• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). There are daily reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Introducing Illusio Audio and its first product, the Alana loudspeaker

sarumbear

Master Contributor
Forum Donor
Joined
Aug 15, 2020
Messages
5,931
Likes
5,348

Curvature

Senior Member
Joined
May 20, 2022
Messages
326
Likes
389
Those designs have been around for a while but they never caught on. I had a Linn Isobarik which had a ceiling firing tweeter. It sounded good for a while until I decided that it’s a bit like the Bose 801, which I also owned.

I prefer point sources.
Those are from different days. The original Revel Salon had a rear firing tweeter, aiming to do a similar thing, too: https://www.stereonet.com/forums/uploads/monthly_04_2015/post-142617-0-09408300-1429442227.jpg

Not fully considered designs IMO.
 
OP
Duke

Duke

Major Contributor
Manufacturer
Forum Donor
Joined
Apr 22, 2016
Messages
1,099
Likes
2,639
Location
Princeton, Texas
As far as I know that's a result very similar to Geddes' Nathan/Abbey/Summa designs, when looking at the available published polars.

Good catch! The resemblance to Geddes' designs is no accident. Are you up for some back-story?

In 2001 I attended the Consumer Electronics Show in Las Vegas. At the time I had been a high-end audio dealer for about a year, and my main product line was SoundLab full-range electrostatic loudspeakers (uniform 90 degree radiation pattern front and back transitioning to a dipole figure-8 at low frequencies, its radiation pattern consistency across the spectrum was remarkable then as well as now). Walking from room to room at the show, I didn't hear anything that I thought rivalled the SoundLabs, until two particular horn speaker rooms: Classic Audio Reproductions, and Edgarhorns. Both had a liveliness that was really engaging and enjoyable. The SoundLabs did some things better, but then these horn speakers did other things better, and it seemed a matter of personal preference as to which was "best". I decided that I wanted a high-quality horn speaker in my lineup.

Based on my appreciation for the natural-sounding timbre of the uniform-patterned SoundLabs, I came up with what I thought was a brilliant original idea for a killer two-way horn speaker: 15" woofer combined with a 15" 90-degree constant-directivity horn, crossed over where the woofer's pattern had narrowed to match the horn's. Little did I know, JBL had done this two decades earlier with their landmark Model 4430 studio monitor. I spoke with one of my few contacts in the professional speaker designer world, Alan Hulsebus, and he told me that the man I needed to get in touch with was Earl Geddes. To say that was a pivotal point in my life would be an understatement.

I contacted Earl and described what I wanted to do, and he said that he'd been wanting to do essentially the same thing for about ten years. He proposed that we make it a joint effort, with him providing the engineering and me paying for parts and materials. This went on until I had some financial set-backs and was unable to uphold my end of the bargain, at which point Earl completed the project on his own and it became the GedLee "Summa". For a little while I assembled Summas for Earl, until a short-notice move resulted in me no longer having an adequate work-space. I had the highly educational good fortune of being present when he designed the crossovers, my lofty role in the measurement process being to rotate the enclosure at so-many-degrees intervals in between sweeps while he recorded the spin-o-rama data.

Incidentally before he made the first Summas, Earl used JBL 4430's!

Most of the things I do trace back to Earl in one way or another, even things he disagrees with like my rear-firing horns. The Swarm four-piece subwoofer system which is sold under the AudioKinesis brand traces directly to a conversation I had with Earl at CES 2006, at a stop-light. He briefly described his idea for using multiple subwoofers intelligently distributed, so I asked if I could license the idea, and he said, "nah, you can just use it." Then the light changed. It was that fast.

Not long ago I described to Earl my recent adaptation of his Oblate Spheroid geometry to a 1.4" throat waveguide (the Alana uses a 1" throat waveguide), and he approved of the direction I took. That particular waveguide will show up in a subsequent model from Illusio Audio, the crossover for which has already been designed, but the enclosure builds were too far out time-wise for the Capital Audio Fest show.

So yeah, Earl Geddes. My mentor. And my friend.
 
Last edited:

Curvature

Senior Member
Joined
May 20, 2022
Messages
326
Likes
389
Good catch! The resemblance to Geddes' designs is no accident. Are you up for some back-story?

In 2001 I attended the Consumer Electronics Show in Las Vegas. At the time I had been a high-end audio dealer for about a year, and my main product line was SoundLab full-range electrostatic loudspeakers (uniform 90 degree radiation pattern front and back transitioning to a dipole figure-8 at low frequencies, its radiation pattern consistency across the spectrum was remarkable than as well as now). As I walked from room to room at the show, didn't hear anything that I thought rivalled the SoundLabs, until two particular horn speaker rooms: Classic Audio Reproductions, and Edgarhorns. Both had a liveliness that was really engaging and enjoyable. The SoundLabs did some things better, but then these horn speakers did other things better, and it seemed a matter of personal preference as to which was "best". I decided that I wanted a high-quality horn speaker in my lineup.

Based on my appreciation for the natural-sounding timbre of the uniform-patterned SoundLabs, I came up with what I thought was a brilliant original idea for a killer two-way horn speaker: 15" woofer combined with a 15" 90-degree constant-directivity horn, crossed over where the woofer's pattern had narrowed to match the horn's. Little did I know, JBL had done this two decades earlier with their landmark Model 4430 studio monitor. I spoke with one of my few contacts in the professional speaker designer world, Alan Hulsebus, and he told me that the man I needed to get in touch with was Earl Geddes. To say that was a pivotal point in my life would be an understatement.

I contacted Earl and described what I wanted to do, and he said that he'd been wanting to do essentially the same thing for about ten years. He proposed that we make it a joint effort, with him providing the engineering and me paying for parts and materials. This went on until I had some financial set-backs and was unable to uphold my end of the bargain, at which point he completed the project on his own and it became the GedLee "Summa". For a little while I assembled Summas for Earl, until a short-notice move resulted in me no longer having an adequate work-space. I had the highly educational good fortune of being present when he designed the crossovers, my lofty role in the measurement process being to rotate the enclosure at so-many-degrees intervals in between sweeps while he recorded the spin-o-rama data.

Incidentally before he made the first Summas, Earl used JBL 4430's!

Most of the things I do trace back to Earl in one way or another, even things he disagrees with like my rear-firing horns. The Swarm four-piece subwoofer system which is sold under the AudioKinesis brand traces directly to a conversation I had with Earl at CES 2006, at a stop-light. He briefly described his idea for using multiple subwoofers intelligently distributed, so I asked if I could license the idea, and he said, "nah, you can just use it." Then the light changed. It was that fast.

Not long ago I described to Earl my recdent adaptation of his Oblate Spheroid geometry to a 1.4" throat waveguide (the Alana uses a 1" waveguide), and he approved of the direction I took. That particular waveguide will show up in a subsequent model from Illusio Audio, the crossover for which has already been designed, but the enclosure builds were too far out time-wise for the Capital Audio Fest show.

So yeah, Earl Geddes. My mentor.
What a story. No idea the connections ran this deep.
 

BlackTalon

Senior Member
Joined
Apr 14, 2021
Messages
359
Likes
578
Location
DC
More incentive for me to make it to the audio show in a couple weeks.
 
OP
Duke

Duke

Major Contributor
Manufacturer
Forum Donor
Joined
Apr 22, 2016
Messages
1,099
Likes
2,639
Location
Princeton, Texas
More incentive for me to make it to the audio show in a couple weeks.

Please introduce yourself if you come by! I might not recognize you from your avatar...
 

Andrej

Member
Forum Donor
Joined
Mar 21, 2021
Messages
54
Likes
55
This is a very interesting approach to sound reproduction! I would like to learn more, in general, and also wrt this particular approach. Assuming rectangular listening spaces.
This is the first and only approach I know of which allows you to "be there" without removing most of the room effects with cumbersome/expensive/esthetically compromised solutions.

- Why 90deg waveguide? Which way should the speaker point to get the best amount of side wall reflections? Can one trade off dispersion vs. pointing angle by the speaker? Is this applicable to any speaker, or only those which have a back firing component like your designs? (I do remember you mentioning a 15deg angle to the speaker, is that assuming an equilateral triangle between the listener and the speakers?)
- How low in frequency should the directionality (90deg) be maintained? In order to play with it I am building a waveguide which can go down to <150Hz:)
- You talked about the backward facing drivers bandwidth and directionality. I have same questions about how low it needs to go, and would adding time delay help/what should it be (how far should the speaker be from the rear wall, and can you compensate with time delay)
- I assume that one can experiment by just adding a drive unit to any speaker (which already has the right characteristics for the front facing part). Am I correct?
- If adding room treatments, where should one put them? They are essentially low pass filters on reflected sound. How does your approach change their application?

Sorry about all the questions, and if you are not comfortable sharing any of the secret sauce, I understand!
 
OP
Duke

Duke

Major Contributor
Manufacturer
Forum Donor
Joined
Apr 22, 2016
Messages
1,099
Likes
2,639
Location
Princeton, Texas
This is a very interesting approach to sound reproduction! I would like to learn more, in general, and also wrt this particular approach. Assuming rectangular listening spaces.
This is the first and only approach I know of which allows you to "be there" without removing most of the room effects with cumbersome/expensive/esthetically compromised solutions.

Thank you for the encouragement. I assume you have some background in the psychoacoustics of reflections in rooms, because usually the best reaction I can hope for is polite skepicism.

- Why 90deg waveguide?

A 90 degree waveguide gives good off-axis soundstaging across a wide area, sometimes with the enjoyable listening area extending outside the speakers. Not every location within this large area will be equally good, and really good soundstaging is confined to up and down the centerline.

Which way should the speaker point to get the best amount of side wall reflections?

I use 45 degrees of toe-in, which (with a 90 degree waveguide) results in the near-side-wall being far enough off-axis that it is only weakly "illuminated" by the waveguide.

Can one trade off dispersion vs. pointing angle by the speaker?

Yes. With a narrower pattern you can use less toe-in, but imo you still want the axes criss-crossing in front of the listener. With a wider pattern (which I haven't tried) you might want to use more toe-in to avoid a significant early same-side-wall reflection.

Is this applicable to any speaker, or only those which have a back firing component like your designs?

The toe-in suggestion is at best hit-or-miss with speakers which do not have a fairly narrow and uniform radiation pattern. An add-on Late Onset Reflection Assist module has been beneficial with other speakers having more conventional radiation patterns, but I don't yet have a wide enough range of experience with it to make sweeping claims. My guess is that "it depends".

The LORA stand was designed with possible use for other stand-mount speakers in mind. It has a switchable notch filter which introduces some dippage in the 3-5 kHz region, to allow for the off-axis energy flare that many stand-mount two-way speakers have above their crossover region, due to the tweeter's wide pattern at the bottom end of its range.

(I do remember you mentioning a 15deg angle to the speaker, is that assuming an equilateral triangle between the listener and the speakers?)

Yes, but that's not carved in stone. Adapt angles and toe-in to your speakers and room.

- How low in frequency should the directionality (90deg) be maintained? In order to play with it I am building a waveguide which can go down to <150Hz:)

Pattern control down to 150 Hz is FANTASTIC!! Dude, you are HARD CORE!!

The lower you can get pattern control the better, all else being equal, BUT all else is NEVER equal (in particular enclosure size and/or complexity). The region from 700 Hz to 7 kHz is apparently the region which matters most (based on David Griesinger's writings). The little Alana misses that 700 Hz target by about an octave, but we still think it's a worthwhile improvement.

- You talked about the backward facing drivers bandwidth and directionality. I have same questions about how low it needs to go, and would adding time delay help/what should it be (how far should the speaker be from the rear wall, and can you compensate with time delay)

The directionality is to minimize any "leakage" arriving during that first-10-milliseconds-window when we're trying to minimize early reflections.

In this application the rear-facing driver starts rolling off in the same region as the front drivers' crossover (1.4 kHz), but the rear-facing driver's rolloff is quite gentle down to about 700 Hz, then accelerates rapidly. This gentle rolloff over an octave or so approximately offsets the pattern-widening of the midwoofer as we go down from the crossover region.

Minimum distance from the wall for the rear-firing driver is a foot or less. The up-and-back angle + the directivity gives us about 10 milliseconds of delay, relative to the direct sound.

You absolutely could use delay for the rear-firing drivers, and in that case you could avoid the complexity of aiming that energy up at an angle.

OR... you might even consider deliberately aiming the additional driver(s) at the theoretically ideal first-reflection zones, along the same-side-walls, delayed to arrive maybe 15-20 milliseconds behind the direct sound. Toole reports that the ideal arrival direction for reflections is about 30 degrees forward of directly to the left and right, or at about 10 o'clock and 2 o'clock. Apparently Ken Kantor felt the same, unfortunately his MGC-1 was a failure in the marketplace:


What we're doing is a passive solution that you can just drop into a "normal" system, but there are arguments for going fully active and taking advantage of the capabilities DSP offers.

- I assume that one can experiment by just adding a drive unit to any speaker (which already has the right characteristics for the front facing part). Am I correct?

Imo there are multiple viable options. In a sense ye olde rear-firing tweeter is a variation on the theme... or maybe we're just doing a variation on the rear-firing tweeter theme. However OUR song-n-dance has MUCH better choreography!

- If adding room treatments, where should one put them? They are essentially low pass filters on reflected sound. How does your approach change their application?

I am NOT an acoustician, so consider this to just be opinion.

The front wall between the speakers becomes even MORE susceptible to illumination because of the aggressive toe-in. I'm not sure you'd necessarily want to treat the entire area between the speakers, but you might want to pay attention specifically to where the first reflections occur (as seen from the sweet spot).

The wall behind the listener is almost always an issue. If it's really close behind your head, that might be a place to put just enough thick absorption to catch the first reflection from each speaker.

I'm less inclined to treat the floor and ceiling bounces.

Personally I'm inclined to re-direct the first reflections as long as the room isn't inherently too lively to begin with. I prefer diffusion over absorption in general, especially along the side walls. In particular, we want to avoid absorption of the rear-firing energy's first few reflections on its way to the listening area.

Sorry about all the questions, and if you are not comfortable sharing any of the secret sauce, I understand!

No problem, the deepest darkest secrets remain hidden, and thanks for your interest!
 
Last edited:

Andrej

Member
Forum Donor
Joined
Mar 21, 2021
Messages
54
Likes
55
As you probably remember, I like absorption to extreme levels in order to minimize room influences to sound reproduction, aiming to preserve the ambiance in the original recordings. However, most recordings which use many close miked sources can be helped with a creation of ambiance illusion, where a more reflective/diffuse environment can help. Appropriate equalization to be used at all times. Your solution seems to be very beneficial in the latter approach. What I would miss the most, I guess, is the decay speed of the room resonances. One can remove the peaks with DSP, but the resonance will keep ringing, which I believe to very much be audible. Why does life always have to be a compromise:)
 

Andrej

Member
Forum Donor
Joined
Mar 21, 2021
Messages
54
Likes
55
Thanks for all the information! Given what you describe, if the speaker is somewhere along the room diagonal, towards a front corner, pointing at 45 deg, I'd probably aim for less than 90deg dispersion, and as for absorption, I'd play with it for a while to see where I find it most beneficial. I am including the picture of a spherical waveguide ( I just like the idea that it acts as a dispersion-limited pulsating sphere:) for an 8" driver (it can be used with a range of sizes) with current width of 39", height of 22" and depth about 28". The height will end up being about 26" after I add the final bit of the rounding at the edges. My biggest problem at the moment is to figure out a way to minimize it's vibrations, as the surface area is large and will be prone to color the sound. One idea is to add a 1/8" outer layer joined with green glue to form a constrained layer to increase damping, or the same but spaced 1" at the edges and 2" in the middle (parallel to the waveguide sides)
20220101_203815.jpg
20221028_091955.jpg
20221028_092006.jpg
and filling the gap with expanding foam. Light and stiff. Still mulling over the benefits and practical (construction and esthetics) consequences. Any suggestions? The HF unit is the Dayton AMT which I like a lot, in a similar shape waveguide, which keeps distortion down 45-55dB in the passband at 103dB/1m.
 
Top Bottom