Thank you very much, Jeffrey! Did we meet? I recall meeting someone who told me he was on WBF (aside from Steve and Ron and Keith), but I either didn't catch his name or promptly fumbled it.
We appreciate the encouragement!!
Something wicked this way comes... (2020 speaker teaser content)
- Thread starter Duke LeJeune
- Start date
Thank you very much, Jeffrey! Did we meet? I recall meeting someone who told me he was on WBF (aside from Steve and Ron and Keith), but I either didn't catch his name or promptly fumbled it.
We appreciate the encouragement!!
We met briefly in your room. Great sound out of a difficult room.
Thank you Marty! It is very encouraging to hear from someone who has been down all of the alternative roads and arrived at the same conclusion.
If all goes well, the native roll-off of the mains will match up well with any subwoofer whose top end is filtered at 24 dB per octave (that's the filter slope in the amp I use for the Swarm). The reason for the steep rolloff is, some of the multiple subs will normally be positioned fairly far from the mains, for the sake of modal smoothing. If those subs are passing upper bass/lower midrange energy loud enough to be audible, they can be localized as sound sources, which we don't want.
Thank you for noticing that!
I went to the corresponding room on another floor, and in my opinion they were having considerably more trouble with the room than we were. I'd heard the speakers in that room sound much better at other shows.
One of the manufacturers we were sharing our room with said to me after the first time we showed together, "This is the first show I've done where we haven't been fighting with the room the whole time." Imo that's one of the things our unorthodox approach does well.
Duke the more I think about what your proposing the more interesting it becomes... very much looking forward to the first pics... btw any chance I can change my vote for you aiming for a Sota $751k sticker price... just sayin.
Another update:
Today I picked up the tooling for the two different versions of our big round horn. Both are 22" in diameter with a 1.4" throat, an Oblate Spheroid profile, and a coverage pattern of 75 degrees. They differ from one another mainly in entry angle, each being optimized for a different, very promising compression driver. But that difference in the entry angle results in a slightly different optimum curve. I designed them using Earl Geddes' equation as a starting point. It turned out that more complex equations were called for (with different constants, due to the different entry angles). College being many decades ago, the math to figure out the correct equation for each horn took me about a day. My brain still hurts.
The actual "horn" itself takes up only about half of the 22" diameter; the rest is made up of the round-over around the inside perimeter of the mouth plus an outer round-over (to avoid edge diffraction). The outer round-over actually wraps around to the backside of the horn, which might end up looking pretty sweet perched atop a big midbass bin. The horn should have good pattern control down to a little over 600 Hz, though I don't yet know that we'll be crossing over that low. Once I get into the actual design stage, the horn and compression driver will tell me where they want to be crossed over.
We will be able to test more than two very promising compression drivers: There is a third, brand new compression driver which will match up well with the entry angle of one of the horns. Months ago I placed an order for what would have been the first pair to enter the United States. Then the factory shut down due to the coronavirus, timeframes became big unknowns, and remained so even after the factory re-opened a few weeks ago. Well this week I found out that those first two drivers have been manufactured and are on a ship that just left Europe. I will be receiving them in a few more weeks.
A bit more about these three very promising compression drivers: One uses a Beryllium diaphragm; one uses a carbon fiber diaphragm; and one uses a polymer diaphragm with a unique geometry which dramatically reduces breakup. All have powerful Neodymium motors, and all have top ends which rival or exceed a really good Titanium diaphragm in extension but with much better smoothness (due to lack of breakup). We can't figure out which is best for our application without measuring and hearing all three on our horn(s).
In the spirit of due diligence, we will also be testing a custom version of one of these three drivers. The "custom" aspect will be a different voice coil, which theoretically should be better suited for a home audio application than the regular voice coil. The regular voice coil has impressive power handling (even by prosound standards), but we don't "need" that, so we are going to try a lighter-weight voice coil.
We are actually planning to do two different loudspeaker models: A big, high-output, high efficiency version with two 15" midwoofers; and a smaller (but still fairly high-efficiency) version with two 12" midwoofers. Both will use the new 22" round horn.
The project is accelerating at last. Completion is still several months away, but there is now a lot less uncertainty in the word "several" than there was even a week ago.
Today I picked up the tooling for the two different versions of our big round horn. Both are 22" in diameter with a 1.4" throat, an Oblate Spheroid profile, and a coverage pattern of 75 degrees. They differ from one another mainly in entry angle, each being optimized for a different, very promising compression driver. But that difference in the entry angle results in a slightly different optimum curve. I designed them using Earl Geddes' equation as a starting point. It turned out that more complex equations were called for (with different constants, due to the different entry angles). College being many decades ago, the math to figure out the correct equation for each horn took me about a day. My brain still hurts.
The actual "horn" itself takes up only about half of the 22" diameter; the rest is made up of the round-over around the inside perimeter of the mouth plus an outer round-over (to avoid edge diffraction). The outer round-over actually wraps around to the backside of the horn, which might end up looking pretty sweet perched atop a big midbass bin. The horn should have good pattern control down to a little over 600 Hz, though I don't yet know that we'll be crossing over that low. Once I get into the actual design stage, the horn and compression driver will tell me where they want to be crossed over.
We will be able to test more than two very promising compression drivers: There is a third, brand new compression driver which will match up well with the entry angle of one of the horns. Months ago I placed an order for what would have been the first pair to enter the United States. Then the factory shut down due to the coronavirus, timeframes became big unknowns, and remained so even after the factory re-opened a few weeks ago. Well this week I found out that those first two drivers have been manufactured and are on a ship that just left Europe. I will be receiving them in a few more weeks.
A bit more about these three very promising compression drivers: One uses a Beryllium diaphragm; one uses a carbon fiber diaphragm; and one uses a polymer diaphragm with a unique geometry which dramatically reduces breakup. All have powerful Neodymium motors, and all have top ends which rival or exceed a really good Titanium diaphragm in extension but with much better smoothness (due to lack of breakup). We can't figure out which is best for our application without measuring and hearing all three on our horn(s).
In the spirit of due diligence, we will also be testing a custom version of one of these three drivers. The "custom" aspect will be a different voice coil, which theoretically should be better suited for a home audio application than the regular voice coil. The regular voice coil has impressive power handling (even by prosound standards), but we don't "need" that, so we are going to try a lighter-weight voice coil.
We are actually planning to do two different loudspeaker models: A big, high-output, high efficiency version with two 15" midwoofers; and a smaller (but still fairly high-efficiency) version with two 12" midwoofers. Both will use the new 22" round horn.
The project is accelerating at last. Completion is still several months away, but there is now a lot less uncertainty in the word "several" than there was even a week ago.
Last edited:
Duke you are in serious tease mode territory now... just sayin.
Thank you sir!!
It's almost like, this bundle of unorthodox ideas chose me, rather than the other way around. It's like they asked me to give them the best possible chance of success. Not that I'm "special" - more likely all the smart guys passed on them, and then they spotted me, with my glasses pushed halfway down my nose.
It's a bit risky posting this much at this early stage, because things can and do go wrong! If that happens, I'll be open about it and learn as much as I can from whatever went wrong, and try again.
Let me try to explain why I do not want to do a three-way, wherein the midrange horn hands off to a treble horn for the very top end:
Briefly, if we get the phase response right, clarity and perceived dynamic contrast benefit. There may be imaging-related benefits, but imo those are not the most important ones. I think it's easier to get good phase response in the region that matters with a good two-way.
Concert hall acoustician and psychoacoustics researcher David Griesinger has found that clarity is not only degraded by early reflections, it is also degraded by poor time-domain response north of 800 Hz. When the harmonics north of 800 Hz (and in particular between 1000 and 4000 Hz) all arrive at the same instant, they combine in a way that the ear perceives as a higher signal-to-noise ratio, which increases clarity and causes us to pay attention. On the other hand if the arrival of the harmonics is smeared out in time, clarity and dynamic contrast suffer, and the music (or speech or whatever) does not easily hold our attention. This difference is a result of the way the ear/brain system processes sound; it will not be picked up by a microphone or SPL meter.
(Note that many if not most speakers have a crossover in this 1000 to 4000 Hz region, which implies a little bit of "room for improvement" over most speakers, IF Griesinger is right... and of course I still have do my part well.)
By covering the region from ballpark 800 Hz on up with a single driver, we think we are getting off to a good start.
The physical depth of our horn will introduce the correct amount of path-length-induced delay to effectively time-align it with the midwoofer, assuming we use filters with the correct slopes (which we do). This also helps line up the harmonics in the time domain.
Unfortunately not very many compression drivers can cover 800 Hz on up to 20 kHz or so. Of those which do, most have significant ringing in the top octave. The compression drivers we are looking at are among the few which do not.
The standard solution is to use a midrange horn plus a tweeter horn. An immediate dilemma arises: Do we align the voice coils for good time-domain response, and maybe space them far apart so that the tweeter's output doesn't early-reflect off the midrange horn? Do we align the mouth of the tweeter horn with the mouth of the midrange horn, thereby guaranteeing that the former's output doesn't reflect off of the latter, but resulting in arrival times several cycles apart? Or, do we use DSP to correct the time domain, and hope it contributes no "signature"?
Obviously most designers of big high-efficiency horn systems have opted to use a midrange horn + a tweeter horn, and sometimes they use more than that. Maybe we'll learn the hard way that the attributes we've prioritized don't matter as much as the attributes they have prioritized. One thing we know we'll be trading off is a bit of efficiency - a good big driver on a good big horn isn't as loud in the top octave as a good small driver on a good small horn. On the other hand, our target curve is not "flat" out to 20kHz; it is a gently downward-sloping curve. So we are okay with the top end being a few decibels down relative to the midband and low end.
And just for the record, I haven't revealed all the tricks we have up our sleeves. Yet.
Tease, tease.
Last edited:
Love it! Duke I am also very much committed to 2 ways (this just looks wrong when I write it) and my current Pap horns with 15 inch dual woofers are proving fantastic and the more I live with these the more I appreciate the virtues of simplicity. They have continuity, coherence and immediacy.
So I’m also just very keen to see how this develops for you... the journey of exploration and the intention to do something lastingly good and build upon the current canon of horns is awesome. Two thumbs up from down here.
Both of my horn systems (my Odeons and my DIY system) are 2-way designs and I find that they don't give up anything with regard to coherence to a single driver system (the DIY system is based on a full range driver so I can go back and forth between 1 way and 2 way rather easily) and it deals with complex music (like orchestra) much better than a single driver.
While I have enormous respect for Beyma, and have made & sold more high-end home audio speakers with Beyma compression drivers than all others combined, no. No Beymas in the finals this time.
Ok, not willing to share what you are using? I noticed on your website you pretty open about what brands you are using. I still want to get a Radian in the system at some point to compare with the two Beymas I use currently (CP350Ti and CP755Ti).
I heard last weekend the German brand Live Act Audio, which is using the Radian Coaxial drivers (everything from 8 inch up to 15 inch) with Be compression tweeters. Very natural sounding with superb transparency and cleanliness in the highs...really superb.
I haven't decided yet whether to say which drivers I'm testing, but at least one of them - whichever I decide to use - will come out eventually.
I've only seen images and one review on the Live Act speakers, but they look superb to me! There's also a Canadian company whose name slips my mind that uses Radian coaxials.
Imo Radian's coaxials use the RIGHT format - the woofer cone is the "horn"; and Live Act's "Reference" line are examples of the RIGHT way to use a coaxial - hand off the low frequencies to a separate woofer section so that the "horn" doesn't undergo long excursions.
I considered using Radian Coaxials, but think I can get closer to my goals using a separate horn. However it is really hard to argue against the inherent elegance of the coaxial format. The fact that Live Act was already doing such a good job with that format actually pushed me away from it; I have no desire to do a "me too" product.
That "superb transparency and cleanliness in the highs" you mention is one of the things I expect to end up with. The two drivers I have listened to so far both have that. I think it is predicted by waterfall plots.
Live Act is actually following the Dynamikks model (who has been doing the two way coax for more than 10 years now) but are arguably using a better coaxial driver (the one from Dynamikks is very good but it doesn't use a Be diaphargm and I think this can be heard). Live Act's lower range is using the Coax full range and that works SURPRISINGLY well. I was really impressed with the smallest model, the 108, with a 16 watt 845 integrated. On paper, this shouldn't have made so much bass as the resonance frequency of the 8 inch Radian driver is not very low...but in practice it rocked! A friend of mine has the next model up (the 110) and he says it is really singing (he has big Odeons as his main room speakers). We also heard on this trip the 312, but since it is using an active crossover between the bass drivers and coax, we didn't feel it was sounding so great (the active crossover was in the circuit not only for the bass but also rolling off the coax...a big mistake IMO as you could easily hear what this active xo was doing...wrong). Finally we listened to the fully passive Reference 408, which uses the same coax as the 108 but a massive cabinet with three 8 inch woofers per channel, and that to my ears was pretty awesome. My friend only didn't like the ultimate bass depth reach (it was fine for me) but the punch of the bass and the timing of transients was pretty much...live. Really good but also a bit steep on the price (list is 50K!!).
VERY INTERESTING!!
Okay, part of what's happening MIGHT be this: The relatively high resonant frequency of the Radian coaxial helps minimize cone excursions from signals well below the enclosure's tuning frequency, which in turn helps to preserve the inherent clarity of the coaxial compression driver.
An alternative coaxial format is to use a separate horn for the compression driver. I have yet to find one with this format which did not have a nasty dip in the frequency response in the upper midrange region from the cone's output reflecting off the back of the horn. And imo it's not so much the resulting measurable frequency response dip which degrades the sound, but the delayed arrival of that distorted reflection. So I prefer the format Radian uses, and they do it very well.
I've wondered if the low frequency bounce the higher ones in a coaxial, because they do on fullrange drivers.
Duke's new speaker sounds amazing. I can't image it won't be something people highly desire if it looks a good as I suspect. And maybe underpriced...
Duke's new speaker sounds amazing. I can't image it won't be something people highly desire if it looks a good as I suspect. And maybe underpriced...
Similar threads
