Lobing =/= Combing? How it physically happens, what's the difference in concretic terms?
Yes of course, for the pictured tetrahedral speakers, the benefit of the shape would be one less corner for the baffle compared to rectangular, wouldn't this mean less equal distances from a centered driver and thus less pronounced baffle gain for the frequencies whose wavelengths correspond to those equal lengths?
And no cabinet reflections frontwards from the "side walls" as they're really back walls.
Combing czauses the lobing. Either between 2 or more drivers in an array or a tweeer to woofer.
If the centre-to-centre of the soiurces is a quarterwave or less there is no combing. One of th ereasons i like WAWs.
dave
Last edited:
I've seen similar speakers online, but never had a chance to listen to one.
Think anyone who could afford some took measurements in-room of the Burmester bending wave omnipolar speakers?
I built a new pair of full range Metronomes (quarter wave quadratic transmission lines) using Fostex FE108EΣ drivers and 18mm Baltic birch; I have to wait for warmer weather to finish two recent builds. I intend to use them as surrounds in my music theater but I tested them up front alone too. I like this driver and have used it in two other pairs of "Onkens" in the system. This system is now all on EΣs. The Mets perform well and don't seem to display as much baffle diffraction step "loss" as I'd expected... even in free space. There's "always" a 1KHz dip with these drivers. The full build process is described HERE if there's interest.
Stand up, sit down. Move side to side, forward & backward will not likely be an issue until you get close.
A lobe starts to develop when the difference in distance from the ear to each of the drivers starts to exceed a quarter-wavelength.
Includes the math and more detail: https://audiojudgement.com/speaker-lobing-polar-response/
In a 2-way speaker this is the general idea. (in theoretical plane on axis.
With 2 FR speakers running FR you don’t have an XO keeping the HF output of one interefering with the other. Below a quarter-wavelength they act as a single unit, as the disatnce increases the combing gets worse (as frequencies go up and and wavelength get shorter).
Here is a conbing diagram (from Taylor) with 4 drivers. As frequencies go down they disappear, as frequecy increases they get worse.
dave
Well, "sounded better" is about as subjective as it gets. Some people get outraged if I mention making speakers for my own use, and tailoring their output to sound good to me. One asked how I could tell it sounded good, and was genuinely outraged enough to repeat back to me- by listening to them- and added several question and exclamation marks .
He was one those whose computer designs a passive crossover, and posts the predicted response graph with zero deviation from a perfectly flat frequency response, and labels it as an in room 1 metre measurement.
Back to subjective, I've seen posts of 4 12" full range drivers in short line-arrays fully enjoyed by their owners.
I would consider all the phantom sound sources- wall, ceiling and floor reflections- introducing lobing and comb filtering in general, especially in a typical multi use room that's not treated for audiophile play.
This might even be a part of why some people don't mind the sound of line arrays at all.
There's comb filtering introduced by the design, yet while (I assume) it maintains the usual sidewall reflections from the common wide dispersion designs, floor a d ceiling reflections are limited by the line array.
I've never done a comparison of a set of measurements for an omnipolar speaker pair from a usual 1 meter on axis response to another set of measurements from what for me is a more usual 4 metre distance, not on axis with where a tweeter would be, and so getting plenty of room reflections. If it's like most speakers, any design will do both good and not so good things.
To my way of thinking, any multiway will introduce some degree of peaks and valleys in a frequency response, simply because the drivers are at different distances from their listener, excepting at that one on axis location usually used for measurement taking.
He was one those whose computer designs a passive crossover, and posts the predicted response graph with zero deviation from a perfectly flat frequency response, and labels it as an in room 1 metre measurement.
Back to subjective, I've seen posts of 4 12" full range drivers in short line-arrays fully enjoyed by their owners.
I would consider all the phantom sound sources- wall, ceiling and floor reflections- introducing lobing and comb filtering in general, especially in a typical multi use room that's not treated for audiophile play.
This might even be a part of why some people don't mind the sound of line arrays at all.
There's comb filtering introduced by the design, yet while (I assume) it maintains the usual sidewall reflections from the common wide dispersion designs, floor a d ceiling reflections are limited by the line array.
I've never done a comparison of a set of measurements for an omnipolar speaker pair from a usual 1 meter on axis response to another set of measurements from what for me is a more usual 4 metre distance, not on axis with where a tweeter would be, and so getting plenty of room reflections. If it's like most speakers, any design will do both good and not so good things.
To my way of thinking, any multiway will introduce some degree of peaks and valleys in a frequency response, simply because the drivers are at different distances from their listener, excepting at that one on axis location usually used for measurement taking.