Beyond the Ariel

I can get my 2 cm throat police sirens to reach 200Hz easy in an elongated Goto S-150. Using such a small diaphragm down to 200Hz seem to give much higher clarity than e.g. a Fane S 8M, which I rank as one of the top contenders to work good in a horn. I still need to fix a proper back chamber for the Fane but I can't imagine it beating the police siren.

Hey Rewind,

I'd like to hear more about your experience with the siren drivers. I bought some San Ming SD-210RX to try in midrange Karlson tubes, but I haven't had the time to get them going yet. What brand are your drivers? How much bandwidth do you get out of them? Do they need any EQ? Here's a link to my 3-way Karlson idea.
 
Member
Joined 2008
Paid Member
Timely

Pano,
I agree and think that most of what people think is the horn loading of the front and back loaded horns is just room gain. I think if you took most of these designs and put them outside and made a correct measurement, even though it is not an anechoic measurement you would see that these claims just don't hold water. The tapped horns which to me are really transmission lines have a better chance of working but those have other issues. I just don't see why the fixation on low frequency horns in homes, the integration of the multiple horns is also a problem that seems to be overlooked in almost all instances. There is nothing wrong with combining a direct radiator for the bottom end with a horn system that starts at a reasonable point. As long as the bass section can integrate with the horn and can produce the necessary spl level to keep up that is the way I would go.

Well, I must say this current topic is very timely for me. In a previous post I considered using my pair of Altec 416-8B's for mid-bass. Now it seems an audiophile friend has offered me his pair of the same drivers. He uses JBL D-130's for total bass duties, and he knows I am fixated on double 15's per side.
So, keeping in mind that I have a SUPERB subwoofer system for operation below 60Hz, my intent is to create a full mid-bass section covering from 60-
500Hz, above which will probably be the large Sierra-Brooks horn. My friend made the suggestion to use my 20" i.d. sonotubes to create a double stack, but that raises the height of the horn up a ways to where I am uncomfortable
with the image height. A side by side orienation has me concerned about a horizontal 500 Hz propogation doublet. So, I am really stuck for ideas.
 
When adding a horn to my Fane 8M I went from normal speaker to hornloaded midbass, not at all tuned or correct, as this was just a prototype, I got something that fit much better with the upper horns. Not hornloading midbass works, but it will have a complete other type of sound than the upper horns. I am listening to a 3-way right now composed of a Beyma TPL-150 + hornloaded JA6681B (1" compression driver) + JBL 2204H (12" midbass cone driver) at the moment, and the midbass is completly off. xD
 
Rewind, if you get the bass horn and low bass horn setup right with your mid treble horn I imagine you'll become another satisfied music lover horny that is amazed with every listen. It's like a hybrid electrostatic mid/treble and dynamic bass, it can be satisfying but sucks compared to a real full range electrostat in your home.
 
Well, I must say this current topic is very timely for me. In a previous post I considered using my pair of Altec 416-8B's for mid-bass. Now it seems an audiophile friend has offered me his pair of the same drivers. He uses JBL D-130's for total bass duties, and he knows I am fixated on double 15's per side.

So, keeping in mind that I have a SUPERB subwoofer system for operation below 60Hz, my intent is to create a full mid-bass section covering from 60-500Hz, above which will probably be the large Sierra-Brooks horn. My friend made the suggestion to use my 20" i.d. sonotubes to create a double stack, but that raises the height of the horn up a ways to where I am uncomfortable with the image height. A side by side orienation has me concerned about a horizontal 500 Hz propogation doublet. So, I am really stuck for ideas.

Well, I've tried a lot of dual-woofer setups. The 2.5 tends to overcompensate for BSC, resulting in a slow, heavy sound. My own feeling about BSC is that it is frequently overdone, giving an exact compensation for outdoor use, but resulting in much too heavy bass-heavy sound in a typical listening room. I prefer to make minor subjective adjustments in the crossover.

A vertical WWMT or MMT array is popular in audiophile speakers (Wilson et al), and I've done designs like these myself, but I've never been all that happy with the resulting sound. I suspect lack of coherence in the arrival-time of the two woofers, but I'm not sure why they don't sound better than they do.

The Ariel uses a slightly modified MTM array, but it took a long time to get the subjective balance to match the measurements. And that's with very small drivers, an intentionally asymmetric driver placement on the front panel (tweeter offset to one side), and a low diffraction enclosure. I tried symmetric MTM's and didn't like the sound at all; the symmetric location of the tweeter emphasizes front-panel diffraction effects, and the upper-midrange diffraction coloration cannot be removed with equalization.

I've yet to hear a commercial high-end speaker with large-diameter woofers in a MTM or WMTMW vertical array that didn't sound very artificial and unnatural. Part of the problem might be subjective; in the real world, bass instruments don't radiate their sound in mid-air, but close to the floor, and you expect to hear and feel the floor sensation with bass. Another possible explanation is the upper woofer sees a substantially different air-load than the woofer closest to the floor (which sees its image), and that unbalances the acoustic pair-matching of the two drivers.

The big pro studio monitors using side-by-side 15" drivers actually have pretty good midbass coherence, with a lot of impact and snap to the sound. I might quibble about the sound of the studio-monitor horn, but have no complaints at all about the sound of the bass, with realistic impact and weight to the drums and piano.

If all the drivers are within 5 degrees of each other at crossover, having them a wavelength apart is an acceptable option, and side-by-side with the horn directly above and between gives the closest possible spacing (like a dot triad in a color CRT). If the drivers have the layout mentioned in the previous pages (each bass driver is 22.5 degrees off-axis from the listener), it should be possible to get a pretty good time-alignment at the listening position.

I grant that much of the foregoing is subjective preference, but when all of the speakers (including my own designs) using a given driver layout sound a certain way, I pay attention. In particular, I am not sure why MMT's don't sound better, but they seem to fall into the "sounds worse than it measures" category.
 
Last edited:
The quest of speaker placement to get the most natural reproduction is maibe worth a own thread. I have had frequencies below 1,2khz reproduced by my 38" lower midrange horn. Since a good part of the voice range is reproduced in this frequency range, a visitor made this remark, and said voices should sound from more above, they were from too low, not sounding natural.Certainly not wrong. But what about different instruments, each radiating in a different kind, environment, different propagation, beside different recordings, some with just one mic, other with various ? How about organ in churches etc ? Does that not bring to a conclusion, that the aim of the most natural reproduction should be only in a certain way and limitation ? The aim of for example the most natural reproduction of voices, the right timbre and tone, is a very valid one. But the aim of reproducing the way, how the soundwaves are propagated at the life event and arrive, the correct size, location etc. is maibe a false one. If you listen to a big orchestra, the origin of the instruments can be 20m, 30m away. That might result in a big, open and deep powerful sound. If you have a multiway system, it might be able to reproduce a orchestra the best way. If you use however a single wideband speaker, erradiating just from two single stereo points, they might struggle harder. If you have however just one Singer, lets say with a guitar, or solo instruments or so, these might be better reproduced with single widebanders, rather than a big multiway speakers. So it will be always a compromise. I was disturbed of having bass only at the half of the room, close to the floor, and it pleases me more to have a line bass array, radiating and filling the whole room with bass. But that is just a subjective preference, without the aim to reproduce the most natural possible.
 
I prefer to make minor subjective adjustments in the crossover.
Also why I like to floor couple, crossed low. Removes the excess, that muddy thick which mars the midbass low mid.

A vertical WWMT or MMT array is popular in audiophile speakers (Wilson et al), and I've done designs like these myself, but I've never been all that happy with the resulting sound. I suspect lack of coherence in the arrival-time of the two woofers, but I'm not sure why they don't sound better than they do.
Would agree about the timing issue. Adding two additional ceiling bounces can't help either. Think of the cascading wavefront. On axis simple timing errors due to the acoustical time alignment, but baffle shape included has an effect, as is listening distance all crucial. But when we look at the upper and lower floor bounce interaction, the interphase of this sum upon the direct arrival is adding/subtracting.


The Ariel uses a slightly modified MTM array, but it took a long time to get the subjective balance to match the measurements. And that's with very small drivers, an intentionally asymmetric driver placement on the front panel (tweeter offset to one side), and a low diffraction enclosure. I tried symmetric MTM's and didn't like the sound at all; the symmetric location of the tweeter emphasizes front-panel diffraction effects, and the upper-midrange diffraction coloration cannot be removed with equalization.

I've yet to hear a commercial high-end speaker with large-diameter woofers in a MTM or WMTMW vertical array that didn't sound very artificial and unnatural. Part of the problem might be subjective; in the real world, bass instruments don't radiate their sound in mid-air, but close to the floor, and you expect to hear and feel the floor sensation with bass. Another possible explanation is the upper woofer sees a substantially different air-load than the woofer closest to the floor (which sees its image), and that unbalances the acoustic pair-matching of the two drivers.

Concur completely...

The big pro studio monitors using side-by-side 15" drivers actually have pretty good midbass coherence, with a lot of impact and snap to the sound. I might quibble about the sound of the studio-monitor horn, but have no complaints at all about the sound of the bass, with realistic impact and weight to the drums and piano.

If all the drivers are within 5 degrees of each other at crossover, having them a wavelength apart is an acceptable option, and side-by-side with the horn directly above and between gives the closest possible spacing (like a dot triad in a color CRT). If the drivers have the layout mentioned in the previous pages (each bass driver is 22.5 degrees off-axis from the listener), it should be possible to get a pretty good time-alignment at the listening position.

I grant that much of the foregoing is subjective preference, but when all of the speakers (including my own designs) using a given driver layout sound a certain way, I pay attention. In particular, I am not sure why MMT's don't sound better, but they seem to fall into the "sounds worse than it measures" category.

I've done some sims using an MLTL placed at various points within another enclosure, that we can call a room. Depending on shape and placement, the interaction of the room has a direct effect. The side by side placement of the midbass's being floor loaded yielded best results so far. What would that acronym be? MTM/MM or WW

Perhaps this CTC distance is critical if crossed high (>100), the same rules should apply.

Also was thinking about antennas again (DUCK) and how the E/H wave propagation travels off of an elliptic paraboloid where the dish is orientation is elliptical in width and narrow vertically. This produces a beam which is shifted 90°. This results in a narrow horizontal but tall narrow beam. Even tho we are working at far lower frequencies this is the same as how we control the off axis horizontal and vertical polars with an MTM. Thus if this holds true for both, then would it be too far off to conclude that at these frequencies would have similar effect down to the very bottom, or rather an interaction of phase due to the constricted wavelengths within the room?
 
Well, it's entirely possible to design loudspeakers that measure well and sound unnatural. The large WMTMW's fall in this category ... every one I've heard sounds disjointed, with a very strange and weirdly unnatural spectral balance. The big woofer perched in mid-air has a lot to do with it; it just doesn't sound right, no matter how it measures. I sometimes wonder if the designers of these speakers have ever heard live acoustic music, or if they are using stadium rock PA systems as a reference. This isn't meant to be a snide comment; it's just that some modern high-dollar high-end speakers have some of the most bizarre colorations I've ever heard.

The problem with a MMT array is much more subtle, a sort of blurred bass that has a noticeable level dependence. I don't think the level dependence is actually there, but is a perceptual-threshold effect, where you start noticing the problems in the time domain if the playback level is high enough. The MMT "sound" is most noticeable if there's a version of the speaker (from the same manufacturer) that uses a single woofer instead of two; the single-woofer version almost always sounds crisper and more coherent. That tells me there's something wrong with how the MMT interacts with the room.

It's easy to forget that although the floor reflection is annoying when measuring a loudspeaker, we expect to hear it with real-life sound sources ... at all times. The only way to avoid a floor or ground bounce is to sit in a tree, or at the edge of a balcony ... neither is a common way of listening to music, singing, or speech.

Close miking, which is pretty much universal practice for pop music and classical soloists, effectively removes the floor bounce by swamping it out with the much louder direct sound from the singer (who is only a few inches from the microphone). The floor bounce is re-created in the listener's living room ... in a way, it is an essential part of the recording, even though it is re-constituted on replay, and is not actually on the original recording.

The essential nature of the floor bounce suggests driver locations that are close to ear level, with the bass radiator below ear level (as it would be in real life), and the HF radiator at or slightly above ear level (again, as it would be in real life).

When I say "real life", by the way, I mean where the sound would come from if real instruments and singers were actually in your living room. The sound from the vocalist would not come from the floor, and the sound from the drums would not come from above ear level (except for the cymbals, which are reproduced by the HF radiator). A grand piano, as it does in real life, should sound physically large and solid, with a noticeable floor component to the sound.

Although I've always loved electrostats, I have to admit they don't do piano very well; the sound is too insubstantial, too airy and wispy, to suggest a real-life grand piano. Really large electrostatic arrays just stretch out the sound, over a whole wall sometimes, while not adding much in the way of weight or solidity.

This is where large direct-radiators and horns rule; getting the big, powerful instruments right. When you hear a Bosendorfer played quietly, it's still very obvious it's a grand piano, and is just a moment away from a startling crescendo. This slight sense of tension is lost with many speakers that miniaturize the instrument.
 
Last edited:
Piano should be monstrous - but recordings are often polite. Go and buy 'Night Kitchen' by Ross Bolleter on ruined pianos "inspired by Goya’s Black Paintings; others by the desert country of Central Australia, especially the area round Ngayurru (Lake Mackay)"

EMANEM 5008: ROSS BOLLETER

I recorded these pianos to reel to reel tape. Think of a saloon bar Jefferson hauled from Chicago in the gold rush and deconstructed by the western Australian desert - bass string pulls that will make your double woofers cry. Ross tells me the horns bring it to life like nothing else. Seriously, this music will test your perceptions and bass reproducers.

Lynn, I sent you a copy - did you try it?

martin
 
The Elephant in the Room.

IF _for the love of tube amplifiers_ someone decides to use a 40" wide bass cabinet with two 16" side-by-side woofers, they would probably also consider a larger horn.

JMLC 270 is about 28" in diameter and setup for 1.4" compression drivers. JMLC 200T is about 38" in diameter and is setup for 2" compression drivers, and is available with a 1.4" reducer.

I have not heard these horns.
 

Attachments

  • elephant_28.jpg
    elephant_28.jpg
    85.5 KB · Views: 578
  • elephant_38.jpg
    elephant_38.jpg
    97 KB · Views: 573
Member
Joined 2008
Paid Member
BMS 4592

Impressive, but to be honest I am at a loss to see the connection to
The Elephant in the Room?
Using the 4592 co-ax (anyone have experience with this unit?) in a large horn like that, with the 2 16's (Altec?) and a true dedicated subwoofer below 60Hz,
well, this looks like the Ultimate 4 way. I mean, the end all, be all !!
 

ra7

Member
Joined 2009
Paid Member
Umm... not quite. Well, not for me anyway. I have a similarly large tractrix horn and I've tried it with a BMS4592. Sure, there is extension, and yes, it is point source. But the horn still determines the dispersion. In other words, it beams like a laser in the HF, and gives unstable imaging.