He was correct with regard to power response, but not (as he later acknowledged) with regard to what he now calls the "acoustic image".
Probably so . . . but mooted by the fact that a rear tweeter active from 7kHz up is dirt cheap and trivial to implement ($40 in LX521, and could be less, I expect). Without it there's a risk (not worth taking) of excessive brightness on axis to maintain uniform room response.
In order: doesn't seem to be an issue in the actual implementation . . . the upper-mid never sounds strained; me too, but there are other issues invloved; fair question, although unless you're into current drive the "control argument" applies (always has applied) more to woofers and subs than it has to tweeters.Well I have and continue to question that based on excursion. Plus, I would have placed the passive between the tweeter and upper mid. Makes more sense from a power spectrum point of view. And, what happen to direct connection to the amp yields better control argument.
I doubt that localization was actually improved. But it's also true that localization blur says nothing about perceived realism.
What you've experienced is most likely caused by specific strong early reflections. I'd like to see ETCs from such a configuration.
As I've said before, I doubt that a dipole speaker is the best way of achieving such perceptions.
Last edited:
Localization was actually improved. And it remained surprisingly "fixed" over a range of listening positions (walking around the room, not just at one "sweet spot").
Yes, "reflections" do seem to enhance imaging, not detract from it. Turns out to be an advantage of dipoles, which turn out (when well done) to produce a better, more defined and more believable "acoustic image" than other (especially "in-wall") designs.
It's not that difficult.
The new frame and "grill" is a good idea, and one I've suggested on many occasions to others.
It's the execution however that needs help.
The vertical angle to the front "panel" is a bad idea - made worse by the fact that it has no "arc". (..that's aesthetically of course, I realize that it's a necessary part of the design.)
The side cut panel on the sub looks really bad. It's a box - clean up the lines *first* and only then make decorative alterations that work with that shape.
If you are going to borrow from a design, consider the Monolith vs. III (..though obviously much smaller.) That design was pretty "clean". The vertical angle can largely be hidden by the new front "arc", and you can add (and should add) a rear "arc" to the rear sub panel..
..yeah, words don't really convey it do they?
Tonight I'll try a basic sketchup to provide a better idea of what I'm *trying* to describe (and will try to post it on your thread).
..btw, the lower cost implementation of the "new" sub drivers makes your design a LOT better from the perspective of value. It's just a lot more *accessible* now.
there is now enough scientific evidence that lateral reflexions are actually good, as long as they are spectrally close to the direct sound, which rules out 99% box speakers.
check this for example:
White Papers - Sausalito Audio
Dipoles do image very well because they have a stronger D/R ratio (3 over 1) and potentially better CD than standard speakers. The nice thing is that with a diffuse rear wall, you extend the scene and get extra ambiance. You still keep a sharp image. Certainly, if you want "razor" sharp stereo, than you can absorb everything behind. Stig Erik does that, his system is just plain scary in imaging department, it has to be heard really. Dipoles are not only good, they are adaptable!
Reading your comments I just feel your experience with dipoles is probably limited, am I wrong?
Last edited:
Yes, but it's the two midrange drivers that are not directly connected. Still, I I have always been on the other side of that equation. The control issue is a red herring when a system point of view is taken.
Back to the 1st order x-o, if it sounds good that is fine. But my experience for sure is that my revised Note system sounds better with a higher order crossover between the mids. That is one thing that is nice about having the Bodzio Ultimate EQ. I can set the system up with different order x-o's with a point and a click and guarantee that I have the same axial response to with in a small fraction of a dB. Difference in the sound are thne totally a result of how the drivers overlap in the stop bands.
The vertical angle to the front "panel" is a bad idea - made worse by the fact that it has no "arc". (..that's aesthetically of course, I realize that it's a necessary part of the design.)
The side cut panel on the sub looks really bad. It's a box - clean up the lines *first* and only then make decorative alterations that work with that shape.
If you are going to borrow from a design, consider the Monolith vs. III (..though obviously much smaller.) That design was pretty "clean". The vertical angle can largely be hidden by the new front "arc", and you can add (and should add) a rear "arc" to the rear sub panel..
..yeah, words don't really convey it do they?
Tonight I'll try a basic sketchup to provide a better idea of what I'm *trying* to describe (and will try to post it on your thread).
..btw, the lower cost implementation of the "new" sub drivers makes your design a LOT better from the perspective of value. It's just a lot more *accessible* now.
Vertical angle?
The side panels are from the NaO II RS. On that system the u-frame woofer is tapered so that the enclosure is hidden by the side panels.
An externally hosted image should be here but it was not working when we last tested it.
The dipole woofer in the new Note can not be tapered and retain the other aspects of the design. I could use the U-frame from the NaO II RS with the SLS woofers but I have decided to go full dipole because the u-frame is a little more complex and harder to set up correctly w/o measurements.
I'm not sure what you mean by Monolith vs. the (Monolith?) III. Just what is it you dislike about the Monolith I and II? I own a pair of II's (well I's updated to II's) and much prefer the look of them to the III's.
Different drivers, different baffle, all that stuff. Probably more a matter of perspective, though. You're thinking of the upper mid and lower mid as two separate drivers, independently equalized and then crossed over. SL treats them as a single "blended" driver (with different baffles and 1st order crossover) and then works with/equalizes the combined properties as if it's just one. It clearly didn't hurt that perspective that he had some control over the driver specs as well . . . they do seem to integrate seamlessly.
*sigh* This migth happen with a specific recording, in a specific room, with a specific speaker but it doesn't happen with any recording, in any room with any dipole speaker.
No, that is not entirely true. In the original Note design the lower ER18, upper ER18, the 10F and the Neo 3 were passively crossover as a unit and then a single active eq was applied to compensate for the dipole roll off and a couple of other issues. Quote form my web site:
Eliminate the tweeter and it is the same approach as the L521. Blend the mids and the apply a single eq.
It really makes no difference though. EQ x (DXo1 + DXo2) = Eq x DXo1 + EQ x DXo2. Look at my comparison between the original and revised:
The original is blend first, apply global EQ. The revise system is treating the drivers individually. The acoustic response is almost identical in both cases except for some fine tuning in the revised speaker.
Ever heard the Beolab 5?
That's the kind of discussion that really doesn't lead nowhere. Looking at actual room data would be a good start.
Ever heard the Beolab 5?
yes I have. It's not bad, but what do you mean with that?
The funny thing is that recently there has been some room data posted here, modulation transfer data to be precise, that clearly showed an advantage for dipoles in the bass. How do you get directivity below 500hz? I'd like to know.. If you are not happy with the rear wave thing, as said, just absorb it and you are done, you get a nice 60° directivity just at the front. Or is that too much?
The discussion doesn't lead anywhere because you just seem to chime in with negative comments every time dipoles are mentionned, whatever the design.
Last edited:
Vertical angle = "slant-back/tilt/rake" of the front panel.
Monolith version III (..not "vs." - my bad again.
) The III had much cleaner lines than either I or II (which among other problems had a noticeable/ugly "step" in the front profile from bass-box to upper panel).What I'm thinking of wouldn't really alter your design much (and almost not at all with respect to your design goals). Again.. I'll try and do a sketch tonight.
BTW, that one with the "grill" in-place looks a lot better than the one in the news section of your site without the "grill" (..though of course it's there in that pic as well, but only in the background).
What if I would tell you that I have heard the Beolab 5 too and it was one of the worst listening experiences I've ever had. What does that mean? It means that there is much more to good sound reproduction than "dipole good, boxed speaker bad" or vice versa.
No the data did not show that. Just follow the discussion.
You get directivity by using a cardioid, not a dipole but does it really matter below 500Hz anyway? We're talking about acoustically small listening spaces, right?
How do you effectively absorb the rear wave? That's just not very practical.
The problem is that you're simply defending your belief in dipoles instead of trying to keep an open mind. In case you've missed it, I did a little subwoofer comparison lately and the dipole sub came out on top. So much for calling me biased.
Vertical angle = "slant-back/tilt/rake" of the front panel.
Monolith version III (..not "vs." - my bad again.
BTW, that one with the "grill" in-place looks a lot better than the one in the news section of your site without the "grill" (..though of course it's there in that pic as well, but only in the background).
A) Monolith I, II vs. III. Matter of opinion. I like the I,II better.
B) Yes, that why the revised Note will have a grill. It is to hide the ugly baffle. As with the LX521, it's a form follows function thing. I have even thought about some type of perforated metal or something to fill the open areas between the baffle edge and the side pieces.
Going back to Dewardh and the 1st order x-o between the mids I can't help but wonder if it also wasn't more about trial and error. Certainly it would be easier to put a cap on the upper mid and a coil on the lower and dick around with the values until the desired measurement results were obtained. With a higher order passive I would imaging that it would require using sims to get the axial response right and then building it and measuring the polar response. That's where digital come in so nicely. I can make sims of the axial response and then emulate them digitally to check the polar. When I get what I want I can then decide whether to try to emulate the filter and eq transfer functions passively, in hybrid mode, or fully active, analog or digital.
But more than that, I started with making polar plots of the naked mids and tweeter and then using that info to guide me in the crossover design. See the latest update to my News page at the bottom.
That is an interesting question. I guess I have come around to Geddes way of thinking. Actually, I always was there but never realized it. This directivity thing at low frequency is a free field thing. There is no such thing as dipole or cardioid radiation in an acoustically small room. All there is are multiple monopoles with different position, phase, delay and eq this way or that. I did some recent measurement of a monopole woofer. I moved it 18" and looked at the response. The degree of differences was not insignificant. So, given that any H dipole woofer has a separation of around 14 to 20" more or less, you can hardly expect the front and rear sources to generate the same response at any point in the room. And what about a large baffle dipole? You can not even begin to model that as multiple point source monopoles. How would you model the effect of the baffle blockage for the source on the back side? It's very complicated.
Assuming the sources can be considered as point sources all that matters is the near filed response of each source, the position of the sources, the position of the listener, the source to listener transfer function for each, and how all this sums at the listening position.
Last edited: