Open Baffle (Dipole) Living Room Line Array Experiment aka OBLIROLA
tmHi, i'm Yuri, from Amsterdam.
1.
The one thing i'm MOST interested in, is the hurdle today's speakerbuilding mostly hasn't taken yet.
The Constant Directivity (CD), Even Power or Polar Response (EPR), Uniform Radiation or Dispersion Pattern, there's a couple more terms, i dunno if EPR is even an acronym. They all boil down to the same:
It is the thing that makes a speaker either sound like a box, OOOORRR, as the final frontier of Constant Directivity is ever more closely approached, like live music you can't find the source of, even when looking straight in the direction you
hear the sound coming from.
Good speakers CAN be like that, but i've never heard a BOXED speaker do it (anywhere even close to) flawlessly.
2.
Regardless, the loudspeaker, any loudspeaker, is the only part of the signal chain still not up to par.
#Great microphones and recordings exist.
#Great listening rooms exist (and are not even needed by some speaker designs!).
#Great affordable gear exists, that most professional listeners in an ABX testing process would not be able to distinguish between, including everything from source devices, to DSPs, DACs, preamps and poweramps.
Think the Bob Carver Challenge.
If you're able to distinguish between electronics devices, either the distinct one or the whole bunch of DUT's is flawed, in way of general best practices.
#All well built gear basically sounds the same (with very few exceptions, on certain very specific issues).
Yet no to two speaker designs sound the same. Ever.Because unlike all the other gear in the signal chain, for loudspeakers, a definitive solution is yet to be found.
Distortion figures are way behind the dot on litterally ALL well designed gear without a single exception, until it comes to loudspeakers.
In those, measured distortion figures between 1% and 5% are very common (among the best built specimens) and figures between 10% and 30% are no exception, even among basically decent products by respected manufacturers.
And, the (professional) subjective listening tests generally agree.
(Fun fact: Loudspeakers that measure the best, are almost exclusively also subjectively judged to be the best in double blind A/B/X listening tests with control group.)
3.
So that's why i became interested in the website of Siegfried Linkwitz (RIP since 2018), an engineer who (together w/Russ Riley RIP a little before) created the presently most used crossover circuit LR4 (=24dB down @1octave, along with LR1, 2, 3, 8, etc. 6/12/18/48dB/etc).
(This is off topic, but the LR4 is a phase linear XO w/very special qualities, and as trade offs go, you must have real good reasons to not implement it, as explained in this part of Rane Note 107:
https://www.ranecommercial.com/legacy/note160.html).
After spelling out every single word of his open baffle dipole R&D, i became more and more interested in this unique approach and it's advantages. More OB giants on whose shoulders we stand include John Kreskovsky, Rudolf Finke, et.al.
Open Baffle Dipoles, depending on frequency, approach a 'figure-of-eight' radiation pattern up to high midrange, with a spherical lobe in front of and behind the speaker, separated by a toot-shaped cancellation null in the 360° vertical plane surrounding the driver at a 90° angle. This shapes a pretty close to Constant Directivity soundfield, especially when well implemented and when also navigating the relevant additional issues involved w/OB.
4.
Quite a few professionals have been hunting the Constant Directivity principle, according to a plethora of academic white papers and published AES lectures, and most studio monitor designers apply various types of wave guides and other measures to even out directivity, especially at XO-range (although the ideal would be at ANY range).
Digging through this subject, i became suspecting that dipoleness of many electrostatic and similar planar drivers of the past (and still today) is the secret to their succes, even when measuring poorly and lacking in other departments.
5.
Also interested in Line Arrays and what they (can) do in large, or even small living rooms.
And seeing some problems in both OB and LA that a combination might partially navigate, next to a few advantages that may also combine favorably.
Open baffle designs somewhat improve on the omnidirective nature of conventional LF reproduction, as mentioned above. Although all reflections will become instantly omnidirectional to my knowledge, the direct sound from the source becomes somewhat directional. This is a step closer to Even Power Response. (For now leaving alone everything that might supposedly be wrong w/OB's.)
Line arrays seem to also improve the Constant Directivity goal, again within a certain frequency range and depending on navigation of all caveats. Like OBs they also diminish room interaction and improve both direct-to-reflected-energy and late-to-early-reflection ratios.
The ideally desired CD could be more closely approached by a wavefield resembling a double cylindrical figure of eight instead of a double spherical figure of eight, predominantly moving in the horizontal plane at yet a wider frequency range. Theoretically so far.
It would interest me to find out if the evenness of the power response (the radiation pattern!) over frequency, will allow us to easier flat-tune further above some 2-5kHz (where beam-shaped wavefields normally start to become an issue) in the amplitude and/or frequency domains. (For now leaving alone everything that might supposedly be wrong w/LA's.)
6.
So that's how i landed at OBLIROLA's (my trademark for Open Baffle Living Room Line Arrays).
It would seem to me that OB config natively supports the cylindric nature of the Line Array theoretical soundfield. Yet the 2Pi nature of the cylindrical waveform should help the dispersed HF to carry a little further than with OB only, where it would normally die out sooner than LF.
Paramount for these assumptions seems to be the 'trueness' of the line source, and the baffle width of HF transducers: with a multitude of small drivers, together forming the source, spacing of the drivers will become a problem at frequencies smaller than the distance between them. Therefore i seek to execute the experiment with true and very narrow line sources: tweeters that are shaped like a line!
The (costly) Mundorf AMT tweeters for instance, present a source approx 30mm wide by 240mm high, with 20-30mm of spacing between each 240mm unit. So approx. 260cm high, it would present just nine 25mm spaces. The not so costly and well measuring GRS PT6816 may present a narrower option, especially on the (open) back side, which i would the use as the front (because directivity in the rear is less critical, especially in the HF range, where diffraction is abound as soon as ANY boundary is hit).
But even then, as Merlijn van Veen aptly reminds us, the cylindrical shape over the whole frequency range will remain at least in part theoretical (
https://www.merlijnvanveen.nl/en/study-hall/193-cylindrical-waves-hyped-or-not).
Nevertheless, in between near and far field, where i personally can often be found hovering, some advantages may be attained, and some more may be in sight with proper weighing of certain trade-offs. Most theory/white papers/etc. on LA's sofar (Toole, Keele, Griffin, Smith, Button, Ureda, e.a.) has been based on large listening venues. I will be looking at smallish, to medium, to rather large living room situations.
Aiming to reproduce the recording as it sounded in the recording venue or studio, without my specific room adding it's own spacial character, UNWANTED on top of the venue's
already characteristic spacial signature?
Having developed toward this over a long period of researching whatever i could find on the topic, a combined perspective from physics and gut feeling tells me i must check this out. In search of the loudspeaker that, although you see it at work, you won't believe the sound is coming from, SO tangible an illusion it creates of a live soundfield.
So gonna experiment somewhat w/that. Love to spar with anyone who is informed (at least somewhat) on the subject. More better. Cheers!
