Beyond the Ariel

Finally a decent commercial system (it's alive!) that I'd like to hear setup properly.

Mono and Stereo High-End Audio Magazine: Viva Audio Master Horn System review

Definitely a hard-core system with a price-tag to match. Unfortunate timing, though. Discretionary luxury purchases tend to evaporate when the financial markets get unsettled, so I hope the manufacturer (and the webzine) are able to ride this out.

The reviewer obviously liked the speaker (and the associated set of amplifiers), but the translation into English made it kind of hard to tell what it actually sounded like, or any of the technical parameters of the system, except the 250,000 Euro price ... well into Top Gear supercar territory. If I had that kind of money, I think I'd get the car, or maybe a Light Sport Airplane, which would be even more fun.
 
Last edited:
Member
Joined 2008
Paid Member
I really wonder about this system too.
90hz for that horn seems low to me. I would have thought it would be more direct radiation and less horn loading around 90hz with that horn...
Still, I would love to hear the system. I had a similar system with a different bass solution and it sounded good.

The big throat and shorter horns work and offer better aesthetics and easier time alignment than a four foot deep horn with smaller throat. I have lived with that compromise a few times myself. In a system of this price I think it might be more "sellable" because of the look of the short horn. My current system I'm building has a depth of around 4' and it looks GREAT from the front LOL

Definitely a hard-core system with a price-tag to match. Unfortunate timing, though. Discretionary luxury purchases tend to evaporate when the financial markets get unsettled, so I hope the manufacturer (and the webzine) are able to ride this out.

The reviewer obviously liked the speaker (and the associated set of amplifiers), but the translation into English made it kind of hard to tell what it actually sounded like, or any of the technical parameters of the system, except the 250,000 Euro price ... well into Top Gear supercar territory. If I had that kind of money, I think I'd get the car, or maybe a Light Sport Airplane, which would be even more fun.

The one percentors should be all over this - it probably does the live sound as well as any other commercial system - What is that in dollars? $400 thousand? :) I think I'll take a vacation home in the Caribbean

Perhaps a bit complex, but, over all a very nice execution. Of special note, is, this system [finally] addresses the very low bass, which is the foundation for which almost all music is built upon.
I "like" it :)

The previous version did not have the upper treble horn. I think that was smart over just the super tweeter. Over the years I find a one inch driver upper treble mates best with a large format below. The four 18's are smart loaded in the corner - they borrowed that from me.. lol :eek:They still need a central mono folded sub and rear channel ambience sub though

I also question the arrangement of the array with the HF's cascading vertically to the top. In my experience it sounds better to have the large format midrange horn above with the treble horn centered between the midbass and midrange. This may be aesthetics too. There is a better way if you are building for yorself and don't have to satisfy the eye of the rich buyer.
 
Last edited:
A very minor update...

Hello,

just by way of an update...

I tweaked the crossover ever so slightly on my main system (described below), bringing the first crossover point down to 800Hz and 'massaging' the curves a bit. This produces an even better integration of the woofer and the compression driver, with almost perfect phase tracking in the overlap region.

Incidentally, I found that the small (-3dB) and low-Q 'dip' centred at around 1.7kHz which is visible in the 1m on-axis measurement of the Mid response (see attachment) *magically* disappears when the system is measured from the listening position (3m distance), and also when the Mid is measured slightly off-axis. The most probable explanation is that it is produced by some destructive interference with the secondary sound emission by the horn 'lips' - a minor anomaly that is essentially inconsequential when actually listening to the system from a typical distance.

Finally, I'd like to share my thoughts on the topic of horn profiles and the improvements & tweaks for the classic Altec radial/sectoral horns (i.e., 311, 511, 811) which were discussed a few posts back.

Two related observations got me thinking:

1) the fact that those tweaks (damping the horn bell, cutting the reinforcement 'fins', etc.) seem to produce positive results;

and

2) the fact that my own horns (which are of a VERY similar design, they too being 'classic' radial horns like the Altecs, but carved out of thick blocks of high-density wood, instead of assembled from thin metal sheets) do not seem to exhibit any of the notorious 'honk' that is typical of the 'stock' untreated Altec horns.

Might it be, then, that in the end, all of the theoretical shortcomings of the classic radial horns designed according to Webster's equation and arbitrarily assuming cylindrical wave fronts... do not PER SE matter that much, and instead the reason for those horns' bad rep lies elsewhere (i.e., in their undamped 'ringing', which is obviated by either several heavy-handed tweaks, or by a radical approach such as carving the horn out of wood)?

Having said this, I do understand the theoretical advantages provided by e.g., the JMLC family of horn profiles. But the fact remains that, whenever I had the chance to listen to any systems using those horns (at the Munich High End show, as well as at other audio shows in Europe), I did not notice any further improvement in terms of 'lack of horn colouration' compared to my own system. (Of course, different systems, with different drivers, different woofers, different crossovers, etc. etc... and of course I may just be biased.... but still...)

Marco


Hi all,

one thing that I noticed in this wonderfully informative and thought-provoking thread is a rather puzzling dearth of actual complete build examples, supported by actual measurements.

In an attempt to buck that trend, I hereby offer my own design for scrutiny, which, even if not fully "beyond the Ariel", still by and large shares the same fundamental principles.

1) DRIVER SELECTION

WOOFER: Fostex FW405N (FW405N | FOSTEX)

This is a modern woofer of recent design, with a hybrid cellulose+carbon fibre cone and a newly developed aluminum die-cast frame based on FEA. Magnet is ferrite, and sensitivity is medium-high at 92.5 dB/W(m) (declared) / 94 dB/W(m) (calculated based on the T-S parameters).

Two features that attracted me were:

(i) its medium-low overall damping (Fs/Qts = 60), which allows to use an overdamped bass-reflex box characterized by very low Group Delay (Vb = 2 * VAS * Qts) and still get an F-3 = 40 Hz;

(ii) its very low mechanical damping (Fs/Qms = 4.0), which in my experience is a recipe for excellent low-level detail retrieval.
By way of comparison, this is roughly the same mechanical damping as that of the GPA 416 (Fs/Qms = 3.4) and the TAD TL-1601a (Fs/Qms = 4.1), and HALF of that of the JBL 2226H (Fs/Qms = 8.0).

COMPRESSION DRIVER: Fostex D1400 (D1400 | FOSTEX)

This is an 'old style' 1" driver with a Ti-alloy diaphragm, half-roll surround and massive AlNiCo magnet. Its internal structure is basically a copy of the TAD TD-2001, with a large back-chamber and an internal conical throat characterized by a low cut-off frequency of 400Hz. These characteristics allow it to perform very well down to much lower frequencies than most modern 1" drivers (at domestic levels, of course, and when coupled with a suitable horn). In fact, its minimum recommended crossover is a low 750Hz (presumably when used with Fostex's own larger H300 horn).

HORN: Fostex H400 (H300/H400 | FOSTEX)

This is where I decided to go 'all Fostex' and try my hand at a 'classic' radial exponential horn (Fc = 455Hz)... but with a few caveats! In fact, this horn differs from e.g. the 'über-classic" Altec 811 in two important ways:

(i) it is carved out of a block of dense high-grade sugar maple plywood (density = 0.67 kg/dm3). This gives it excellent self-damping and prevents the notorious 'ringing' of the classic metal horns.

(ii) its (cast aluminium) throat adapter is perfectly matched to the driver's internal 400Hz flare rate, and operates an extremely smooth transition from the driver's 1" (25.4mm) circular throat to the horn's 35x35mm square throat.

Of course, all this does not change the fact that this is still a 'flawed' design in the other classical ways (calculated based on Webster's equation assuming cylindrical wavefronts, producing some diffraction at the adapter/horn throat interface and at the horn mouth, etc.). But my hope in selecting it was that the two positive points listed above would go a long way in minimizing any unwanted 'horn honk'...

Super-TWEETER: Fostex T925A (T925A | FOSTEX)

A classic 'bullet' tweeter, with an aluminium ring diaphragm and large AlNiCo magnet.


2) PASSIVE CROSSOVER:

For the Woofer-Mid crossover, I got inspiration from the design used by Pioneer in their TAD/Exclusive monitors (EXCLUSIVE model2402), and later reprised by Shozo Kinoshita in his Rey Audio monitors (RM Monitor), and I designed an asymmetrical 6th-order low pass / 2nd order high pass.

The crossover frequency was chosen to be 850 Hz, so as to:

(i) match the horizontal directivity of the Woofer to that of the radial horn (110 degrees);

(ii) stay approx. one octave above the horn's cut-off.

When both the Woofer and Mid are connected with positive polarity and the Mid is phisically set back so as to create a suitable offset between the two acoustic centres, this crossover results in the emissions of both drivers to be in phase over a relatively wide frequency range around the crossover frequency [see actual measurements in FIGURE 1].

This type of crossover is handy in several ways:

(i) the front-to-back offset allows the convenient positioning of the horn-loaded mid atop the Woofer box without requiring any form of delay (digital or otherwise);

(ii) the same offset also happily results in the two impulse responses to be almost perfectly 'time aligned', i.e. both traces leave the horizontal line simultaneously [see IR measurements in FIGURE 2];

(iii) the 6th-order low pass effectively does away with any unwanted resonances at the top of the woofer's operating range.

For the second crossover between Mid and super-Tweeter, I opted for a symmetrical 2nd-order Buttwerworth at ~7.5 kHz, which gives a Constant Power response.

The super-Tweeter is not time-aligned because:

(i) at such high frequencies, the human auditory system's sensitivity to phase is reduced;

(ii) the resulting comb filtering pattern is essentially inaudible since the narrow notches are within the ERB (https://en.wikipedia.org/wiki/Equivalent_rectangular_bandwidth);

(iii) having an intentionally misaligned constant power crossover at such high frequencies results in a smoothed summed response that is essentially invariant over a +/- 30 degree listening angle, since any additional misalignment introduced by moving the head laterally is swamped by the the >360 degree original misalignment.

I realize this design choice may be somewhat controversial, but IMO while it looks bad in a simulation, it actually sounds better than any other option (i.e. better than just going 2 way and making do with the compression driver's response over the top octave, and better than trying to obtain a time-aligned all-pass crossover between the mid and the supertweeter, and then living with the 'hole' in the summed response that is produced off axis).

From a practical point of view, the complete crossover was built using quality parts (Mundorf coils and Jantzen MKP capacitors), and housed in two external wooden boxes [FIGURE 3].

The woofer low-pass section [FIGURE 4] uses a semi-balanced topology to reduce crosstalk with the mid and tweeter sections [FIGURE 5].

Both the compression driver and the super-tweeter are attenuated using variable units by Fostex (R80B/R82B/R100T | FOSTEX), respectively the transformer-based R-100T for the CD and the R-80B potentiometer for the sTw.

The resulting system impedance plot is shown in [FIGURE 6].


The final system's measured in-room response (unequalized and not corrected in any other way) is shown in [FIGURE 7], with superimposed the well-known Bruel & Kjaer 'target' function, as well as a +/- 3 dB tolerance band.

And to conclude, a picture of the full system taken from the listening spot [FIGURE 8].

C&C welcome!

Cheers,
Marco
 

Attachments

  • Wf+Mid+Tw filtered (gated).png
    Wf+Mid+Tw filtered (gated).png
    31 KB · Views: 533
  • 01 Loudspeaker.jpg
    01 Loudspeaker.jpg
    179.4 KB · Views: 539
Last edited:
I've been wondering why a bass horn, or even a not-so-great A7-style horn, sounds so different than an array of direct-radiator drivers. On paper, both are similar articles; lower distortion and greater headroom are the expected tradeoff for larger size, cost, and complexity. If anything, the horn should be worse, thanks to mouth reflections and somewhat uneven diaphragm loading vs frequency.

But in practice, they sound different. A lot different. The horn shares traits with electrostats; quick and snappy, although with more weight and a sense of in-the-room physical presence. Direct-radiator arrays usually just sound big, and can be noticeably more blurry than single drivers (particularly if the rear volume is shared between drivers). They should sound like horns, but don't.

Just spitballing here, but I wonder if the difference is the cabinet. There's a lot of stored energy in the cabinetry for a bass array, while a bass horn typically only has a very small rear enclosure. The ear is extremely sensitive to the characteristic sound of vibrating woods and metals, and relatively small amounts of cabinet coloration are noticeable when the cabinet is removed from the picture. The brute-force solution to cabinet coloration is an open baffle, but that throws away efficiency and power-handling, and demands long excursions from the driver ... pretty much the opposite of what a horn does.

I've heard the same thing, and I say that from experience with long MF/HF lines (not arrays; true lines).

Long-gone Kinergetics Research once made a black box to address what they found was the inherent wake problem of lines or arrays operating as lines. IIRC according to them, the line, being a 2D radiator, cannot reproduce a proper step function.

Apparently the roughly cylindrical waveform doesn't form the required geometry of a natural source correctly and the result may be that characteristic audible blurring of softness you get from lines and especially planes. While the gigantic sensaround effect is not unpleasant, ultimately it's not guaranteed to sound natural even with the purported qualities of big long exotic sources.

Whether this affects a stack of bass drivers I can't say. You'd think low frequencies theoretically shouldn't have this problem when they're already radiating more or less spherically from an array many times shorter than line function calls for, but maybe the true spherical form can't develop unless each driver from center is stepped back along a curve, even for say, 40Hz.

Searching hasn't turned up the Kinergetics white paper but I'm curious to know what it delves into...
 
Marco,
I once long ago pulled a mold off an Altec 311 horn sans the fins. I copied that horn shape exactly but changed the material the horn was made of to a glass balloon filled polyester and we compared the two horns in the exact same configuration. The difference in sound was instantly obvious, and don't forget that the 311 is not sheet metal or the thin die casting like an 811 or 511, the 311 was a very heavy cast aluminum horn with damping material on the outside. So yes the material that the horn is made from makes a major change in the final sound.

POOH,
I think you idea of the high horn being in the middle is the way to go, that is what I would do with a large multi=horn array, it just makes sense. As you say I think many horn systems are just made to look a certain way visually but the thought pattern of what you are really trying to do is lost in convention. I have had a design in my head for many years of a concentric three way horn system, a tri-axial horn design. Perhaps one day I will build it just to prove my point. I keep looking at some dual 18" throat horn sections I have sitting that I originally did for PA use. Four bass horns stacked per side gave a true 50hz 1/4 wavelength mouth size, something that I haven't heard in any other bass enclosures. They would go as low as the fs of the 18" drivers but of course you lose directivity control below the 50hz frequency but who cares really. They made the best PA bass I ever have heard. Just don't stand in front of a stack like that, definitely capable of destroying your hearing instantaneously.
 
Hello,


Might it be, then, that in the end, all of the theoretical shortcomings of the classic radial horns designed according to Webster's equation and arbitrarily assuming cylindrical wave fronts... do not PER SE matter that much, and instead the reason for those horns' bad rep lies elsewhere (i.e., in their undamped 'ringing', which is obviated by either several heavy-handed tweaks, or by a radical approach such as carving the horn out of wood)?

Marco

I'm hooked on "good" radial horns myself. They just sound so right and alive to me. I never liked the 811 511 altecs much (still have some Flamenco's buried around here) but grew up with a damped pair of 311's and are my long term "reference" from the 70's lol - I have heard many horns since but they are the one's that got me hooked. The big round horns need to be be spaced too far away from lower and upper ranges and show up as incoherent lobes in my room. Not as natural and cohesive as the radials.
 
Last edited:
diyAudio Moderator
Joined 2008
Paid Member
The big round horns need to be be spaced too far away from lower and upper ranges and show up as incoherent lobes in my room. Not as natural and cohesive as the radials.
I'd like to take this at face value.. I want to build a radial but for the wrong reasons and I'm looking for an excuse. I believe they sound good but can a radial style tweeter throat be as good as a round horn? I'm with you on the spacing but what about the lobes, would a ceiling absorber take care of it?
 
You make the initial throat round and morph it to conic, then radial. If you use Altec, GPA, the round conic is built in to cover much of the HF BW, then the transition to conic isn't too critical due the size of the WLs involved as it transitions to radial.

One interesting thing is that if this transition is done right, the rectangular pattern will 'hold' for very long distances, so while 'sound is round', as long as there's good pattern overlap, a radial system is just as good as a round one plus you don't have to be nearly as far away for it to sound as one large point source.

Some folks obsess over this pattern 'distortion', though IME, it's not the rectangular pattern, but the transition section. That said, the pioneers concluded in anechoic listening that ideally one should limit it to a 1.00:1.273 ratio, though rarely practiced. Back in my youth I couldn't tell any difference in comparison when BW limited until run 'full-range', so concluded it had to be due to an imperfect transition and judging by Dr. Geddes's much more recent 'round to round' WG transition observations about how critical they are, I consider it 'carved in stone' now.

GM
 
I'd like to take this at face value.. I want to build a radial but for the wrong reasons and I'm looking for an excuse. I believe they sound good but can a radial style tweeter throat be as good as a round horn? I'm with you on the spacing but what about the lobes, would a ceiling absorber take care of it?

It's the interference between the drivers, the round horns are spaced further apart - if you use a steeper crossover it helps because there is less overlap between the drivers but steep crossovers have their own set of problems. I never built a radial horn because there are good ones out there already done right (and a lot done wrong like JBL with the 2328 throat or them big nasty 2360/65 biradials. If you have a 2" driver and want to try one on the cheap a good example is the JBL2386 - disregard the narrow radiation and price and just measure it and listen. Used as a mid horn 600 - 6k they are real sleepers. Another is the Community BRH90 "big radial horn" - If you like what you hear build a nice pair and enjoy.

http://clairusedgear.com/products/jbl-2386-horn
 
Last edited:
"The big round horns need to be be spaced too far away from lower and upper ranges and show up as incoherent lobes in my room"

Couldn't agree more, I don't understand the thinking where this is concerned. .

Round horns can sound pretty good but in my experience in multi way systems they have some problems - a 2 way probably will work best. I have some big wooden 180 flare round tractrix horn and if you load them with TAD 400* drivers and sit in one exact spot with a 2 way they are really good - I don't like the beaming though.

What it is is the "high end" audio crowd went back into horns with the Avantgarde round horns and the hyped spherical wave front and a lot of people bought it hook line and sinker - me included LOL

Horn Loudspeaker - Avantgarde Acoustic? Hornloudspeaker GMBH
 
POOH,
i remember when those round horns hit the market, I knew right away they were selling image not sound quality. Yes with a two way you can get away with one large horn but as you say it has to beam as the frequency rises. That is the reason that Early has stuck with his constant directivity and all the room treatment it takes to make that work. I designed radial horns that solve the throat section design a long time ago but never produced them, I just moved away from horns at one point. I still enjoy them and know how to make them work but I didn't see a real large market for that. I gave up on chasing the audiophile market, it just cost to much in marketing cost to play in that market.
 
diyAudio Moderator
Joined 2008
Paid Member
I'd be coming from a round 80 degree OS with hypex transition to 180 degrees at a mouth diameter of 80cm for crossing at 700Hz so I fear I have far to fall..

One of the 'wrong' reasons I want to consider radial is so I can try going co-entrant using a pair of 8" mids I have, one per side so that one dimension is less than 1/4wl above the crossover and I can mount the mid on one wall (maybe entering at two places.. eg either side).

I'm prior convinced that round is an appropriate way to commence a 1/4pi corner loaded tweeter horn despite radial seeming ideal.. only to avoid the transition too close to the throat (I won't mind the loading or the flat(ter) mounting surface for the mid though).

You make the initial throat round and morph it to conic, then radial. If you use Altec, GPA, the round conic is built in to cover much of the HF BW, then the transition to conic isn't too critical due the size of the WLs involved as it transitions to radial.
This is where I'm getting stuck. I want the horn to have the flattened profile before the smaller dimension exceeds 6cm, but I want the horn to be round until the mouth exceeds 6cm.

The 1" tweeter is also probably more than an inch deep before the mounting surface FWIW. I can hold any dimension as small as 13 degrees out from there if necessary.

I feel it's safe to hold the round cross section to at least 15cm across but I'd like to cut this as far as I can.

One interesting thing is that if this transition is done right, the rectangular pattern will 'hold' for very long distances,
I'm good with rectangular, so are my walls, naturally the dome isn't. Some choose to make the transition right near the throat, where would be best?

Some folks obsess over this pattern 'distortion', though IME, it's not the rectangular pattern, but the transition section.
If by that you mean the independence of H and V (and hence a continually changing aspect ratio) isn't bad if the source is transitioned well that way, then that's good.
judging by Dr. Geddes's much more recent 'round to round' WG transition observations about how critical they are, I consider it 'carved in stone' now.
I've always felt it best to do aspect transitions either very early or very late, now I'd like to do one in the middle. :eek:

@POOH, 6k? what about higher?
 
@POOH, 6k? what about higher?

In that particular 2386 horn I never used the TAD and the rest of the 2" drivers I have used are more intended for midrange low treble. I would not use a large format driver past 6K anymore because I have found one inch drivers that handle the treble with a lot less "splash" and sizzle. Some drivers I have found that work great in radial horns is the Community M200 (to 3k), Gauss 4000 (4k) and the EV DH1A (5k) with the aluminum DH1012 diaphragms. I have been told by a very reliable source the TAD 4001 can be used higher in radial horns (the TAD radials) without the up front sound you get in the round horns but never tried it. I'm sold on a separate treble horn with 1" drivers. If you have the TAD drivers I suppose you may be happy going higher.
 
diyAudio Moderator
Joined 2008
Paid Member
I'm all B&C but happy at this point. I'm also fairly firm on separate treble but the thought of a single horn two way to 200Hz is shaking my foundation. I thoroughly considered separate horns and each way came out near identical except the mid starts later down the throat.

I want the horn to have the flattened profile before the smaller dimension exceeds 6cm, but I want the horn to be round until the mouth exceeds 6cm.
I want to rephrase this. I have the old plan written down somewhere.. mid entry 7cm from the mount where the mouth is at 12-15cm across, 60-70 degree round OS before that point then hypex T=2 or 3, 80 degrees at 1k, 90 at 700, H out by 500, V out by 350 or thereabouts.

If radial then H would be the same.
 
diyAudio Moderator
Joined 2008
Paid Member
Just make your mid/high horn the same aspect ratio as your bass horn - and Bob's your uncle!
That's where I was at. The mid horn ended up being the hypex section on its own with a lowish compression ratio. I was going to make a mould so I could do four mouth sections.

I read Poohs comments and I must have regressed about a month. Don't worry Pooh, this is just a hobby (though I've been known to help schools and churches). Indecisive is just a state of mind.. I think.