So - why DON'T people go with open baffle?

[Deward Hastings]

Re: So - why DON'T people go with open baffle?

A question for the folks who own dipoles: How far away from the wall do you need to put 'em?

Mine ar four feet from the front and five feet from the back of built-in wall-to-wall bookcase (with a top shelf at five feet high) . . . I guess you'd call that five feet from the wall, with diffusers . Sometimes I move them out a foot further, sometimes (social occasions) a foot back (that's the limit of motion . . . the cables feed through the floor to amps in the basement, and there's not much slack). It doesn't make a lot of difference, but further out from the wall seems "better".

Also, what's to be done about tweeters? Using a monopole tweeter seems just a bit inelegant for this particular application, but two domes is kludgy and expensive.

It's a problem, and one of those "compromise" things. We'd all love a (good, cheap) planar with sufficient output to below 1500 Hz. that didn't beam in the vertical axis (or the horizontal). Let me know if you find one. Going to a "two-way tweeter" has some advantages and some disadvantages . . . you can get a better horizontal polar at the cost of the vertical polar . . . we can argue endlessly about which gives the better overall response in the listening area. You could probably substitute RS28 tweeters for the Millenniums and save a bunch of money with little to no loss of performance, but the RS28 is even deeper, which would make the mounting even more problamatic. Scan has shallower tweeters that would certainly work, but they're just as expensive. You can *never* optimize for just one thing and come out with a good speaker.

some subwoofers capable of a higher xover point than the XLS subs he used.

It's the same "compromise" question again. The XLS drivers are fine at least an octave higher than they're used in ORION, what limits them in ORION is cavity resonance in the H frame. And it is as deep as it is so the speaker can be as narrow as it is (to match the midrange baffle and present a smaller footprint in the room). They could easily go higher if the baffle was wider . . . but . . .

Interesting. It seems that the big reason for non-dipole speakers is mostly just that dipoles are about as cost-effective as using gold-plated Ferraris as delivery vans.

It depends on what you're optimizing for, and what compromises are acceptable. Full range dipoles are quite cost-effective *for what they do best* . . . accurate reproduction of acoustic music in a "medium" sized listening room. I'd put them far from "first choice" for a dedicated HT installation, with it's very different acoustics and SPL requirements. They're not "party speakers" either . . . "dipole" and "disco ball" may both start with "d", but they don't belong in the same room. Different musical tastes and different listening environments point to different speakers . . . the wrong speaker is never "cost effective", the right one, in the right place, may be "cost effective" regardless the price.

[Deward Hastings]

Re: So - why DON'T people go with open baffle?

Here's an excellent paper on dipoles/baffles and some of the design compromises they entail. There's an interesting aside on the ORION+ rear tweeter towards the end . . .

http://www.dipolplus.de/frameset.htm?/thema11.php

[johnk...]

Re: So - why DON'T people go with open baffle?

It depends on what you're optimizing for, and what compromises are acceptable. Full range dipoles are quite cost-effective *for what they do best* . . . accurate reproduction of acoustic music in a "medium" sized listening room. I'd put them far from "first choice" for a dedicated HT installation, with it's very different acoustics and SPL requirements. They're not "party speakers" either . . . "dipole" and "disco ball" may both start with "d", but they don't belong in the same room. Different musical tastes and different listening environments point to different speakers . . . the wrong speaker is never "cost effective", the right one, in the right place, may be "cost effective" regardless the price.

FWIW. My simulations of an H frame circular cross section of equivalent area to the Orion H yield a resonance of arond 300 Hz with about a +6dB peak and fairly low Q. My measurements of the Orion H resoance were even less.

[Deward Hastings]

Re: So - why DON'T people go with open baffle?

So, what is the advantage, if any, of dipole over bipole

I’ve been thinking about your “rear tweeter” question, and wondering why it’s been gnawing at me, and I think I have an answer . . . an answer broader than just “the phase doesn’t matter”. There’s a much bigger answer that addresses why we (some of us, anyway) do dipoles in the first place. There is, after all, nothing “magic” about “dipole sound”, and it makes no sense to fetishize a polar pattern. Our ears hear direct and early reflected sound from the loudspeaker, and reflected/reverberant sound from the room (as excited by the loudspeaker) . . . they don’t hear a polar pattern, only the effect of how that pattern excites the room, and (on the listening axis) the ear directly.

We favor dipole woofers (despite their size and/or inefficiency) because of the uniformity of the way they excite the room (constant pattern) and the way they avoid exciting at least some room modes. We favor them also because it is easy to avoid resonances, and easy to adjust the Q of the eventual LF roll-off to “critically damped”. And we favor them because they blend seamlessly with dipole mid-bass and midrange.

We favor dipole mid-bass and midrange because of the uniformity of the way they excite the room (no “baffle step”) and because we want to eliminate box resonance, and once a driver is “out of the box” it is inherently dipolar (so really there’s not much choice). Just accepting the dipole pattern and running with it makes for a “cleaner” midrange, again not because a dipole pattern is “magic”, but because it avoids all sorts of negative effects of “speaker-in-a-box”. And the uniform room illumination from the dipole midrange blends seamlessly with the uniform room illumination from the dipole woofer.

We favor dipole tweeters because (in for a penny, in for a pound) we want uniform excitation of the room to match the woofer and mids . . . but here physics starts to work against us (a not unusual occurrence, I'm afraid). Drivers and baffles behave differently as they become large relative to wavelength, and other radiation patterns begin to dominate the dipole pattern, and getting uniform excitation of the room becomes more difficult. To maintain uniform horizontal dispersion the dipole driver/baffle needs to become narrow. To maintain output it then needs to become tall. Once tall there is beaming and lobing, and we don’t have uniform illumination of the room (to match the midrange) any more. Small “point source” dipole drivers that would avoid beaming simply don’t exist, so we’re into the world of compromise. Matching the forward pattern of the midrange suggests uniformity on all axis . . . a conventional dome tweeter does that, and over a very small portion of its range has a half-dipole-like radiation pattern. Above that it beams, and below that it “blooms” . . . clearly a waveguide (big enough to control the bloom) would help. Of course that raises other problems (like lobing at the M/T crossover), so there are tradeoffs, and things that just have to be lived with until something better comes along. We’re still waiting . . .

Getting a tweeter radiation pattern to the rear that matches the midrange is even more difficult, and where even more compromise must occur. Tweeters have depth, back to back mounting means acoustic center displacement, bad things can happen. It’s probably not as bad (for uniform radiation pattern) as a long-ribbon beaming, but it’s not what we want, either. About the only thing we have going for us is that the rear sound will be de-correlated anyway (by diffusion and reflection) . . . if we can get smoothly through the M/T crossover with uniform total radiated energy .

In the end the proof of the pudding is in the listening area. Do we have uniform direct and early (first) reflection sound, and do we have uniform reflected/reverberant sound, *for the listener*. It hardly sounds like I’m talking about “dipoles” any more, eh? But that’s the point. A dipole radiation pattern is not the end in itself, it’s just a means to the end, and just one of many means in the complete speaker system at that. There are other performance considerations that favor full range dipoles, and there are practical considerations that work against them. Full range omni speakers satisfy many of the same goals that dipole designs aim for. There are a few other designs that can maintain good uniformity over their full range . . . corner horns, for example. Dipole line arrays avoid vertical pattern issues, at least if you’re far enough back. Design is all in the details, and the necessary compromises.

And yes, I think a lot of “conventional’ speakers would be improved if pulled out into the room and made bipole (quasi omni). It’s not enough to just add a rear tweeter, though, they’d have to be bipole all the way down to baffle step . . . and they'd still be "in a box" . . .

[johnk...]

Re: So - why DON'T people go with open baffle?

We favor dipole woofers (despite their size and/or inefficiency) because of the uniformity of the way they excite the room (constant pattern) and the way they avoid exciting at least some room modes.

Not that I agree or disagree with what you have expressed here, but this is just repeating hype that just doesn't hold up in the real world, or even when modeled correctly. The the idea that dipoles woofers excite room modes in a vastly different manor than other sources is basically poppycock. There are so many caveats associated with that statement that it is simply misleading. I'm not attributing the statement to you; we all knows its origin. So let's look at the situation a in a little more depth rather than base any conclusion on 1-D equivalent circuit models with strategically place source and listener.

First of all, the best that can be said is that dipoles excite modes differently. The result is that under very specific conditions a dipole woofer won't excite certain modes. Those conditions are, in part, that the room be rectangular; that the axial modes be orthogonal; and the dipole axis be aligned with a room axis. But how often does the listening room meet those conditions? And when it does, how often are the speakers aligned so the dipole axis is aligned with a room axis (for example, the axis of the woofer is perpendicular to the wall behind the speaker when the room is rectangular. A more typical placement is to angle the speakers toward the listener so that the axis of the stereo pair crosses in the vicinity of the listeners head, perhaps behind, perhaps in front. Under such conditions the dipole woofer is capable of exciting just as many modes as a monopole or cardioid, and it will. The modes may be excited differently, but they will be excited just the same. I performed a study of this based on a full 3-d room mode analysis years ago. Skipping the math and jumping to the bottom of the page the results are as shown below.

Room configuration:



Result for monopole, dipole and cardioid:



So, other than the case where the sources are in the corner where the dipole can not produce low frequency (remember the previous discussion of dipole woofers close to reflecting walls?) I ask, how can it be concluded that a dipole is any better or worse that a monopole or cardioid based on modal analysis when the positioning is more typical off actual placement in a room? The only conclusion which can be made with any certainty is that below the room fundamental the dipole woofer does not pressurize the room, as is apparent by the relatively flat tail below the first mode. The monopole and cardioid do pressurize the room. Consider that an advantage or disadvantage as you like. It might also be of interest to note that as the position of the sources moves around, the level of the first room mode changes considerably for the dipole and monopole woofer. But for cardioid woofer the level of the first mode remains relatively constant. So, at least for the fundamental mode, the cardioid appears to be least sensitive to position.

There is another thing to consider with stereo speakers positioned in a rectangular room. In such a case with the speakers set up symmetrically many of the side to side and floor to floor modes, being asymmetrical, will not be excited simply because one woofer excites the mode 180 degrees out of phase with the other. This also has impact on many of the oblique and tangential modes.

Thus the idea that dipole woofers are, in general, in any way superior to other woofer formats, based on excitation or room modes, is pretty far from the truth as shown by 3-D modal analysis.

[Deward Hastings]

Re: So - why DON'T people go with open baffle?

Thus the idea that dipole woofers are, in general, in any way superior to other woofer formats, based on excitation or room modes, is pretty far from the truth as shown by 3-D modal analysis.

In my experience when I replaced Thiel 3.6 with ORION at almost exactly the same location in my listening room the excitation of room modes was reduced and the perceived "quality" of the bass improved considerably. And I do know from discussion with others who have made a similar transition from omni to dipole woofers that mine is not a unique perception.

Of course a few contrary examples do not disprove a theory . . . there are a lot of confounding influences in bass reproduction, and there may well be examples out there where switching from omni to dipole did not make a difference (and I just haven't encountered any of them), or where the switch even favored the omni (pressurization being one explanation, especially where the room is small). I'm certainly not recommending dipole bass for a car cabin.

And I've got no objection to "theory" . . . most of my life, in chemistry and electronics, has been governed by it. But . . . I've also learned that when theory and practice diverge it's something wrong with, or missing from, the theory, not something missing from reality, that is at fault. It is dangerous to use theory as polemic . . . it is far more productive to use it as a guide in seeking (and perhaps eventually explaining) what "really works". In my experience dipole bass really works, and that leaves me . . . puzzled . . . about a theory that says it doesn't.

[Sydney]

Re: So - why DON'T people go with open baffle?

From this observers opinion: I do see some agreement in this discussion:
Modeling does not include the variables that prevent a simulation from being an accurate predictor of the the end result.
It's safe to say that no one lives in a theoretical rectilinear model; where the room can be sealed to allow pressurization, and the boundaries reflect everything, with no absorption.
And that every design has to take the surroundings into consideration:
A bad room beats the best design, a design is constrained by it's implementation.

[johnk...]

Re: So - why DON'T people go with open baffle?

In my experience when I replaced Thiel 3.6 with ORION at almost exactly the same location in my listening room the excitation of room modes was reduced and the perceived "quality" of the bass improved considerably. And I do know from discussion with others who have made a similar transition from omni to dipole woofers that mine is not a unique perception.

How do you know they modes were "reduced". What we hear and measure is the summation over all the excited modes, all modes contribute to all frequencies. Exciting fewer modes may not even lead to smother response over all. Dipole woofers may well sound different then monopole and monopoles and dipoles different from cardioids. Different just means different modes are excited in different ways. It also means that transients are different.

I'm not knocking that different woofer systems will sound different. I'm knocking the ideal that the simple statement that dipoles excite fewer modes is not consistent with anything but some very specific cases.

I would also point out that measurements support that the in room response of dipole, monopoles and cardioids generally show similar behavior, as predicted. The position of dips and peaks may move around but the response isn't all of a sudden the result of only axial modes being excited.

To make a legitimate comparison what is necessary if to compare the different woofer formats when they all have the same free field axial response.

[Deward Hastings]

Re: So - why DON'T people go with open baffle?

How do you know they modes were "reduced".

Measurement, and listening . . .

transients are different.

Yes . . . any modal excitation represents a resonance that degrades transient response.

I'm knocking the ideal that the simple statement that dipoles excite fewer modes is not consistent with anything but some very specific cases.

One such "very specific case" is the floor/ceiling mode, which is generally excited by omni radiators (and in theory by cardiods) but less so by dipoles. Since most rooms have floors and ceilings that "very specific case" is fairly common . . .

To make a legitimate comparison what is necessary if to compare the different woofer formats when they all have the same free field axial response.

To make a "simple" model what is necessary is to make significant assumptions (idealizations) about room shape and boundary behavior, and about the "idealized" (not actual) on and off axis response of the woofers being compared. In a case where such simplifications do not apply (not simple rectangular rooms), or where the actual room parameters (like boundary reflection/absorption) are difficult to establish, we're left with observation and experience. In my experience, in my listening room, dipoles work better. In my experience the *simplest* models (of where the axial nodes will occur) also work (pretty well). Beyond that . . . confusion. Your experience, in your room, may be different.

[dantheman]

Re: So - why DON'T people go with open baffle?

I think what we are ignoring in this discussion about dipole/omni bass and the differences we hear are SBIR issues. Maybe they don't effect 'sub' frequencies, but they definitely effect more audible frequencies where the ear still has a long integration time. You'd have to think Dipoles have the advantage there and I know many people think dipole midbass is clearly better(including me in every system I've heard it). Lateral interference should be minimized with OB.

Dan

More about it here:

http://dtmblabber.blogspot.com/2010/12/little-more-into-boundary-conditions.html

[spasticteapot]

Re: So - why DON'T people go with open baffle?

A very nice local DIYer let me listen to his "NaO clones". He used some ScanSpeak woofers in place of the spec'd HDS woofers and a standard W-frame in place of the woofer-in-a-tube setup. A DEQ2496 was used for the tweeters, mids, and midbass arrangement, with the woofers on some rather crude active crossovers. The top section was EQd razor-flat.

A few things I noticed:

1. Plenty of bass. Anyone who is complaining about the bass needs his head examined. But maybe it's just me.

2. The bass sounded pretty much the same as a standard boxed woofer, to be honest.

3. The sweet spot was surprisingly narrow. This may have been a result of the non-standard waveguide on the tweeter, but the claims of an enormous sweet spot I've heard from some Pluto reviews are invalidated.

4. There was a vast amount of room interference, which surprised me. This was an incredibly un-ideal room full of hard surfaces and no room treatment, but it did a good job of illustrating the need for a dipole to be placed in a suitable room. My speakers will likely be used in less-than-ideal circumstances, so reduced dispersion may not be a disadvantage.

Looks like I'll be going for standard boxes, I guess. Thanks for the help!

[benchtester]

Re: So - why DON'T people go with open baffle?

A very nice local DIYer let me listen to his "NaO clones". He used some ScanSpeak woofers in place of the spec'd HDS woofers and a standard W-frame in place of the woofer-in-a-tube setup. A DEQ2496 was used for the tweeters, mids, and midbass arrangement, with the woofers on some rather crude active crossovers. The top section was EQd razor-flat.

A few things I noticed:

1. Plenty of bass. Anyone who is complaining about the bass needs his head examined. But maybe it's just me.

2. The bass sounded pretty much the same as a standard boxed woofer, to be honest.

3. The sweet spot was surprisingly narrow. This may have been a result of the non-standard waveguide on the tweeter, but the claims of an enormous sweet spot I've heard from some Pluto reviews are invalidated.

4. There was a vast amount of room interference, which surprised me. This was an incredibly un-ideal room full of hard surfaces and no room treatment, but it did a good job of illustrating the need for a dipole to be placed in a suitable room. My speakers will likely be used in less-than-ideal circumstances, so reduced dispersion may not be a disadvantage.

Looks like I'll be going for standard boxes, I guess. Thanks for the help!

Nice write-up.

(However, one comment about #3, the Pluto is a quasi omni-directional. Linkwitz's "Orion" is the dipole.)

[MSaturn]

Re: So - why DON'T people go with open baffle?

Indeed, the enormous 'sweet spot' is due to the fact you're already totally off the L16's axis, so very little response change happens as you move around.

[spasticteapot]

Re: So - why DON'T people go with open baffle?

Nice write-up.

(However, one comment about #3, the Pluto is a quasi omni-directional. Linkwitz's "Orion" is the dipole.)

That was a typo - I meant to say "Orion."

[Deward Hastings]

Re: So - why DON'T people go with open baffle?

I think what we are ignoring in this discussion about dipole/omni bass and the differences we hear are SBIR issues.

Hardly being ignored . . . the need to place dipoles well away from the walls (particularly the front wall) is well understood, as is the need for a diffusive front wall for simultaneous good imaging and broad soundstage. The ability of the "mind's ear" to tune out the room when it is uniformly illuminated is only partial . . . it certainly doesn't mean that dipoles or omnis can be placed in just any room and always sound their best. In some very bad rooms otherwise horrible speakers may well sound better than dipoles or omnis. The speaker is just one part of the reproduction system . . . it must work with the room in which it is placed.