Audiovector R 3 Arreté loudspeaker

My first encounter with an Audiovector loudspeaker was at the 2019 Toronto Audiofest. Driven by colorful (both sonically and visually) Alluxity electronics, the R 3s sounded pure and very fine. As I sat listening to the R 3 Arretés, the R 8 Arreté, their big brother, sat quietly in the corner, seemingly pleased with the performance of its smaller sibling.

I ended up reviewing the R 8 instead of the R 3, which in retrospect hardly seems fair:It was the R3 I heard that day, the R 3 that attracted my attention and got me interested in the brand. I didn't hear a peep from the R 8 Arreté—although I do recall being impressed by the larger speaker's gorgeous, glossy wood finish, which Audiovector calls Italian Walnut Burl Piano.

The R 3 is a 2.5-way bass-reflex speaker. Its high-frequency driver is an air-motion transformer (AMT). Its two front-firing 6.5" drivers, with carbon-fiber/aramid fiber/synthetic wood resin cones, appear identical but have different motors: "The magnet strengths, the coil windings, the suspension parts—spiders and roll surrounds—are meticulously matched to optimize the exact function of each driver, bass, mid bass, bass/midrange, etc.," said Audiovector CEO Mads Klifoth in an email. (Mads's father, Ole, is the designer and also the company founder.) The two drivers overlap at the lowest frequencies, but only the upper driver directly crosses over to the tweeter. The lower crossover frequency—the lower driver's upper rolloff—is specified as 320Hz.

Not unlike an auto company, Audiovector offers many of its loudspeakers—all but the top models—in various trim levels. The R 3 is available in Signature, Avantgarde, and Arreté. While they share a common cabinet, in some ways the three speakers are quite different from each other. The Signature, for example, has a soft-dome tweeter, while the Avantgarde and Arreté have AMT tweeters—different ones. All are 2.5-way speakers, with the same woofers and midrange drivers, but each of the three models has different crossover frequencies and power-handling abilities.

Audiovector likes to present its technologies in discrete, branded nuggets, which it calls "Concepts," each with a nifty logo. I admit that at first I found this perhaps a bit slick, but when I started paying close attention, I realized that most if not all of Audiovector's Concepts correspond to something not just real but unusual—even, in some cases, surprising, such as what at first glance appears to be a power cable, on a fully passive speaker (more on that in a bit). The key to understanding Concepts is to recognize that Audiovector does a few things differently than other loudspeaker manufacturers, then wraps it up and ties it with a bow.

At least two of those Audiovector concepts focus on what I'll call, more casually than they do, the pursuit of relaxation—a concept I endorse in lifestyle as well as in loudspeakers. Some high-mass loudspeakers can seem a bit, well, constipated; they seem to struggle to let go ... of the music. Audiovector speakers are designed to address that issue—to let go of music lightly, to let the music be. The most obvious "Concept" example is "NES," short for No Energy Storage. "Drive units sound best if they do not have to 'carry' the weight of the speaker cabinet to which they are fixed"; that's one way that Audiovector sums up "NES." NES dictates enclosures that are stiff but—in contrast to many other high-performance speakers—relatively lightweight. Each R 3 Arreté is just under 41" tall and according to Audiovector weighs about 50lb, making them exceedingly easy to handle. (Having handled them quite a lot, I'd swear they were lighter than that, but I didn't weigh them.) For comparable stiffness and material complexity, lighter weight means less stored energy. An analogy to Rega turntables (against mass-loaded turntables) might be applicable here, although it's probably more important with speakers than it is with turntables because, by design, speakers are exposed to far more energy than a turntable is.

What is Audiovector's rationale for NES? It's "a question of transient response," their literature says. NES allows their speakers "to achieve an even higher level of realism and open soundstage from our speakers." "Open" is the word.

A second example of Audiovector's relaxation aesthetic is what the company calls "LCC"—the Low-Compression Concept. (Just as I am an advocate of relaxation, I am an advocate of decompressing.) This, I think, gets at a different kind of energy storage: compressed air resonating inside the loudspeaker cavity. It's rather well-known that loudspeakers derive some of their sound from the vibrating air inside; the way that vibrating air is absorbed or released is critical to loudspeaker performance. The R 3 is a bass-reflex speaker like any other—you can't see it in the photos because the port faces the floor, venting front and rear through that aluminum platform that you can see—but in one way, the cabinet is unusually open. Instead of absorbing the AMT tweeter's back wave in a circuitous chamber, as other high-performance loudspeakers do, the R 3 vents the backwave directly out two ports at the top of the cabinet's rear panel. The backwave of most tweeters has nowhere to go except into the enclosure that contains it. That means compression in the short term, then absorption—brief storage—before it is released either as heat or as sound (footnote 1). Don't bottle up those highs. Let that energy fly.

The other Concept the open-backed tweeter implements is the one called SEC, for Soundstage Enhancement Concept. In broad terms, this means that Audiovector endeavors to make speakers with a wide soundstage—that sound good no matter where you sit (or stand) in the room. This is aided, I think, by deploying energy in various directions, including out the rear of the cabinet. In the tweeter's frequency range—the specified crossover point is 2.9kHz—the R 3 has some dipole character, although the volume of sound exiting the rear ports seems small.

Back to relaxation: There's a third way Audiovector works to minimize what I will call sonic or listening stress, employing that word in very broad terms: by dissipating electromagnetic energy stored in the driver chassis. They call this FGC, for Freedom Grounding Concept. FGC, Audiovector says, "addresses the movement-induced distortion of and between the drive units in a loudspeaker. The currencies [sic] running between the chassis are being processed and dealt with through a new separate crossover, which routes the signal to the ground terminal of your wall socket or your grounded mains distributor. These currencies cause coloration and distortion between the drive units. By balancing and filtering these through a dedicated separate filter and by offering the possibility of connecting them to earth/ground, we achieve a clean, very accurate, much more realistic sound with low noise floor." This second "crossover" is not, of course, in the signal path.

This grounding "crossover" terminates at an extra binding post on the back of the loudspeaker, near where the regular cables connect. A special cable runs from there to the wall, terminating in a standard plug, though only the "earth" ground is connected. No power is drawn from the wall. I'm still not clear on what's going on with this concept; maybe it has something to do with eddy currents.

The R 3 Arreté shares much with the R 8 Arreté, which I reviewed in 2021. Both speakers use Audiovector's highest-grade AMT tweeter, for example. The 6.5" midwoofers are shared. Both speakers have tweeters that vent out the back. For more information about the R 8 Arreté and about R-series Arreté speakers generally (there are three other models), readers are referred to that review/

Despite the similarities, the R 8 and R 3 Arretés are very different speakers. Start with the fact that the R 8 Arreté is much larger, has several additional drivers, and weighs three times as much. The R 8 is a complex speaker, which, as I wrote in my review, "sends energy out into the air via no fewer than nine sound-radiating openings, only four of them—the three midwoofers and the tweeter—firing forward. The others are the down-firing woofer (which radiates through those base slots to the back and sides), the rear-firing midrange driver (diffracted through those five horizontal grooves), the vent that allows the tweeter's back wave to emerge from the rear, and the two aforementioned ports." That's a lot for a designer to integrate, though with the R 8, Ole did a superb job: As I wrote in the review, "what I heard from the R 8s was anything but complicated: It was pure and coherent and easily grasped."

The R 3, in contrast, is simple not only in sound but also in design. Apart from the aspects already mentioned—its open-back tweeter; the grounding system—it's a straightforward if innovative bass-reflex loudspeaker in a lightweight cabinet, with, as previously mentioned, a downfiring port.

The detailed setup story for this review would take much too long to tell, because it involves exploring a brand-new space, which just a few short months ago was unfamiliar to me, acoustically and otherwise (footnote 2). The long process of learning and acclimation—living in the space, getting used to how it feels and how it sounds (these being closely related characteristics)—is, strictly speaking, part of the setup process. For nearly a year, I've been gradually building a system from different components, becoming familiar with each piece in context, determining its unique contribution to the experience of music in the space—work that will continue for months or years. Renovations are scheduled, and equipment will be swapped out as I gradually—hopefully—approach the sound I'm seeking. That I'm only in the space part of the time adds to the challenge. All this gives new meaning to Floyd Toole's phrase "circle of confusion."

It's a work in progress, but I have reached the point where I have sufficient confidence to pick apart the various influences on what I'm hearing, including the contribution of the loudspeakers.

After weeks of adjustment, I ended up with the R 3 Arretés positioned 90" apart tweeter to tweeter, the front baffles 32" from the front wall. Because of an annoying sofa, which is much too deep and will soon be removed, the speakers were forced away from one wall and toward another; the left one, regrettably, ended up just 8" from the left sidewall (which will eventually be removed). That's hardly optimal, but the R 3s took it in stride.

The tweeter on the R 3 was about 37" off the floor, slightly above the height of my ears at my listening seat, which was roughly 10' from the plane of the speakers. Even though that height put my ears near enough to the desired listening axis, for good measure I used the spikes to tilt the speakers forward by a couple of degrees.

For a description of the system, see the Associated Equipment sidebar at the end of this review.

Footnote 1: I use "sound" broadly here, encompassing air vibrations that may or may not be in the human audible range.

Footnote 2: Two summers ago, my wife and I bought a small house in Maine, where we lived for some 20 years before moving to New York City. We intend to retire there—though not soon. We go there when we can, usually on holidays, when my wife's students have gone away; she is a professor of chemistry. If all goes well, we'll start a major renovation soon, though the changes to the listening space will be modest—importantly, the removal of the partial wall that the left speaker is too close to.

F3/Audiovector ApS
Mileparken 22 A
DK-2740 Skovlunde
+45 3539 6060

supamark's picture

that huge (~8dB) suckout near the crossover point. It has to be audible since it gets worse off axis, and should not exist in a speaker at this price point.

PMC are the same way, always have a huge suckout, and cost too much to have it.

Mark Phillips,
no longer writing for any publication, nor working in the audio industry.

DaveinSM's picture

Yes, and that FGC AC grounding plug sure seems gimmicky to me. I kinda doubt I’m the only skeptic here…

Glotz's picture

God bless you in whatever endeavor you engage in.

supamark's picture

First, like when I was a recording engineer I spent most of my time listening to sounds and equipment instead of the music. Then after work I didn't want to listen to music. That's no good.

My last review (about 90% written, because it required two versions) was about one of the many THX AAA headphone amps out there, and that circuit doesn't sound good to me at all (and has I believe a fatal flaw, but I don't have the AP-555 receipts to publicly say what the flaw is - SoundStage does though which is why I wrote two versions, I might be wrong about the flaw) but they were not interested in it. It took me over a year to figure out why I hated listening to that amp so much. What I *can* say is it sounds terrible with Sennheiser HD600 headphones because the cone's distortion is way higher than the amp's distortion and it's annoying. I mostly use planars so that wasn't my issue. My issue was, if you're familiar with Heinlein, TANSTAAFL and I didn't like the price.

I will never intentionally listen to another amp that uses the THX AAA circuit, it's as bad as class D just in an entirely different way. With the caveat that I've never heard any of the GanFET amps, all class D amps I've heard (IcePower, NCore, Eigentakt) sound bad to me with high frequency phase issues, they sound grey/hard like bad 80's digital - synesthesia for the win. Writing a quality critical review of something one doesn't like is f'ing hard work. Slagging is fun and easy, but not professional. This is probably why you seldom see negative reviews from the major publications, they're very hard to write well. In the same time you could review 2 or more enjoyable products. Which makes more financial sense?

On top of that my mom died a few years ago which messed me up for a while, but left me in a very different tax bracket meaning my niche at SoundStage would have to change because I am far less constrained in the equipment I can afford (at retail price). They don't need that niche filled - the stuff I reviewed tended to top out around $1k and I have no interest in the budget stuff any more. There's been a fair bit of churn there the past year or so and I may have been the only "subjectivist" there as well... although my review of a power cable would probably be 1 sentence - I plugged it in to the wall outlet and it worked, good job guys. YMMV but I think that would be hilarious to print. I do have a hard science degree (biochem) so I got some objectivist in me.

Now I have time to persue what I *really* want to do, produce. More looking at the big picture with musicians - performance, arrangement, telling the engineer how I want it to sound so they get to listen to sounds instead of music lol, and of course corralling the cats and helping bring their artistic vision to life. If you're curious what a Mark Phillips engineered co-production sounds like listen to The FuckEmos albums "Can Kill You" or "Tape II". Strong artistic vision and guidance to get them there. It's on all the streaming svcs though I think the current mastering is not, uh, optimal. It's 16 track digital (ADAT) recorded and mixed in 2 days in my house at the time. It's also the only mix I ever used a compressor on the main mix (2 track) bus. It's funny, or offensive, depending on your sense of humor. I also personally remastered Tape II because I hated my own mix so much that I didn't listen to it for almost 30 years (not publicly available).

Indydan's picture

I have seen (but not heard) these speakers in person. The build quality is outstanding.
All the more surprising that the Audiovector engineers left in that huge 8db suckout. It baffles the mind.

hb72's picture

we are probably not as sensitive, hearing wise, to such rel. narrow band dips, as we are able to spot a flaw on a frequency response plot.

Having said that, I see one good reason to chose a proper 3-way concept over a 2 or 2.5way type, which is the possibility for a dedicated midband chassis: rel. low excursions permitting optimization potential not available to bass/mid Jack-of-all-trade chassis, resulting in lower mass and often a bit smaller dimensions & enhanced spatial radiation pattern. Using such high-excursion capable chassis that shall also cover high frequency at high quality (see price of a pair) is quite a stretch, and an earlier roll off perhaps the compromise to take..?? Not sure what the gains are though ..??

Of course, it would be interesting to compare the R3s against speakers of B&W, Triangle, ATC, Focal, .. which put lots of efforts in their midband chassis - but may have their own compromises where the R3s shine?

DaveinSM's picture

Yeah, but 8 db is a LOT, and 3khz is in the range in which human hearing is supposedly most sensitive. It may be a relatively narrow and deep dip, but it’s not a notch either, significantly suppressing the 2khz to 5khz regions up to 3db or more, according to the quasi anechoic response graph, fig. 5. There isn’t an in-room response test, and they even tried reversing polarity on the tweeter unit to fix it, with no success.

Considering many much less expensive designs don’t have this irregularity at all and are objectively much more neutral in frequency response, this is troubling at best.

The test sample was given more than generous leeway in accounting for this flaw, but at least this speaker’s colorations were acknowledged in the listening notes. Kudos to Stereophile for providing this test data anyway, despite what I can imagine is some trepidation about how some prospective buyers could interpret the results. That 8db dip would frankly scare me away.

supamark's picture

-8dB and over an octave wide is not a relatively narrow dip, especially when it's in the area to which human hearing is most sensitive. It does not fill in to the sides. It's audible and JCA said as much when he said the R3 sounded relaxed. It will literally suppress the vocals in most music. The R8 also has a weird frequency response but nothing so bad as the R3.

It's a flawed design, whether that flaw bothers you is a different question; but it's flawed none the less and at a price point where it is not acceptable to me. YMMV.

Kal Rubinson's picture

The gains are in the increased bass capabilities due to the manufacturer's decision to stick to 3 drivers. Their choice was to offer more extended and/or potent bass using two woofers versus the option to optimize the midrange with a dedicated driver. Potential buyers have that same choice.

hb72's picture

..quite obvious, actually :) and JA describes the R3s as a well rounded product with meat on the bones.

Trondi's picture

… that is exactly my experience with these Audiovectors. The extended bass is wonderful, not bloated or exaggerated but just there, right down deep, without the need for a sub imo. The top end is a delight too.

Okay, now to that notch. It is a gentle listen at those frequencies, giving you a wide but ‘rear stalls’ audio experience. This is not a bad sound at all, but one that I like to improve with DSP.

I use a NAD M33 with Dirac Live - full range - and I am able to incrementally lift the notch. As I do this my audio listening position changes and I appear to get ‘closer and closer to the stage’. With a near flattened target curve I appear to end up in the middle of the musicians. Astonishing.

My hifi buddy prefers DL switched off in my listening room. He is a classical music enthusiast and enjoys being ‘rear stalls’. I prefer to adjust DL to get to the ‘conductors podium audio experience’. For me this is the sweet spot with all types of music that I listen to.

Other than that these were the only speakers I demoed that my wife didn’t roll her eyes and grit her teeth on first appraisal. They are easy on the eye.

monkeybouie's picture

Sibilance, a side effect of close miking a vocalist, will generally be centered around 6-8kHz. That being the case, their filter is way off the mark.

Also, instantiating a filter like that will globally effect all the music that will be played by that loudspeaker seems a very strange choice.

As Pee-wee Herman once said, “I meant to do that.“