I agree with what Dr Gedlee's last post completely.
When comparing my OB speakers to my friend's Thiel speakers, The Thiel sounds smoother (they were the smoothest among all Thiel models). But that is almost the only area that the Thiel sounds better than my OB. On the other hand, no matter how the thiel sounds, it sounds like good music produced in a box, while my OB disappears, as if "musicians are right here" vs "musicians are there". There is a big difference. Perhaps this is due to "no box" sound. It is not just related to radiation patterns, as both speakers are placed in fairly damped, treated rooms.
But I suspect the less smoothness of my OB is due to greater baffle edge diffraction and the back waves wrapping around the speakers joining the front waves creating peaks and dips (ballooning). Move your head a few inches to notice the change of sound (in a very well damped room). This is the area that OB won't be as good as boxed speakers. Careful design will reduce ballooning (by using narrow baffles and driver directivity at high frequencies), but not eliminate them, and no matter how careful, diffraction seems worse than boxed speakers, as Dr Gedless described. My brief measurements made me believe in this.
Everything seems to have some compromises and nothing is perfect. There are always pros and cons.
To have the best of both (OB vs Boxed) worlds, I am currently working on a new pair of large U-frame speakers. The idea is to have the same air impedance in front of and behind the drivers, so this is the same as OB (flat baffle) with reduced "boxed sound". But the back waves are heavily damped (above 300Hz) so the ballooning is basically eliminated (reflected in my measurements). The U-frame allows a large front edge round over (70mm radius) to reduce diffraction. Damping the back waves also produce far more even total power response across the audio frequency spectrum, comparing to boxed speakers and OB without rear tweeters.
Is that perfect? No. In order to have 115dB SPL in OB I use MTM so the vertical lobbing is clearly evident. My measurements show due to floor bounce the response is rough around 1kHz. That could be the only large limiting factor of my speakers. I have been thinking about waveguides, as they may provide better directivity at 1kHz to reduce the floor/ceiling bounce effect. But how much can it reduce? Would it introduce a new set of problems I need to deal with?
Enjoy music and DIY.
Regards,
Bill
When comparing my OB speakers to my friend's Thiel speakers, The Thiel sounds smoother (they were the smoothest among all Thiel models). But that is almost the only area that the Thiel sounds better than my OB. On the other hand, no matter how the thiel sounds, it sounds like good music produced in a box, while my OB disappears, as if "musicians are right here" vs "musicians are there". There is a big difference. Perhaps this is due to "no box" sound. It is not just related to radiation patterns, as both speakers are placed in fairly damped, treated rooms.
But I suspect the less smoothness of my OB is due to greater baffle edge diffraction and the back waves wrapping around the speakers joining the front waves creating peaks and dips (ballooning). Move your head a few inches to notice the change of sound (in a very well damped room). This is the area that OB won't be as good as boxed speakers. Careful design will reduce ballooning (by using narrow baffles and driver directivity at high frequencies), but not eliminate them, and no matter how careful, diffraction seems worse than boxed speakers, as Dr Gedless described. My brief measurements made me believe in this.
Everything seems to have some compromises and nothing is perfect. There are always pros and cons.
To have the best of both (OB vs Boxed) worlds, I am currently working on a new pair of large U-frame speakers. The idea is to have the same air impedance in front of and behind the drivers, so this is the same as OB (flat baffle) with reduced "boxed sound". But the back waves are heavily damped (above 300Hz) so the ballooning is basically eliminated (reflected in my measurements). The U-frame allows a large front edge round over (70mm radius) to reduce diffraction. Damping the back waves also produce far more even total power response across the audio frequency spectrum, comparing to boxed speakers and OB without rear tweeters.
Is that perfect? No. In order to have 115dB SPL in OB I use MTM so the vertical lobbing is clearly evident. My measurements show due to floor bounce the response is rough around 1kHz. That could be the only large limiting factor of my speakers. I have been thinking about waveguides, as they may provide better directivity at 1kHz to reduce the floor/ceiling bounce effect. But how much can it reduce? Would it introduce a new set of problems I need to deal with?
Enjoy music and DIY.
Regards,
Bill
The Orions are OB (as opposed to dipole) only between 1.4 kHz and about 40 Hz.
I have made many OB speakers, and for some time, every time I make a new design I try it with and without a rear facing dipole tweeter, unpadded and padded to different levels.
In all cases, my wife "just loves music", and my brother "an audiophile", and I, prefer the sound without the dipole tweeter.
Are we perhaps finding dipole diffraction above, say, 1 kHz objectionable, a correlation perhaps with the baffle size?
Quad ESLs (which I used to own) used a layer of felt over the rear surface, which effectively substantially, and progressively, reduced back radiation above about... 1 kHz.
Is the inherent diffraction caused by a dipole design less ojectionable (unnoticable?) the lower the frequency, due to the wavelengths being substantially larger than the size of the baffle??
Maybe taking the concept of the Summa as a comparison, an OB speaker that crossed over to a waveguide in the 800-900 Hz region, or probably less, and gradually attenuated with increasing frequency at the rear with appropriate "curtain" absorption, might achieve the subjective OB notion of "boxlessness" without the negative effects of diffraction???
(I don't believe panel colourations with boxes are the issue, as it requires a great effort to reduce baffle vibrations down to near the level of a well damped box).
(My wife, my brother, and I, all prefer the effect of OB, and my brother maintains that he can detect "boxiness" when we have done a line-up of a number of OB and boxed speakers. Not at all rigorous I know).
David
I have made many OB speakers, and for some time, every time I make a new design I try it with and without a rear facing dipole tweeter, unpadded and padded to different levels.
In all cases, my wife "just loves music", and my brother "an audiophile", and I, prefer the sound without the dipole tweeter.
Are we perhaps finding dipole diffraction above, say, 1 kHz objectionable, a correlation perhaps with the baffle size?
Quad ESLs (which I used to own) used a layer of felt over the rear surface, which effectively substantially, and progressively, reduced back radiation above about... 1 kHz.
Is the inherent diffraction caused by a dipole design less ojectionable (unnoticable?) the lower the frequency, due to the wavelengths being substantially larger than the size of the baffle??
Maybe taking the concept of the Summa as a comparison, an OB speaker that crossed over to a waveguide in the 800-900 Hz region, or probably less, and gradually attenuated with increasing frequency at the rear with appropriate "curtain" absorption, might achieve the subjective OB notion of "boxlessness" without the negative effects of diffraction???
(I don't believe panel colourations with boxes are the issue, as it requires a great effort to reduce baffle vibrations down to near the level of a well damped box).
(My wife, my brother, and I, all prefer the effect of OB, and my brother maintains that he can detect "boxiness" when we have done a line-up of a number of OB and boxed speakers. Not at all rigorous I know).
David
I don't think a rear tweeter in OB works well in a small room. However, I do believe it works well in a large room. The rear tweeter is important because it gives more even power response. However, in a small room, since rear tweeter is better not to be used, the power response is weaker at high frequencies, which introduces a type of colouration that is quite audible, especially in a live (vs damped) room. That is yet another compromise one has to make with OB.
Boxed speakers are not better. For flat frequency response at listener's position, the power response can be 6dB higher below the baffle step. This again produces fairly uneven power response. And perhaps this gives it a "boxed speakers" sound?
That is why I am now trying the damped U-frame OB without a rear tweeter. Above baffle step, the front waves radiate into 2Pi space. There is no rear waves from the tweeter, while the rear waves from the midwoofers are heavily damped to be insignificant. Below baffle step (300-600Hz), dampping materials become less effective (due to limited thickness comparing to the wave lengths) so the radiation gradually changes to cardiod then dipole. Not only does it provide directivity but also reduces the power response comparing to sealed box. Below about 150Hz-200Hz, a sealed woofer box takes over. Due to the floor boundary, it again radiates into 2pi space. In this way, the power response of the speakers are the most even, when comparing to normal OB speakers (with or without a rear tweeter) and sealed boxes. Diffractions can be substantially reduced, rear waves, which often cause more harm than good in a small room, are substantially reduced, and power response is a lot more even. I hope this retains most benefits and removes most defeats of the typical OB speakers.
Boxed speakers are not better. For flat frequency response at listener's position, the power response can be 6dB higher below the baffle step. This again produces fairly uneven power response. And perhaps this gives it a "boxed speakers" sound?
That is why I am now trying the damped U-frame OB without a rear tweeter. Above baffle step, the front waves radiate into 2Pi space. There is no rear waves from the tweeter, while the rear waves from the midwoofers are heavily damped to be insignificant. Below baffle step (300-600Hz), dampping materials become less effective (due to limited thickness comparing to the wave lengths) so the radiation gradually changes to cardiod then dipole. Not only does it provide directivity but also reduces the power response comparing to sealed box. Below about 150Hz-200Hz, a sealed woofer box takes over. Due to the floor boundary, it again radiates into 2pi space. In this way, the power response of the speakers are the most even, when comparing to normal OB speakers (with or without a rear tweeter) and sealed boxes. Diffractions can be substantially reduced, rear waves, which often cause more harm than good in a small room, are substantially reduced, and power response is a lot more even. I hope this retains most benefits and removes most defeats of the typical OB speakers.
bwaslo said:
Hey Bill
But what if you do hear something? What does that indicate? You have to somehow scale it to the direct sound and then see if its audible above the masking. Thats another level of problem.
I think that I've said this before, but I have tried my darndest to find these effects and can't. I gave up. What more can I say ... "box sound" is just not a significant effect if it exists at all.
The diffraction is different, the dispersion is different over a region, the panel resonances are different, the room sound is different. Those things surely form a distinction package. Don't they? No matter better or not, someone can distinguish the general fingerprint of a box type and an OB type in my experience.
P.S. The strong diffraction problem when opposites hit the baffle edge of an OB, can be ameliorated either with randomly perforated edges, or strongly curved ones.
P.S. The strong diffraction problem when opposites hit the baffle edge of an OB, can be ameliorated either with randomly perforated edges, or strongly curved ones.
Ummm, donuts ....
(With apologies to Homer Simpson) I thought that radially symmetrical shapes are worst possible case, even if they may solve diffraction. There is going to be a very strong on-axis interference pattern (caused by delayed back wave). Now, making an egg shaped donut with driver correspondingly offset ... is going to be very tough indeed. Even harder than oval waveguide.
(With apologies to Homer Simpson) I thought that radially symmetrical shapes are worst possible case, even if they may solve diffraction. There is going to be a very strong on-axis interference pattern (caused by delayed back wave). Now, making an egg shaped donut with driver correspondingly offset ... is going to be very tough indeed. Even harder than oval waveguide.
I hope this is not going off-topic for this thread. What I wish to do is this: I have a number of Audio Nirvana's (don't ask why.....
), both 6's and 8's.
I have decided to make a dipole using them, thusly: crossover in the tube amp (before the cathodyne stage) at 310 Hz (linkwitz PCB from Elliot);
send everything above the crossover through the tube amp to the Audio Nirvana in some sort of baffle open to the back, send everything below 310 to a mosfet plate amp driving a subwoofer. (Yeah, I know, there will be intersecting curves).
The front panel would have the AN mounted up top, the subwoofer mounted down lower. I was thinking the Dayton 12" 23 Hz to 600 Hz.
The question is, how wide to make the front panel, how deep to make the side panels, and how to contour either or both....aaaand how far apart to mount the subwoofer and the AN, aaand how far above the floor to mount the subwoofer...and the AN up above...and what angle to put the face at. OK, 8 questions!!!
The Orion appears to have more depth for the subwoofer, then the sides curve toward the front and don't have much depth for the mid and tweeter. 'Course, in mine, the mid and tweeter are combined in the AN and I don't feel the need to add a tweeter 'cuz the amp starts to round off the treble at 15 KHZ anyway (and I don't seem to hear much above 15 KHz anymore anyway! .
suggestions?
Best, Charlie
), both 6's and 8's. I have decided to make a dipole using them, thusly: crossover in the tube amp (before the cathodyne stage) at 310 Hz (linkwitz PCB from Elliot);
send everything above the crossover through the tube amp to the Audio Nirvana in some sort of baffle open to the back, send everything below 310 to a mosfet plate amp driving a subwoofer. (Yeah, I know, there will be intersecting curves).
The front panel would have the AN mounted up top, the subwoofer mounted down lower. I was thinking the Dayton 12" 23 Hz to 600 Hz.
The question is, how wide to make the front panel, how deep to make the side panels, and how to contour either or both....aaaand how far apart to mount the subwoofer and the AN, aaand how far above the floor to mount the subwoofer...and the AN up above...and what angle to put the face at. OK, 8 questions!!!
The Orion appears to have more depth for the subwoofer, then the sides curve toward the front and don't have much depth for the mid and tweeter. 'Course, in mine, the mid and tweeter are combined in the AN and I don't feel the need to add a tweeter 'cuz the amp starts to round off the treble at 15 KHZ anyway (and I don't seem to hear much above 15 KHz anymore anyway!
.suggestions?
Best, Charlie
I also proposed an open baffle torus here three years ago but with variation in the thickness of the torus to give an cardiod shape variation in the path length. http://www.diyaudio.com/forums/showthread.php?postid=698312#post698312
After posting that I realized it should be more like two cardiods overlapping to make it left right symetrical. This could even form a dipole elipitcal waveguide, thicker donut walls above and below the speakers with narrower waveguide angle here, tapering to thinner cross section at the left and right sides (and wider waveguide angle. The catch: what driver would work well in a thing lke this?
After posting that I realized it should be more like two cardiods overlapping to make it left right symetrical. This could even form a dipole elipitcal waveguide, thicker donut walls above and below the speakers with narrower waveguide angle here, tapering to thinner cross section at the left and right sides (and wider waveguide angle. The catch: what driver would work well in a thing lke this?
If we have to choose between being at the performance, or having the performers in our room, surely,as a rule, the former is preferable.
In other words, minimal room interaction is desirable, making directive monopole designs a better choice than dipoles, especially in average size listening rooms. Of course, Dr Geddes already figured this out years ago.
In other words, minimal room interaction is desirable, making directive monopole designs a better choice than dipoles, especially in average size listening rooms. Of course, Dr Geddes already figured this out years ago.
Hi David,
You seem to be giving the argument FOR OB speakers.
In my experience, with "real" properly miked classical concerts, for example, an OB sets the performance space and ambience cues behind and separate from the speakers. i.e you are at the concert. (what OB speakers give you the sound you are talking about?)
David
You seem to be giving the argument FOR OB speakers.
In my experience, with "real" properly miked classical concerts, for example, an OB sets the performance space and ambience cues behind and separate from the speakers. i.e you are at the concert. (what OB speakers give you the sound you are talking about?)
David
I never heard a "bad" dipole, even using $4 jaycar midrange on a baffle sounds good. There is something inherently correct about their sound reproduction capability. I think it is the holy grail of tranducer setup. Something Zen-ish like only the simplest can bring the truth blah2.. WHy... I even play metallica on OB and they sound great! Jazz, blues.. ooh too easy.
So to answer the question, yes I think being a dipole is one major factor why Orion would sound good (never heard them). I just wish they weren't soooooooooooooo expensive to begin with
I got the Pluto but their space was soon taken over by another pair of OBs after few months so you can say I'm biased.
So to answer the question, yes I think being a dipole is one major factor why Orion would sound good (never heard them). I just wish they weren't soooooooooooooo expensive to begin with
I got the Pluto but their space was soon taken over by another pair of OBs after few months so you can say I'm biased.
gainphile said:I never heard a "bad" dipole, even using $4 jaycar midrange on a baffle sounds good. There is something inherently correct about their sound reproduction capability. I think it is the holy grail of tranducer setup. Something Zen-ish like only the simplest can bring the truth blah2.. WHy... I even play metallica on OB and they sound great! Jazz, blues.. ooh too easy.
So to answer the question, yes I think being a dipole is one major factor why Orion would sound good (never heard them). I just wish they weren't soooooooooooooo expensive to begin with
Gainphile,
Why did the Pluto's get put to the side for another OB? I have not heard the Pluto's but they are supposed to sound very, very similar to the Orion's if I am not mistaken.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.