are needed.
The last picture in post 4 looks like a flat-walled conical horn (like a Synergy) that has 4 spherical caps stuck onto it.
...so, what if you did exactly that?
Start with a flat-walled conical horn, but instead of directly cloning the Synergy (mids that fire through a port), you mounted some appropriately shaped midwoofers into the walls.
Would the HF off the tweeter "see" the surface of the woofers as solid?
Pic is of a Tymphany GBS woofer.
...so, what if you did exactly that?
Start with a flat-walled conical horn, but instead of directly cloning the Synergy (mids that fire through a port), you mounted some appropriately shaped midwoofers into the walls.
Would the HF off the tweeter "see" the surface of the woofers as solid?
Pic is of a Tymphany GBS woofer.
Attachments
This is also precisely what I do as well. The speakers should not be modified for the room except in the LF modal region.
There's nothing wrong in modifying the speakers for the listening room, if they are used only in this room. I mean, it's not important how a high fidelity reproduction is achieved, even if it's much more preferable to build the speakers universally usable. But to modify the speakers is as legitimate as modifying the room and furniture to achieve the desired sound. Nowadays it's much easier though to get the speakers right and then adjust/EQing them right to the room. That got very strict borders when it got to reflections, dispersion characteristics is something no DSP can correct. The question now is, what modifications do actually make sense?
I completely disagree. What one wants is first and foremost a neutral direct field, first arrival, response. That means that there is no "room" in this most critical aspect of sound perception. So correcting a speaker "for the room" will inevitably ruin this direct response if the speaker has been EQ'd for neutral in a free field.
No, that's absolutely not what everyone wants. Most, yes, but not everyone. There are a lot of examples where you want an even sound in most of the room and not only somewhat nearly-perfect in just one spot. You cannot assume your standards are the measurement for EVERYONE.
The free field is in 99% of the cases irrelevant because 99% of the listerners aren't using the speakers there (free field). See, while I agree that many hifi enthusiasts will benefit from 'correct'/'right' speakers, you cannot lay that strict standard upon anyone. Once you're away from the sweet spot, so-called 'ideal' speakers do not perform that well anymore. That means, optimizing your speakers depends strongly on what you're trying to achieve and that can include room dependent issues. If you ignore that, you'll have probably a virtually perfect speaker but still a terrible sound in the room. Now please tell me how that's better just because it matches your requirements and impression on how a speaker should look and sound like.
Hi ICG,
I have to say, IMO free field is the beginning of the name of the game.
It's the way we separate variables....room from speaker.....
this is crucially important...there's an old adage....you can't make chicken salad out of chicken ****
Gotta start with the best chicken salad (free field speaker) you can get !!!....
And it's also crucially important for maximizing our odds at having great sound in any room, at any "sweet spot".
Put a refined free field speaker in a room, and you can know then, that the only ways to make better sound in that room are with speaker placement, acoustic treatment, and low frequency management...ie squelching room modes via eq.
I have to say, IMO free field is the beginning of the name of the game.
It's the way we separate variables....room from speaker.....
this is crucially important...there's an old adage....you can't make chicken salad out of chicken ****
Gotta start with the best chicken salad (free field speaker) you can get !!!....
And it's also crucially important for maximizing our odds at having great sound in any room, at any "sweet spot".
Put a refined free field speaker in a room, and you can know then, that the only ways to make better sound in that room are with speaker placement, acoustic treatment, and low frequency management...ie squelching room modes via eq.
Last edited:
Well, even if you didn't expect that - I totally agree! BUT! There are circumstances where a perfect free field speaker does not perform well because of reflections, resonances, imperfect positioning or flutter echo. How comes? Free field does not include any reflections, that means, wide or uncontrolled dispersion of the sound isn't 'punished' in any way. In real world situations on the other hand, it looks a lot different. You'll not encounter any real world listening room without any reflections or uneven absorpitons. Heck, even if everything's perfect in the listening room, the human body of the listener doesn't absorb/reflect evenly throuought the whole frequency range. Okay, I know that's completely theoretical but in reality - or real rooms - you'll get a lot of reflections. To understand reflections aren't bad per definition is very important since they attribute a lot to the room impression of the reproduction. The problem is, the room reflects or absorbs differently, depending on the frequency. While the location and impression of the stage mainly dependend of the first (direct) impulse, the tonal reception of the sound actually is dependent on the direct and reflected/delayed sound.
That shows dispersion control is much more important in some rooms than in others and also shows why a perfect free field speaker performs that bad in some rooms. That's why some speakers sound as great in one location as another pair of speakers do but they differ a lot in a different location/room. The more even it disperses the energy into the room, the less the speaker changes the characteriscis from room to room. That's dependent on the absorpiton, the reflection, the room dimensions and also the speaker positioning and the listening locations, that all contributes to the reflections and absorptions.
Or in short: Perfect free field speakers only perform perfect in free field. Changed room/circumstances make the speaker sound much worse and adaption/modification can improve the reproduction a lot if it's optimized for that location. And don't forget, optimizing for one location may result in worse sound in other areas - it's all a compromise.
That shows dispersion control is much more important in some rooms than in others and also shows why a perfect free field speaker performs that bad in some rooms. That's why some speakers sound as great in one location as another pair of speakers do but they differ a lot in a different location/room. The more even it disperses the energy into the room, the less the speaker changes the characteriscis from room to room. That's dependent on the absorpiton, the reflection, the room dimensions and also the speaker positioning and the listening locations, that all contributes to the reflections and absorptions.
Or in short: Perfect free field speakers only perform perfect in free field. Changed room/circumstances make the speaker sound much worse and adaption/modification can improve the reproduction a lot if it's optimized for that location. And don't forget, optimizing for one location may result in worse sound in other areas - it's all a compromise.