Scott, you are correct... 6+12=18, but that is not an accurate picture of what is going on at the low end of the passband of a real world OB subwoofer. The driver is a typically low Q, low Fs (sub)woofer with high excursion. The dipole cancellation is indeed 6dB/octave, however, the driver rolloff has not yet reached the ultimate rolloff of 12dB/oct, which will only happen well below the resonance frequency. In a low Q response, around resonance the slope is slowly increasing but will be about 6dB/oct at Fs. This is why I noted that the measured rolloff at 20-30Hz in my H-frame dipole was 12dB/oct, because that what it actually is, even down to 10Hz. I used a subwoofer with Fs~20Hz.
Also, this brings up a point I forgot to mention earlier about why dipole bass can work well: The resonant frequency of the system is not being "raised" by the compliance of the air in the box that is containing the back wave. In boxed loudspeaker systems this causes Fs and Qts to increase compared to the free-air response. It's the same (or similar) effect for closed box or bass reflex systems. But in a dipole, especially open-baffle but also U- and H-frames, the driver is largely acting as if it is in free air and there are only relatively minor changes to Q and Fs. This results in the driver's resonance frequency being lower than boxed systems, which also helps to keep distortion down, and the low Q results in the less-than-ultimate low end rolloff slope that I mentioned above.
As another poster mentioned, when Q is low and the rolloff not steep, the transient response is without overshoot or ringing and group delay low. These are some clear advantages of OB/dipole systems compared to boxed systems in the time domain.
Actually, there are a number of manufacturers building OB/dipole systems, so I think the trend is an increasing one. Off the top of my head, apart from the LX521 there is Emerald Physics, Pure Audio Project, Jamo, Kyron Audio, dAudio, Manzanita Audio, and probably others I did not find with a 5 min Google search.
Same as the sides, cavity resonances. As the simple flat baffle becomes more and more like a box, so it starts to sound more and more like a box. That's not all bad, of course.
Using four drivers creates a longer front-to-back pathlength, so together they only work up to about 1.2kHz (judging by the 2D heatmap) before there are some off axis holes forming and the front lobe is narrowing. The quartet is starting to beam. But I think Gerrit was looking at how distortion performance was influenced, not trying to make a better dipole radiator:
Please correct me if I misunderstood his intention.
Yes, loudspeaker building involves lots of trade offs.
Maybe someday I will get to hear your system, and someday I can have you listen to the one I am building, too.
Yes, that's his stated intention, not sure I understand. Perhaps it would be interesting to use just two drivers, for symmetry, it seems a simple solution, if not perfect?
Scott, download The Edge (for Windows), it is a relly nice little simmer, just fine to check baffle diameter effects with dipoles. basically H/U-frame can be spread to one plane, just measure distance from driver midpoint to edge around the bend.
U/H frame forms a cavity, but usually that is quite short and resonance is above it's passband. If a U-rame is used up to mid freq, there might be a problem. But every one that I have seen are laying on the floor and are obviously used only as sub.
By the way have you noticed that SL took mostly just nearfield mesasurements of his bass modules, and calculated the farfield response, without taking floor and walls effect in calculations. Thor sub.
I think Gerrit used reversed drivers first of all to get better front-back symmetry. Compared to a single large driver it measured much better above dipole peak.
U/H frame forms a cavity, but usually that is quite short and resonance is above it's passband. If a U-rame is used up to mid freq, there might be a problem. But every one that I have seen are laying on the floor and are obviously used only as sub.
By the way have you noticed that SL took mostly just nearfield mesasurements of his bass modules, and calculated the farfield response, without taking floor and walls effect in calculations. Thor sub.
I think Gerrit used reversed drivers first of all to get better front-back symmetry. Compared to a single large driver it measured much better above dipole peak.
It just gives the sound another parallel surface to reflect off. It's that bouncing back and forth that is the resonance. Just like a room. With no cavity, no cavity resonance.
Yes, embrace randomness, irregularity, and no character of its own for a speaker to reproduce music (maybe the opposite for a music instrument).
This discussion seems to diss'ing the The Big Five Box approach by endorsing all the in-between variants.
In the late 60's I used Bozak speakers which had lots of stuffing but with the backs off. H, U, or Baffle, all can work. Aperiodic BR and well-stuffed TLs and pipes are likewise in-between efforts to sequester or otherwise manage the rear wave.
While The Big Five kinds of boxes may hold some appeal for those who like straight-line thinking (and the comfort of sim's), there is no reason for rest of us to be trapped there. Seems better to craft a good speaker and then fine-tune acoustically by ear and mic with the great power of DSPs.
B.
Last edited:
Just my way of saying it is not productive to think in terms of the familiar (and readily sim'ed) box solutions (like BR, sealed, TL, horn, mis-named tapped-horn, etc.*). Better instead to think-through the problem you think you are trying to solve (such as rear-waves in a woofer in relation to your room and your sound priorities).
Long pipe to sequester rear wave
B.
* DIY people should also recognize that some solutions are familiar because they are natural for manufacturers to build and ship. Motional feedback, large OBs, or bi-amp'ing are not as natural for commercialization.
Long pipe to sequester rear wave
B.
* DIY people should also recognize that some solutions are familiar because they are natural for manufacturers to build and ship. Motional feedback, large OBs, or bi-amp'ing are not as natural for commercialization.
Last edited:
Just wanted to add some authority to my comment. Here's a quote from SL that I've been seeing for years but didn't understand until today (emphasis added):
"What you hear is not the air pressure variation in itself but what has drawn your attention in the streams of superimposed air pressure variations
at your eardrums"
In other words, Linkwitz is saying that the acoustic stream ("superimposed air pressure variations") can not be interpreted in the absence of knowing something about consciousness. That is not to deny that the place of physical measurement* only to say it is far from the whole story.
So in comparing speakers by physical measures. Even looking at Toole's great work, we can ask, "Were people choosing speakers that sounded like "great speakers" or sounded like "great orchestras"?
B.
* nobody could be ranting more than me in this forum more about "show us the measurements"
Ben, you have got to be the only member who regularly uses footnotes in your forum posts! [1] How academic of you 
I have to agree that you are on to something in the above message... yes indeed, the ear-brain hearing mechanism has many tricks up its sleeve. A microphone measurement is not a representation of "hearing". Wringing of hands over how the speaker will interact with walls must be viewed within the context of how the brain will respond to the multiple "copies" of sound it receives in a reverberant space. Known effects like the precedence effect [2] show that the brain is very a sophisticated processor of stimuli from the ears, perhaps a necessary evolution for making sense out of the auditory world around us.
Dipole loudspeakers can in a sense be viewed as a generator of two copies of the audio input, one to the front, and one to the rear. It has been said [3] that when the brain detects a "copy" of a sound heard recently (within the past few msec) and the spectral balance of the "copy" is like the original sound, they are considered to be from the same source and that this gives some extra "space" to the perceived sound. This is one reason why I am trying for the same front and back radiation in my current dipole loudspeaker projects. [4]
[1] I prefer end notes.
[2] See: Precedence effect - Wikipedia
[3] possibly by SL within his vast website.
[4] As outlined in "In Pursuit of a 20-20k Dipole Loudspeaker"
I have to agree that you are on to something in the above message... yes indeed, the ear-brain hearing mechanism has many tricks up its sleeve. A microphone measurement is not a representation of "hearing". Wringing of hands over how the speaker will interact with walls must be viewed within the context of how the brain will respond to the multiple "copies" of sound it receives in a reverberant space. Known effects like the precedence effect [2] show that the brain is very a sophisticated processor of stimuli from the ears, perhaps a necessary evolution for making sense out of the auditory world around us.
Dipole loudspeakers can in a sense be viewed as a generator of two copies of the audio input, one to the front, and one to the rear. It has been said [3] that when the brain detects a "copy" of a sound heard recently (within the past few msec) and the spectral balance of the "copy" is like the original sound, they are considered to be from the same source and that this gives some extra "space" to the perceived sound. This is one reason why I am trying for the same front and back radiation in my current dipole loudspeaker projects. [4]
[1] I prefer end notes.
[2] See: Precedence effect - Wikipedia
[3] possibly by SL within his vast website.
[4] As outlined in "In Pursuit of a 20-20k Dipole Loudspeaker"
Last edited:
+1... couldn't say it better.
You mention precedence, and there is timing, loudness, head shadow, pinna effect, ear canal resonance, Haas, Fletcher-Munson, and more are known about.
But the real challenge are phenomena like "learning" a room so as to hear things right in that room. Whatever algorithms your brain is processing to hear optimally in a room, it is applying related processing in discerning stereo images from two speakers. In turn, do those algorithms find it easier to "make sense"* of the stereo tracks when coming from dipoles or point-source boxes?
B.
* what Helmholtz first identified as "unconconscious inference"
You mention precedence, and there is timing, loudness, head shadow, pinna effect, ear canal resonance, Haas, Fletcher-Munson, and more are known about.
But the real challenge are phenomena like "learning" a room so as to hear things right in that room. Whatever algorithms your brain is processing to hear optimally in a room, it is applying related processing in discerning stereo images from two speakers. In turn, do those algorithms find it easier to "make sense"* of the stereo tracks when coming from dipoles or point-source boxes?
B.
* what Helmholtz first identified as "unconconscious inference"
Time for a couple more anecdotal points - comments. After 40 plus years of fooling around mainly with traditional box woofer alignments, I kept going back to OB - Cardioid constructs. What I realized is that the OB seemed to perform well in most spaces as long as it was properly designed AND you could get the rear of it out (away) from a rear wall by at least 36". 48" or even more is better.
It was obvious that the OB's were far less interactive with the space they were working in. WAY less than ported and somewhat less than sealed. Be it a small or large hotel room, various residential environments, out doors, where ever. The OB would be amazingly consistent. Way more than any traditional closed box system. And I have brought along many closed box creations to RMAF (twice) LSAF 12 years running (I think?) and so forth. The boxes change with the space. The OB's don't much. Small OB or Monster OB. With the boxes, the bigger they are, the bigger the change usually is.
Charlie... your goal of equal front and rear output is a great one, but I fear, if you get close to it, you will not prefer it in all but the largest closed spaces. It is about time vs frequency. Will not elaborate here, but know if you think about it, you will come to the same conclusion. As soon as you add distance and a reflective surface, well, your rear output (radiation) no longer equals your forward radiation from the typical listening position. So many reasons why this will be the case with dynamic drivers in an OB application. My $0.02 worth!
It was obvious that the OB's were far less interactive with the space they were working in. WAY less than ported and somewhat less than sealed. Be it a small or large hotel room, various residential environments, out doors, where ever. The OB would be amazingly consistent. Way more than any traditional closed box system. And I have brought along many closed box creations to RMAF (twice) LSAF 12 years running (I think?) and so forth. The boxes change with the space. The OB's don't much. Small OB or Monster OB. With the boxes, the bigger they are, the bigger the change usually is.
Charlie... your goal of equal front and rear output is a great one, but I fear, if you get close to it, you will not prefer it in all but the largest closed spaces. It is about time vs frequency. Will not elaborate here, but know if you think about it, you will come to the same conclusion. As soon as you add distance and a reflective surface, well, your rear output (radiation) no longer equals your forward radiation from the typical listening position. So many reasons why this will be the case with dynamic drivers in an OB application. My $0.02 worth!
When looking at this it is:
Listener > Direct Sound.
..most of the reflected sound is effectively "processed-out" except as it pertains pressure levels near the Direct Sound average or highly impulsive sounds (like a "click") at higher freq.s a bit lower than the Direct Sound average (and of course when higher than that in sound pressure as well).
The whole thread on the "Preference for Direct Radiators" seems to be looking at it ***-backwards (source > reflections > and (if rarely discussed) listener). But that's what usually happens.
Listener > Direct Sound.
..most of the reflected sound is effectively "processed-out" except as it pertains pressure levels near the Direct Sound average or highly impulsive sounds (like a "click") at higher freq.s a bit lower than the Direct Sound average (and of course when higher than that in sound pressure as well).
The whole thread on the "Preference for Direct Radiators" seems to be looking at it ***-backwards (source > reflections > and (if rarely discussed) listener). But that's what usually happens.
Last edited:
Positioning of dipoles must be different than normal monopoles Linkwitz tells about this a lot with good links and references. First reflections are more or less in nulling-area of horizontal radiation.
I prefer placement on the long wall, and generous toe-in. This way also the back-wavefront gets maximal delay. The resultant sound is extremely life-like and 3-dimensional. This is good for classical. Problem is that most hifi-enthusiasts don't go to concerts and are not used to hearing "living voice and sound". They prefer what they are accustomed to at home - monopole speakers with increasing directivity towards highs.
I have had dipoles and monopoles side-by-side for a few weeks now and the difference is obvious. For certain kind of studio-trickery-effect "music" monopoles do sound better.
I prefer placement on the long wall, and generous toe-in. This way also the back-wavefront gets maximal delay. The resultant sound is extremely life-like and 3-dimensional. This is good for classical. Problem is that most hifi-enthusiasts don't go to concerts and are not used to hearing "living voice and sound". They prefer what they are accustomed to at home - monopole speakers with increasing directivity towards highs.
I have had dipoles and monopoles side-by-side for a few weeks now and the difference is obvious. For certain kind of studio-trickery-effect "music" monopoles do sound better.
Hi all,
a lot has been discussed about dipole bass, but not much about dipole treble yet. Looking at the speakers by John Busch, they seem to have a wide baffle and, therefore, not a "nice " dipole dispersion at the top end. Is that right and what are your thought on dipole treble?
Best, Erik
a lot has been discussed about dipole bass, but not much about dipole treble yet. Looking at the speakers by John Busch, they seem to have a wide baffle and, therefore, not a "nice " dipole dispersion at the top end. Is that right and what are your thought on dipole treble?
Best, Erik
This my opinion and experience as well. More general, once a speaker achieves that its off-axis sound at any angle always is a precise copy of the on-axis response (complex, mag&phase), except for level, then such speaker has the best chances to dissapear in the room response signature, it sounds like a real source playing at that location as the room coloration is natural. With a nice stereo signal this gives the recorded ambience an additional but very natural reverberant space to "breathe in" and this sounds extremely well and life-like in most circumstances (unless the room is too small).
The choosen main pattern (omni, cardioid, dipole, beaming sector, line-source, waveguides etc) then affects the perceived amount of "liveliness" (the color of it is correct already), of course. How much of it is considered "just right" is a personal preference and also dependent of source genre and playback levels. Too live a room signature (as exited by the speaker) may not be ideal to project speed metal at concert volumes but will shine on a small string quartet, jazz combo or "girl-with-guitar" type of music giving us a near-perfect "they are here" experience.
The dipole construction now is one of the easiest ways to get a room signature that is a) quite a bit "dryer" than with true omni and b) very neutral, helping to create a convincing auditory scene with a nice balance of direct vs room sound. And this room sound needs less acoustic treatment than with omni let alone speakers with skewed off-axis sound.