Beyond the Ariel

[CLS]

Originally Posted by cuibono

....

And below 40-50Hz, IMO, is sealed/bandpass sub territory.
...

Ability to pressurize the room is the key to the very bottom LF, yes!

However, with larger cone area (then a single 15" per side), or some more excursion (I prefer the former), you can get down to high 20's ~ low 30's in room, and with a shallow rolloff below that.

For those really serious bass freaks who want and need to get teens, OB can't deliver. But for almost all kinds of music, there's barely any compromise with high quality dipole high 20's ~ low 30's bass.

In fact, most speakers on market can't do this. And a pair of real bass horns which can get down to here might be the size bigger than a bathtub. So OB is a real treat.

[ScottG]

Originally Posted by Lynn Olson

But the increase in excursion (below baffle peak) is no small matter. The added excursion (compared to a closed box) is another 6 dB per octave - of course, it can equalized and additional watts thrown at it, but the extra distortion is there to stay (unless servo feedback is used). The OB increase in excursion is [Ion top of[/I] the normal 12 dB increase in excursion per octave of a closed-box speaker.

Ideally, this would be offset by an increase in cone area - but as you can imagine, the cone area necessary to keep excursion low at the lowest frequencies would truly be immense - the size of a wall.

The baffle peak is a little troubling as well. It is caused by the rear wave coming around the baffle to the front - so it appears in the time domain as well as the frequency domain. Subjectively, it is much less offensive than box modes and panel vibrations, because those have much higher Q's and take a lot longer to ring down to zero - that's the droning sound you hear coming from the side of a conventional box.

Ah, but that's where you design around those limitations. Namely:

1. Cut your low freq.s either with filter and/or by driver limitation, and increase sd (..at least beyond conventional designs).
2. Cut you high freq.s *below* the dominate baffle-step transition.
3. Utilize free gain from a rising impedance when interacting with an amplifier that reacts to it.

..but that's just the OB suggestion.

There is also the Olimpia Audio W416. In a 2 cubic foot "lossy" box with the right amount of flow resistance could work wonders as well.

Much cheaper, but with less gain (and less extension near the average) for a similar aperiodic enclosure could come from:

1. Monacor SP-38/300 (neo like the Olimpia) 1.5 cubic ft.
2. Weber 15A200 (an alnico based motor) 2 cubic ft.
3. Eminence Delta Pro 15A (ferrite) 1.5 cubic ft.

[TrueSound]

Originally Posted by lynn olson

well, that last post certainly set the cat amongst the pigeons!

I'm not married to any one approach - but i do think it's useful to partition deep bass (20~80 hz) from the 80~800 hz range. The solutions are different, and the interactions with the room are different as well. Deep bass is mostly below the lowest-frequency room mode, and is the domain of hoffman's iron law, which is inescapable.

Setting aside the choice of technology, the dominant problems in the 80~800 hz range are room modes, followed by box modes and panel vibration. The drivers are inherently flat, since they are in the piston band, so they are essentially blameless here. There are worries about delta-inductance and arcana of magnet construction, but those are minor compared to room modes, box modes, and panel vibration.

If anyone doubts that box modes or panel vibration are audible, by the way, just put your ear against the side panel of conventional loudspeaker, listen for a while with bass-range or vocal music, than move back to your chair. The droning coloration you heard while listening to the box will still be audible from the sitting position - in fact, now that you know what it sounds like, you won't be able to ignore it!

The marvelous thing about ob's is that this coloration is finally, and wonderfully, all gone. But the increase in excursion (below baffle peak) is no small matter. The added excursion (compared to a closed box) is another 6 db per octave - of course, it can equalized and additional watts thrown at it, but the extra distortion is there to stay (unless servo feedback is used). The ob increase in excursion is [ion top of[/i] the normal 12 db increase in excursion per octave of a closed-box speaker.

Ideally, this would be offset by an increase in cone area - but as you can imagine, the cone area necessary to keep excursion low at the lowest frequencies would truly be immense - the size of a wall.

The baffle peak is a little troubling as well. It is caused by the rear wave coming around the baffle to the front - so it appears in the time domain as well as the frequency domain. Subjectively, it is much less offensive than box modes and panel vibrations, because those have much higher q's and take a lot longer to ring down to zero - that's the droning sound you hear coming from the side of a conventional box.

Each type of approach has measurable, and with listening practice, audible faults. Once you know what box modes and panel vibrations sounds like, they are hard to accept. Once you get used to the effortless quality of horn-loaded bass - especially with piano and symphonic music - direct-radiators don't sound as good. Different people are sensitized to different types of coloration, thus the preference for different types of loudspeakers, electronics, and source materials.

I have had the misfortune of working with loudspeakers for a while, and notice odd sounds that most others don't hear. That doesn't increase my appreciation of hifi, it makes more difficult to listen, since much of the time all i'm aware of is a checklist of faults. The objectivists might say that it's all in my head - and well, yes, it [i]is[/i} all my head, since i can't hear what others are hearing. Who's head would it be in, anyway, if not my own? How can i imagine what anyone else is hearing? I'm not good at that kind of thing.

Less facetiously, i usually hear an improvement when a measurable fault is corrected - although not always. I'm not very sensitive to phase distortion, and others are. Some people go nuts when the absolute phase (of both channels) is inverted, but all i hear is a subtle - sometimes very subtle - timbral change on some instruments, less than a change in seating position, or a track-to-track difference in eq.

What i do hear - in other words, what i care about when listening to a loudspeaker - are cabinet colorations, as well as fairly obvious dynamic-range limitations. These appear to be the major tradeoffs.

lol!

[Lynn Olson]

Originally Posted by ScottG

Ah, but that's where you design around those limitations. Namely:

1. Cut your low freq.s either with filter and/or by driver limitation, and increase sd (.. at least beyond conventional designs).
2. Cut you high freq.s *below* the dominant baffle-step transition.
3. Utilize free gain from a rising impedance when interacting with an amplifier that reacts to it.

..but that's just the OB suggestion.

There is also the Olimpia Audio W416. In a 2 cubic foot "lossy" box with the right amount of flow resistance could work wonders as well.

Much cheaper, but with less gain (and less extension near the average) for a similar aperiodic enclosure could come from:

1. Monacor SP-38/300 (neo like the Olimpia) 1.5 cubic ft.
2. Weber 15A200 (an alnico based motor) 2 cubic ft.
3. Eminence Delta Pro 15A (ferrite) 1.5 cubic ft.

Hmm - all good things to think about.

Going backwards, Point (3): the Zout of the Karna amplifier is slightly less than 1.5 ohms (on the 8-ohm tap). No, I am NOT re-designing the Amity or Karna amplifiers, which have source impedances comparable to many SET amplifiers. That project is done.

The most likely crossover with the horns I have is between 700 and 900 Hz. Referring back to Point (2), that would mean a pretty small baffle, or rather, a pretty small driver - maybe 8" or so? The efficiency is now getting pretty low, not to mention power-handling. The alternative is a much bigger horn for the compression driver - um, no thanks.

Going back to Point (1): Increase Sd - yes, the way to go, solving many problems with distortion, power-handling, and headroom. It conflicts with Point (2), though - a baffle peak at 800 Hz or preferably higher.

The way to square the circle (staying with OB for now) is overlap or frequency-split a large and a small baffle. So - a single efficient 8-incher (this is starting to look like an Orion), with a large (at least 2x15) bass array beneath it. That would cover the desired 80~800 Hz range, at the expense of complexity, since EQ is needed everywhere in the operating range, and the EQ preferably done at the low-level stage ahead of the power amplifier(s).

Hmm. Not saying it's good or bad, just looking at all of the ramifications. I'm not an op-amp guy, so the low-level EQ ain't just a matter of slapping in a Behringer and letting the thing auto-adjust the whole system. I know auto-digital-EQ is standard pro practice these days, but that's now what I want to do. Other folks are fine with it, I'm not.

The Linkwitz Orion is the way to go for people who want a moderate-size speaker, and are fine with a complex array of op-amp equalizers and a bank of 200-watt power amplifiers (the equipment used at SL's demo at the RMAF). Moderate-efficiency drivers that are popular in mainstream high-end loudspeakers, very flat response thanks to SL's careful work and thorough methodology, and a well-packaged system.

However - this is not a system with the headroom of a Klipschorn, Altec A5, or the Oswald's Mill Audio speaker shown at last year's RMAF. No. Headroom of the Orion, not surprisingly, is similar to mainstream high-end speakers using the same drivers. The 6 to 20 dB of boost EQ completely precludes use with SETs or PP DHT amplifiers - unless you want to get all kooky with transmitter-tube amplifiers running at 1000~1500V.

For those of us of the DHT-triode persuasion (please do not try to argue us out of our delusion), an Orion-alike is not an option. It just isn't. Anywhere there's 6 or more dB of EQ, a 8 to 20-watt triode amplifier is not going to be happy. These are very expensive watts, believe me, and throwing them away hurts.

The tradeoff has been with us since the Thirties - Triode watts versus Pentode, true thermal Class A versus Class AB (static, dynamic, or programmed bias), tube versus transistor, and now, Class D, the modern dollar/watt champion. Some people have chosen different stopping points on this spectrum - well, for better or worse, I'm on the low-power end, not because low power is wonderful, but because I like the devices.

Which has delivered me on the doorstep of efficient speakers. I think flow resistance - another Fifties revival - opens up new possibilities for OB speakers, since it reduces the need for equalization and offers interesting possibilities for reduction of baffle peaking, one of the nuisances of OB speakers.

The BFM speakers are absurdly light for their size, thanks to neo-magnet drivers and 1/2" Baltic-Birch plywood construction. They're musical-instrument speakers, after all, and luggability makes a big difference when you're the one schlepping the PA system from gig to gig. As a former set-em-up guy at many CES shows, I sympathize, I do. It would be ten times worse being a roadie for a band, carrying all this heavy junk from show to show.

So it's entirely possible the BFM OmniTop 15 or 2x15 would be a disaster for HiFi applications. Much too thin, much too resonant, too much the PA speaker. But then - I think of how an old-school Klipschorn (the Fifties oldies, not the modern versions) or Altec A5 sounds. Despite the appalling construction and frequency response, they sound a lot better than you'd expect. Frankly, I have no idea what a flat-response Klipschorn or A5 would sound like - different, I guess. Their virtues aren't in what they're doing wrong - the horrific response, etc. - but in something they do right.

Panomaniac can speak to what these kinds of speakers can do, since he re-worked the big Altecs in Paris. I'd like to keep the virtues while sidestepping the worst aspects of late-Forties design practice. If the Olimpia Audio W416 is at the RMAF, it should be interesting to audition it - if my low-fi recordings don't get me thrown out the room first (which is what happened - twice - at last year's show). Maybe I'll just bring the new Beatles reissues - oh wait, then I'll get to hear CD vs LP, Capitol versus Parlophone, and mono versus stereo mix arguments.

Well, it just so happens I have original English-pressed Parlophone LP's (bought in Hong Kong in 1966) in addition to the recent 09-09-09 Beatles re-issue CDs. I bet I could completely sidetrack a demo with pointless A/B comparisons between the formats.

[Telstar]

Originally Posted by gainphile

OB bass don't need to be big. Just look at orion or Phoenix. Mine uses 2x10" /ch, just cheap $14 woofers and I equalise them to 20Hz Q=0.5.

Don't know how loud you guys listen to your system, but mine is loud enough for the other half to complain.

The thing here is: active line-level eq. Passive system can forget OBs as they won't have the ability to eq the bass.

I dont know your exact setup, but my U-frame with single 12" if i give more than 10db boost at 20hz it starts to make the other freq dirty (it goes up to 300hz). Do you use the T-bass circuit as well?

It may be culprit of Waves EQ (I EQ in digital before conversion), I'm planning to try Algorithmix PEQs and Ozone 4.
Or it's just that a single driver cannot go that much low. Anyway from 50hz up it sounds beautifully (better than the Orions that I have listened to), so it's a keeper. To reinforce the very low, I'm thinking of a sub with very open mind (i.e. OB with servo and multiple drivers, tapped horn and even closed box).

[Telstar]

Originally Posted by Lynn Olson

Well, that last post certainly set the cat amongst the pigeons!

I'm not married to any one approach - but I do think it's useful to partition deep bass (20~80 Hz) from the 80~800 Hz range. The solutions are different, and the interactions with the room are different as well. Deep bass is mostly below the lowest-frequency room mode, and is the domain of Hoffman's Iron Law, which is inescapable.

I totally agree with you, Lynn (although I would try to go lower, lets say 60 or 50hz with them idbass OB for sheer quality and immunity to room modes).
This is often neglected because a satisfying bass in the low 30s or even 40s (like the one in OB can be) is usually so pleasant that one dont think that something is missing down there.

If one checks the response of many expensive speakers it is dramatically lacking in the bass foundation. Exceptions are few and usually sound bloated.

Originally Posted by Lynn Olson

Setting aside the choice of technology, the dominant problems in the 80~800 Hz range are room modes...

I have room modes all the way between 40 and 80 hz as well, and much more than above 250 This is probably the main thing that keeps me from doing a closed box sub...

[Lynn Olson]

Originally Posted by agent.5

I think what you guys should consider is using

DR280 (size 24" x 24") from around 130hz up to 1000hz. I just guess that a horn lower mid may integrate better with an AH-425 than open baffle.

From 130hz and below, use BudP's designed cardoid sub and rotate it 90 degree. So, it will sit below the DR280.

I suggest open baffle servo subs by GR-Research and Rythmik.

GR-Research will have its V1 at RMAF; supposely it sounds very good and the open baffle servo sub is good up to 200Hz. Maybe Lynn can listen to it there and let us know what you think.

This is worthy of serious consideration, with the modification of using Gary Pimm's V-Bass below the DR280's. The V-Bass subwoofer works just fine up to 200 Hz - I trust Gary on this, and his measurements are good - and the DR280's make the grade between 130 Hz and 1 kHz.

I've heard the V-Bass - it shakes the room, and has outstanding tonality, especially for a subwoofer. The only downside I see with the DR280 is the requirement for time alignment, which would put the AH425 pretty far back - probably actually behind the DR280, on its own stand.

Thanks for the tip about GR-Research and Rythmik; if they're at the RMAF, I most certainly will give them a listen. Decent subwoofers are kind of rare ducks.

[CLS]

Originally Posted by Lynn Olson

...

The way to square the circle (staying with OB for now) is overlap or frequency-split a large and a small baffle. So - a single efficient 8-incher (this is starting to look like an Orion), with a large (at least 2x15) bass array beneath it. That would cover the desired 80~800 Hz range, at the expense of complexity, since EQ is needed everywhere in the operating range, and the EQ preferably done at the low-level stage ahead of the power amplifier(s).

....

If very uniform directivity is not your first priority (I remember it's not), then you might take some bigger than 8 incher drivers for, say, 300~1kHz. 10 or 12 inchers in small open baffle can cover this range with very little EQ, so there won't be a bank of OP amps. Sensitivity can approach 100dB/w. Dynamics, headroom sort of things will never be an issue.

2 x 15 inchers cover 300Hz and below will need some EQ and/or some more power, but no big deal, either. EQ requirement will also be very small for the 80Hz target, or even none if high Q drivers are used. (If more EQ is tolerable, then this portion should be capable of reaching 30's.)

If the ragged responses of K-horn, A5... etc. can be accepted because of the 'right' things they do, then there'll surely be a doable way in between them and that 'over-engineered' Orion -- a much less EQ'ed, coloration-free, still highly sensitive and dynamic, and almost fullrange system.

Of course there're so many faults everywhere, but in reality they are not that bad as expected and all can be dealt with. If box/panel resonances are so annoying and can't be tolerated, then OB is the only way!

[Lynn Olson]

Here's an interesting thread comparing the OmniTop 12 to other BFM models - at the end, there are graphs comparing the DR280 to an OmniTop 2x12, with both models using the same drivers. (OmniTop with dual 12-inchers in blue, and a DR280 with a single 12-incher in black. Both graphs normalized to 1 watt, not 2.83 volts.)

By the way, an OmniTop 2x12 is not a large or particularly heavy speaker, measuring 28"H x 22"W x 18"D and weighing 60 pounds, which includes the weight of a pair of Eminence Deltalite-II 2512 drivers. At 106 dB/metre/watt, efficiency is in the Klipschorn range, with much better flatness. With Altec, GPA, and Radian large-format compression drivers 110~111 dB efficient (requiring only moderate attenuation), a SET with a single 45 power tube would be an entirely practical power amplifier.

[gainphile]

To OB proponents here, do you find the lack of physical impact disturbing?

I find it funny sometimes that there is total lack of pressure/slam. Somewhat not real. I know very deep, clean OB organ sounds realistic but I think the brain expects that for something that low there should be some real vibration in the air/furniture.

I know there's "vibrator" equipments out there. I wonder if it makes impact if I install it on my sofa. I am aware of studies such as human multimodal sensory perception as blogged by Harman here:

http://seanolive.blogspot.com/2009/0...with_7561.html

Failing that perhaps sealed sub from 20-30Hz is the way to go, although I'm very skeptical about the omni radiation that it will project upon the existing dipole radiator.