This to me is an excellent approach to take, in fact what I have always done instinctively, I have never ever worried about testing for distortion with instruments - if I can hear it, it's significantly there - and the more unpleasant or irritating it is, the greater the effort should be put into minimising it.
People will have different takes on what disturbs them - for me it is excess distortion in the upper midrange and treble, why most replay just sounds like "hifi" when I listen to other systems, rather than a musical event. From experience I know this can be subjectively eliminated, so that's always my first priority.
The only woofers I've heard that really keep up with good midrange horns is the Onken W with Altec 15s. No other direct radiator I've heard comes close. I didn't actaully know how inept most woofers are, until I heard that big box (with the right amp). It's one of those things that makes you shake your head and say "Speakers can't DO that!"
The first time I heard Don Keele's CBT arrays it was obvious that the woofer he was using could not keep up with the array. It seemed flabby and boomy to me. Like a speaker. No one else I asked had any problem with it.I hear a lot of speakers that sound like very good speakers, but they still sound like speakers. Maybe that's what most folks expect.
Yes, Frank - subs can have a "style". Most sound just like subs. The rare good ones sound like natural sounds, no speakers. I think the servo subs would make a lot of people very happy, based on the big grins and "Wow!" comments that go around the room when they are demo'd. They are certainly a large step in the right direction, but still not really the sound I'm looking for.
A little backstory: the two systems that I heard that made me curious about horns were the Altec A5's at the San Francisco audio club meet and Christian Rintelen's Blue Thunders in Zurich. I never liked Altec A7's, and the sound from the A5's, as crude as they were, was a revelation. Frequency response was obviously not flat, the bass cabs were flimsy and boomy, and the speaker obviously needed a real crossover, not the 12 dB/octave joke from Altec.
But ... the big multicells were sweet and powerful, an extremely rare combination in audio, and had wonderful tonality. The woofers matched just fine, obviously all of a piece, from the same design philosophy. Nothing like the abrasive harshness of the A7; what I had disliked about it was the really bad sectoral horn and possibly the small-format compression driver. The A5 was relaxed, effortless, and musical in the 500 Hz to 2 kHz frequency range where the A7 sounded like a train-station PA system.
So ... the stock A5 was nothing close to hifi, but could obviously be made to sound a lot better with a little care and attention. So I took the Sound Practices article about the system at La Maison de l'Audiophile seriously; better construction and good EQ was exactly what the A5 needed. I never heard it, but I heard the raw ingredients that went into it.
The 2004 experience, as the invited guest speaker of the 2004 ETF, and hearing Christian's system in Zurich, had a similar effect. It was a TAD wood horn and TAD compression driver, combined with a 15" bass driver, the system was superbly set up, with a fantastic EMT 930 turntable and Ortofon SPU cartridge, and Christian's great taste in music ... the ultimate personal DJ experience, regaling our little band of Americans on our first evening in Switzerland. I have still have warm memories of that early December evening, with Christian playing one great record after another ... what a great way to return to Europe after a long hiatus (the last time had been 1975, a long time ago).
My takeaway from these experiences was that large-format compression drivers, the right horn (with low diffraction and energy storage), and the correct EQ could take things to an interesting new place. Small-format drivers seemed to get in trouble somewhere below 1.5~2 kHz, and didn't take kindly to being "stretched" to lower frequencies.
I didn't realize until a couple of years ago that our moderator "Pano" was the man who physically built the renowned La Maison de l'Audiophile system for Jean Hiraga, who was at the 2004 ETF. I bow down to their experience. They've been there, done that, and own the T-shirt factory. Without Jean Hiraga, I don't think there would have been a DHT revival in the Japan and Europe, and certainly not the USA, which was ten years late to the party.
Does this mean I'm going to copy the La Maison de l'Audiophile system or what Pano has now? No, of course not. But ... I'm curious what they've found along the way. Very curious. You don't get a certain kind of sound by lucky accident; that almost never happens. It takes a lot of work, trying to find out the technical reasons for this or that sound, and seeing if it can be replicated.
Bass horns? The strongest recommendation is from my friend Martin, with the plans for the J-horn pictured on the Azurahorn site. The only way I'll be able to try them would be if I relocate the hifi system down into the unfinished basement. The basement isn't as grungy as the basements in old houses, thank goodness, but the HVAC machinery whirs and rumbles on and off since Colorado swings between too cold (-10F last week) and too hot (102F last summer), and that detracts from hifi enjoyment.
The mismatch in timing between a medium-to-large bass horn and the MF horn is a matter of concern. No, I don't want to use a digital equalizer and time corrector. It's enough of a hassle finding a single good-sounding DAC; I don't want to mess with building my own multichannel PCM1704-based DAC, just to correct the time misalignment of the bass horn. Better to find the technical reasons for the shortfall in direct-radiator sound, and see if they can be resolved.
A bass array is an obvious solution, but there can be problems with rough-edged polar patterns and asynchronous arrival times ... what I find unsatisfactory about line arrays.
Band-splitting, and tight control of excursion for the upper-bass/low-mid driver, is another approach. The perceived limitation in dynamics has a lot to do with excess excursion, which after all is what a horn controls. As it turns out, the natural cutoff of a short, Altec-sized horn is about the same as the desired crossover frequency, in the 150~200 Hz range.
Last edited:
A minor point. With the 150~200 Hz cutoff in mind, it should be possible to design a short horn (no deeper than the AH425) that takes advantage of the floor image; in essence, a Tractrix, exponential, or LeCleac'h that is a dual-driver horn, with the lower half of the horn a virtual floor image. That should retain pattern control, along with reduced driver excursion, down to the cutoff frequency.
I understand what the bass horn guys are getting at. The diaphragm in a horn-loaded system is constant-velocity, not constant-acceleration, like a direct-radiator. Translated into English, that means less excursion, more headroom, lower IM distortion, etc.
The point about IM distortion is important. Since direct-radiators are constant-acceleration devices, that means excursion increases 4X for every octave drop in frequency. (Constant-velocity means excursion only increases 2X for every octave drop in frequency.) The practical effect is that low frequencies always intermodulate higher frequencies, creating close-in sidebands. With real musical sources instead of test tones, with a far more complex program spectra, IM distortion sidebands greatly outweigh harmonic distortion.
The worst-case condition for IM distortion is loud, spectrally dense music played through a small full-range driver; bass content will strongly intermodulate higher-frequency content. Perception of distortion tends to follow the Fletcher-Munson curve, so distortion in the 1~5 kHz range is most audible (and annoying), while the peak energy in the musical spectrum is around 200~800 Hz. Bass intermodulates the higher frequencies on a program-dependent basis; not so bad with sparse spectra, not so good with symphonic and choral music. There's a reason that hifi demos at shows control the program content.
The strongest reason to band-split is reduction of IM distortion, since the MF or HF reproducers are never exposed to the energy of the 200~800 Hz region. Horn loading can also make a difference, since excursion in the 200 Hz region is significantly reduced. A bass array does the same thing, since mutual coupling raises efficiency (reducing excursion), and there are more drivers to share the load.
I understand what the bass horn guys are getting at. The diaphragm in a horn-loaded system is constant-velocity, not constant-acceleration, like a direct-radiator. Translated into English, that means less excursion, more headroom, lower IM distortion, etc.
The point about IM distortion is important. Since direct-radiators are constant-acceleration devices, that means excursion increases 4X for every octave drop in frequency. (Constant-velocity means excursion only increases 2X for every octave drop in frequency.) The practical effect is that low frequencies always intermodulate higher frequencies, creating close-in sidebands. With real musical sources instead of test tones, with a far more complex program spectra, IM distortion sidebands greatly outweigh harmonic distortion.
The worst-case condition for IM distortion is loud, spectrally dense music played through a small full-range driver; bass content will strongly intermodulate higher-frequency content. Perception of distortion tends to follow the Fletcher-Munson curve, so distortion in the 1~5 kHz range is most audible (and annoying), while the peak energy in the musical spectrum is around 200~800 Hz. Bass intermodulates the higher frequencies on a program-dependent basis; not so bad with sparse spectra, not so good with symphonic and choral music. There's a reason that hifi demos at shows control the program content.
The strongest reason to band-split is reduction of IM distortion, since the MF or HF reproducers are never exposed to the energy of the 200~800 Hz region. Horn loading can also make a difference, since excursion in the 200 Hz region is significantly reduced. A bass array does the same thing, since mutual coupling raises efficiency (reducing excursion), and there are more drivers to share the load.
Last edited:
This in fact can work work very well, I've been listening to music this way for a year or so - the main thing is to keep the deep bass frequencies away, via one means or another. Why most people may not realise it's possible, is that the amplifier is not sufficiently sorted out, or the speaker drivers, and associated electronics are not adequately conditioned before serious listening ... typically, the sound is downright awful when one starts from cold, it's in the "how can you stand listening to that racket??!" category - one needs to be patient, aware of the procedure being followed.
The harder and longer you drive the simple full-range speaker the better, after an hour or so it rises very nicely to the task at end - so, for example, one can put on a full blown choral production at close to maximum volume, and it will handle it without audible distress or overt distortion ...
The Fletcher-Munson hearing threshold effect also helps the 90x40 SEOS waveguide folks like me to put a "coherent" crossover ~1.1-1.4Khz to 10"-12" midbass drivers. Vocal range is defined as 80-1,100Hz, which can be covered by a single midbass in this design philosophy. 90x40 polar response is perfect for HT couch families. The F-M sensitivity peak is ~3500kHz.
I am also trying to understand why accepting a physically larger woofer does not also make larger horns like the JMLC-350 or JMLC-270 which still use 1.4" compression drivers more attractive. If you purchase a $700 Be compression driver, and add a $400 big woofer, the horn cost cannot be the issue.
Attachments
Yes, I can see that as a major reason for them sounding like speakers. They do sound like they are working, so you notice them.
But don't forget the box. Without nearly heroic efforts, boxes talk - and we hear them. The Onken W is a big box with a lot of surface area (outside volume of about 700 liters - 25 ft^3) but the very robust construction with sand filled double 25mm marine plywood walls counts for a lot. And at that size, the box should be fairly low pressure. Combined that means the walls don't talk.
While the Altec A5 (Mr. Hiraga's, not mine) remains my favorite speaker overall, there are ways to do better bass. Gary Pimm has found a very good technique with nested 15" woofers in an H frame. I admire Lynn for going down the road to find a design that works for him. The rewards may almost be worth it.
Yes, as soon as you're aware that something mechanical is happening to make the sound then it's game over. A force cancellation technique along the lines suggested by Lynn, combined with adequate damping, effectively rendering the enclosure totally inert while the drivers are pumping hard would be the ideal.
Hi All,
I´ve been reading through this thread with great interest I do not know how many boxes i have built over the years but i have always returned to some form of horn loading of "full" range cone drivers. As it seems my taste in music and my way of listening is much in agreement with Lynn´s and a few others here
What i have struggled with over the years is to find a bass system that can keep up with the horn dynamics and i have been living with a system for a few years now that uses 4x15" IB for bass and it is over all the best I´ve had. It cannot really compete with the few (2 or 3) fully horn loaded systems that i have heard that really worked. They were all based on the article "Acoustic Compensation" by Rex Baldock from the mid-sixties. The bass from these systems are truly weightless and they can, fully time aligned, give full justice to a 32´ organ pipe. I have not found this article on the web but can scan and post it tonight.
I have always found CD based systems lacking energy and dynamics in the 300-500Hz region except on a few rare occasions and have stuck to my horn loaded cones, a force of habit and comfort as much as anything i guess
I have recently relocated and had to give up the old IB/horn system and is therefore setting up a little winter project looking something like below. The drivers used for the mid/high will be the same as in the old system, Feastrex D9e and for bass i have four GPA 416-16B. The aim is to get an as seamless integration as possible in the crossover region and radiation pattern that is smooth and well integrated with the room below say 5k.
I will have a go without sub to begin with and take it from there.
BR,
Anders
I´ve been reading through this thread with great interest
I do not know how many boxes i have built over the years but i have always returned to some form of horn loading of "full" range cone drivers. As it seems my taste in music and my way of listening is much in agreement with Lynn´s and a few others here What i have struggled with over the years is to find a bass system that can keep up with the horn dynamics and i have been living with a system for a few years now that uses 4x15" IB for bass and it is over all the best I´ve had. It cannot really compete with the few (2 or 3) fully horn loaded systems that i have heard that really worked. They were all based on the article "Acoustic Compensation" by Rex Baldock from the mid-sixties. The bass from these systems are truly weightless and they can, fully time aligned, give full justice to a 32´ organ pipe. I have not found this article on the web but can scan and post it tonight.
I have always found CD based systems lacking energy and dynamics in the 300-500Hz region except on a few rare occasions and have stuck to my horn loaded cones, a force of habit and comfort as much as anything i guess
I have recently relocated and had to give up the old IB/horn system and is therefore setting up a little winter project looking something like below. The drivers used for the mid/high will be the same as in the old system, Feastrex D9e and for bass i have four GPA 416-16B. The aim is to get an as seamless integration as possible in the crossover region and radiation pattern that is smooth and well integrated with the room below say 5k.
I will have a go without sub to begin with and take it from there.
BR,
Anders
Attachments
Hmmm....I have been toying in my head with an idea involving passive radiators. I have also only heard more or less bad implementations of the concept but it does have some appealing assets.
What if we use an ordinary bass driver with large excursion capability and low Fs as a passive radiator? Then we could apply tailored damping by loading the VC.
//Anders
I just had another possible idea, and I thought I'd share it. It is based on the fact that Lynn is toying with the idea of a front horn for the mid-bass. How about making a hexagonal front horn, and the turning it into a type of Unity-horn. Start with a side length (for the sides of the hexagon where it meets the baffle of the mid-bass) of 26cm per side, the inner radius of the hexagon will be 22.5cm, giving an inner diameter of 45cm- just enough for a 15" woofer. Increase the length of the sides so that at the mouth the length of each side is say 46cm, then the inner radius will be 39.8cm, giving a inner diameter of circa 80cm. True, if you add a round-over to the outer edge of the horn-mouth you'd end up with a box of about 40" wide and about 44" high, but that is the one of the compromises. Now add a 12" woofer on each of the sides of this horn mouth. That will be six 12" drivers, enough to ensure good low bass. Not only that, but there will be a partial force cancellation between the opposing bass drivers. The sound will be also co-incident. My only problem with this idea is horn-mouth reflections from the bass drivers could be a problem somewhere in the mid-bass region, so I am not sure if this will work. Adding a round-over to the horn-mouth will also reduce reflections. I added an pic to better illustrate my idea, but again I apologize for the crudeness of my drawing. I only have MS Paint to do my drawings in.
Regards,
Deon
PS. I got so caught up in the idea of a hexagon, I never gave thought to just a plain, square horn-mouth. Same idea, just fewer sides. That could also work, but from my own personal perspective, still need some help in the low bass. I really think 3 pairs of 12" drivers should be the minimum, but that's just me.
Regards,
Deon
PS. I got so caught up in the idea of a hexagon, I never gave thought to just a plain, square horn-mouth. Same idea, just fewer sides. That could also work, but from my own personal perspective, still need some help in the low bass. I really think 3 pairs of 12" drivers should be the minimum, but that's just me.
Attachments
Last edited:
At this point I would also like to relate the experience of a friend of mine with regards to front horns on large drivers (I am talking about large cone drivers.) My friend, Chris Templer (you can look him up on TNT-Audio where he is one of the new reviewers - Chris on TNT-Audio), had a pair of the big Tannoy Autograph cabinets built for his Tannoy dual-concentric drivers. But instead of the standard conical front-horn, he decided to make it an exponential-type front horn. The exact expansion profile was 'toilet-seat cover' (his words). Basically he took a toilet-seat cover, traced to curve onto paper, and that became the curve for the front horns. The result was that they never sounded right. He said one of the things that was wrong was that they tended to throw the images to the ceiling. He later on had the Autographs rebuilt with the factory-spec conical front horn, and he said they sounded excellent that way. Since then he has upgraded to a double Autograph with a sub consisting of four 18" woofers in 6 meter tube (see a pattern here ). Nevertheless, his experience was that a conical front horn is best with a large driver and that is what he still uses.
Deon
). Nevertheless, his experience was that a conical front horn is best with a large driver and that is what he still uses.Deon
Generally damping in a passive radiator is not desirable, so this idea would be an expensive addition to solve a non-existent problem.
I have used passive radiators in a bandpass configuration and they work well, but I don't see much advantage in them as "ports". In a bandpass system they are very good at reducing noise from "air" ports, which can be a big problem in a bandpass design.
Still, nothing beats several small closed boxes for simplicity and, when properly setup, extremely good bass performance. No single bass location system is going to compete with a distributed one. And when you do that, size is not a big factor anymore.
I thought of it as the PR equivalent of adding flow resistance to port, sort of a Q-trimmer.
//A
Nothing beats a bass corner horn. GF knew what he was doing! Nothing new, and cheap to build.
It may change the "Q" of the tuning, but it does this by lowering the efficiency, which is the last thing that you want to compromise in a sub. All resistance in a "port" are loses in sound radiation. The system "Q" should be controlled by the driver where lower Qs do not come from efficiency lose.
The thing I liked about the bass horn was having the in phase coherent in a perfect world) radiation propagating out across the end wall and flank walls. I think phase in the bass is as important as the rest of the delivery must be in phase with the mid mid/treble horn. I still appreciate your subs where you can tune and phase by placement. And a corner is an otherwise wasted area in many rooms. I prefer the horn effect with organ stops in particular giving a synergy of being delivered down a pipe.