In the inaugural issue of the new online publication "HFi Zine", I have an article on the controlled-pattern offset bipole loudspeaker configuration. I use this configuration in several of my commercial designs, including one that received a Golden Ear award from The Absolute Sound in 2008. Near the end of the article is a list of suggestions for DIYers.
http://www.hifizine.com/2010/06/the-controlled-pattern-offset-bipole-loudspeaker/
http://www.hifizine.com/2010/06/the-controlled-pattern-offset-bipole-loudspeaker/
Thanks for the article, very informative.
I can't stop wondering how a sealed three-way design would work in an offset bipole configuation using the crossover between the midrange and midbass to help control some of the wrap around deficiencies.
My thought would be a three-way that is designed to maximize efficiency, quality of the midbass and midrange, while using four distributed sealed subwoofers from 60Hz or so to provide low end bass and control room nodes?
Any thoughts?
Thanks
Norris
I can't stop wondering how a sealed three-way design would work in an offset bipole configuation using the crossover between the midrange and midbass to help control some of the wrap around deficiencies.
My thought would be a three-way that is designed to maximize efficiency, quality of the midbass and midrange, while using four distributed sealed subwoofers from 60Hz or so to provide low end bass and control room nodes?
Any thoughts?
Thanks
Norris
Last edited:
Norris, the wrap-around notch pretty much disappears in the reverberant energy, so I don't believe in trying to "fix" it by changing the speaker's frequency response. Sort of like I don't believe in baffle-step compensation in most cases (I don't use it in any of my monopolars), as it solves one problem but introduces another that I think is worse.
My understanding is that, in their first-generation bipolars, Mirage notched the rear-firing woofer's output at the wrap-around frequency. This avoided the cancellation notch, but introduced a dip in the power response. Mirage stopped doing that in their second-generation bipoles, and just relied on their wide, shallow enclosure geometry (which is what I do).
I think that Genesis takes a very interesting approach with their four-way Model 5 series bipolars: The wrap-around notch region is covered by a pair of midwoofers, and the rear-facing one is wired in reverse polarity. This way, ithe rear midwoofer's output is reinforcing rather than cancelling at normal listening positions at the wrap-around frequency, but then there is cancellation to the sides.
Definitive Technology has patented their approach, which is to use a single side-firing woofer and confine the bipolar arrays to frequencies above the wrap-around.
Now there's no reason one cannot do a three-way bipolar; my reason for going with a controlled-pattern two-way is that's what I know and do best.
My understanding is that, in their first-generation bipolars, Mirage notched the rear-firing woofer's output at the wrap-around frequency. This avoided the cancellation notch, but introduced a dip in the power response. Mirage stopped doing that in their second-generation bipoles, and just relied on their wide, shallow enclosure geometry (which is what I do).
I think that Genesis takes a very interesting approach with their four-way Model 5 series bipolars: The wrap-around notch region is covered by a pair of midwoofers, and the rear-facing one is wired in reverse polarity. This way, ithe rear midwoofer's output is reinforcing rather than cancelling at normal listening positions at the wrap-around frequency, but then there is cancellation to the sides.
Definitive Technology has patented their approach, which is to use a single side-firing woofer and confine the bipolar arrays to frequencies above the wrap-around.
Now there's no reason one cannot do a three-way bipolar; my reason for going with a controlled-pattern two-way is that's what I know and do best.
Thanks Duke,
Just as I thought something as simple as adding a crossover in the right frequency range to control the wrap around difficiencies would be a fix, life gets complicated.
I guess simple is never truely simple. If it looks too got to be true, it is.
Norris
Just as I thought something as simple as adding a crossover in the right frequency range to control the wrap around difficiencies would be a fix, life gets complicated.
I guess simple is never truely simple. If it looks too got to be true, it is.
Norris
I just saw this. If i didn't have too may projects already, I'd definitely give this a try.
Superb article, thanks Duke
Superb article, thanks Duke
A very interesting approach. When I came across this, an idea occured to me. Another thing to try would be to install the front firing speakers into a bass traps which acts like a soffit, flush mounted. Or simply flush mount. Now add a second version instead of the rear drivers, put them on the side wall pointing towards the listening position. Now delay this second "ambience speaker" digitally, rather than doing it via placement out from the walls. I may actually try this at some point, no idea how it would turn out, if it's a good idea or not.
Thanks for the article. I hope more people get interested in trying out bipolar designs due to their very real superiority over conventional monopoles and dipoles.
I think my next project will be to try a bipole Karlsson enclosure.
I think my next project will be to try a bipole Karlsson enclosure.
In what way are they superior to dipoles with appropriate eq and headroom? The dipole pattern seems to distribute modal behavior pretty well, and can exhibit pretty uniform directivity to low frequencies, impossible with anything else except ginormous horns. The Bipoles will always be compromised in terms of power response by the bass peaks to the sides.
Extrapolation from the no-baffle-step monopole case can be a decent starting point, but the crossover should still be "voiced" with all the speakers in the box they'll be used in. Also, you'll need to take into account the in-phase summing below the wrap-around frequency region which boosts the low end somewhat, relative to the semi-random-phase summing above the wrap-around frequency region.
Appropriate headroom and EQ for a dipole can be more expensive and/or complex to incorporate into a DIY project, relative to the greater cabinet construction burden of a bipolar. A controlled-pattern offset bipolar system incorporates some of the advantages of a good dipole (well energized, spectrally correct, late-arriving reverberant field; smoother room interaction in the bass region) with the bass efficiency and impact of a good box speaker. It's not the solution to all problems audio, but it is an interesting combination of tradeoffs that may be useful in some cases.
Just for the record, my inspiration for the controlled-pattern offset bipole format is the SoundLab A-1. A longtime hardcore SoundLab owner (continuously since the late 80's) told me that my bipolar came the closest to emulating the SoundLabs of anything he has heard. Better than SoundLabs? No, of course not. A possible alternative if cost, amplifier compatibility, and/or high SPL are major considerations? Maybe so. Most of my manufacturing cost goes into the enclosure construction, so eliminate that by DIYing and you could come out well ahead of the game.
Naturally, re: starting point and voicing, I'm just wondering if there are any additional issues I may be missing.
The in-phase summation is just the wraparound which makes bafflestep unnecessary, no? I'm sure in many rooms there's also some boundary gain at LF, but it doesn't sound like there's anything I'm overlooking.
Understood- but usually when we talk about advantages around here, practicality is not the issue 😎. Seems to me that from an in-room performance standpoint, a large and proper dipole has the edge over anything shy of a large and proper fully horn-loaded CD system.
Certainly that's not meant as a knock against the offset bipole, as you say, it's an interesting combination of tradeoffs. In my setup, the dining room is behind a speaker and the other can be designed to fire into the kitchen, so either a dipole or bipole is very desirable in that I can get good sound when NOT on the sofa.
That's my hope- the impetus for the build was that I have in my posession 4 3/4" thick, 5.75" radius (or so) semicircular columns of 75" length, paid only $80 for all 4 ($90 per normally) and thus have the perfect start of an extremely smooth wraparound .
I am assuming you don't mind my playing with your concept?
I am totally flattered that your Bad self thinks there's enough potential merit in my concept to put time into trying it.
Take a look at the simulated early-reflection in-room response shown in my article, in particular Graph 2 (monopole with no baffle step compensation) vs Graph 4 (offset bipole). Note not only the smoother in-room response, but also the relative increase in energy in the low bass region. If you can factor those trends into your initial design choices, your chance of success goes up.
Last edited:
In your article you're very clear on wanting the rear woofer at the floor height- I was planning to do an inverted assembly, with my taller speakers having space for the woofer higher up. Is there any reason that the floor is special, or is it just that with normal cabinet heights that's the best way to distribute the woofer locations in the vertical?
Rear woofer down near the floor helps to fill in the floor-bounce dip and reduce the octave-higher floor-bounce peak, and also reduces the depth of the wrap-around notch. However if you plan to listen from either side of the speakers, then imo that takes priority, and you might even go with a symmetrical bipole. Just try to make it considerably wider than it is deep.
In the October 2011 issue of Stereophile the Vivid Audio B1 is reviewed, and it's a symmetrical bipole south of 500 Hz or so (with the rear-facing woofer rolling off gently north of 200 Hz). You can clearly see the wrap-around notch in quasi-anechoic measurement, and you can clearly see it fade away in the in-room measurements, so in practice it's not a big deal.
Okie dokie. Width is fixed by the driver panel size and half-round width, and depth is fixed by the half-round at abt 11.5". The overall width will have to be between 25-29" depending upon whether I use JBL 123A (12") or a 15".
Thanks for revelation.
Without noticing your findings I have built symmetrical (without offset) bipole bass pseudo-horn aka Voigt pipe (W54cm, H110cm, D30cm, 6 degrees sloped front baffle, half-folded horn mouth opening internally towards the floor terminated through omni-slot made by elevation of the speaker, two vintage 12" 95dB/1W Qts ~0.35 drivers mounted at maximum possible height with magnets facing eash other - so called push-push setup).
The tweeters are connected as dipoles (rear polarity inverted) and have "naked" round 16cm waveguides without separate baffle. I did very simple in-room directivity measurements of the tweeters to see what happens if tweeter polarity is inverted or not. Turned out it makes significant difference with naked or narrow baffle (in this case waveguides): in dipole configuration on frequencies under 8kHz dispersion turns increasingly lateral, while in dipole (inverted) it keeps near perfect figure eight all the way down.
After I attenuated rear tweeter a couple of decibels with L-pad I found the whole speaker having spectacular balance between direct and diffuse sound despite small room size (2x3x2.5m). The jazz singers felt not just like being in the same room just in front of speakers but also enveloped with aura of much larger space than the room. Compared to other setups I have ever built this one have gave me very pleasant balance of both presence and envelopment qualities.
While reading your article about Controlled-pattern offset bipole I found all mentioned bonuses present also in my symmetrical bipole design - extended, box-less bass and less interaction with room. Yes, I have deep notch at 200Hz and broad elevation at 500Hz, probably both attributable to effects of wrap-around interference. But notch isn't noticeable even with downward sweeping bass notes, perhaps coincidentally room modes come in to compensate them. Elevation around 500 Hz somehow sounds just right as this range often have spectral dip (at least for my taste).
What still is problematic is sensitivity of tweeters which is too low against woofers especially after waveguide compensation passive EQing. Also there is a dip at 1kHz from HF driver roll-off that steals some attack from midrange. So I'll try to go with pro tweeters and controlled directivity waveguide.
I hope you find this information useful. If you are in Riga in ocassion, you're welcome to have an ear on the system.
Last edited:
Hi Duke,
Why should the box be shallow and wide? The reason why I ask is that delay can be achieved by both shallow/wide box as well as deep/slim box?
Thanks,
Goldy
The wide, shallow format is intended to reduce the magnitude of the wrap-around cancellation notch, and it does so by two mechanisms:
First, the width of the cabinet pushes the frequency at which wrap-around begins lower, just like a wide cabinet pushes the baffle-step frequency lower.
Second, the shallower the cabinet, the shorter the wrap-around path length, and therefore the higher the frequency at which the notch occurs. Remember this notch is centered on the frequency where the wrap-around rear radiation arrives 1/2 wavelength behind the front radiation. The higher the frequency, the more directional the woofer (or more precisely woofer + baffle) will be, so the less energy will actually wrap around to cancel out the front woofer.
You can see this illustrated in my HiFi Zine article. Look at Graph 3 (narrow & deep) and Graph 4 (wide & shallow):
The Controlled-Pattern Offset Bipole Loudspeaker
The wrap-around notch is at 200 Hz for the narrow/deep box (Graph 3), and at 250 Hz for the wide/shallow box (Graph 4). Compare the width and depth of those two notches, and while you're at it, eyeball the rest of each curve as well.
Thanks Duke for the latest information.
I hope to eventually get an opportunity to listen to your latest creation. Hopefully we will make it to RMAF this October. We missed you this past weekend in Dallas at LASF.
I hope all is happy and well for you and yours.
Norris
I hope to eventually get an opportunity to listen to your latest creation. Hopefully we will make it to RMAF this October. We missed you this past weekend in Dallas at LASF.
I hope all is happy and well for you and yours.
Norris
- Status
- Not open for further replies.