This review page is supported in part by the sponsors whose ad banners are displayed below

Having heard the Weiss/Klangwerk combo at many shows, I knew its pedigree. When Markus and Alain delivered the review pair, first impressions weren't that favorable however. As had been the initial case with Boenicke's B10 and German Physiks' HRS-120, they sounded bassy, dark, opaque and shy on harmonic lucidity. I thought it likely because Markus had set up his Ella wide just as I tend to do. Due to his sidefiring mid/woofers—the B10 does the same with dual 10ers, the HRS-120 uses a true 360° omni—I figured this unduly intensified interaction with my room's midband behavior. I assumed that insufficient sidewall distances impacted overall clarity by compacting the time differential between direct and reflected sound. Instead of allowing my ear/brain to separate main event and echo into two things, they summed as one. This created undue darkness and thickness.


Moving the speakers inward in steps toward the central throw rug atop our sisal mats lifted a lot of that heaviness whilst also narrowing the soundstage. But even at its best the sound still lacked top-end energy. Whilst the built-in electronics afford some bass adjustments, they don't for the treble. The baffle angle also had the tweeter fire well above my ears. Due to preference and design goal Markus further crossed speaker axes in front of my seat. I thus heard the tweeter quite off-axis horizontally and a bit less so vertically. Standing up altered this balance a bit to suggest I either sit higher or angle Ella forward to reduce baffle rake. Propping up the plinth's rear with two small risers verified that I wanted to hear more of the tweeter to overcome prominent midband energy. As Markus' subsequent MLSSA measurement confirmed, my room is well damped up to ~500Hz. Above that it gets more reverberant into the presence region before becoming absorptive again.


Now he explained that the trim pots of Relec's all-pass filters do allow for treble modifications, just not to the end user. Though simple if you know where and how, he didn't want to perform those in situ. He needed to be in his shop to properly calibrate these changed settings with measurement gear.


A few days later he returned with the 'retuned' boxes joking they now measured very close to Alain Roux's very first Ella prototype straight out of the anechoic chamber many years back. Set up in the same final spot as before, Markus had indeed done it. Now there was more energy and sparkle in the top end even if the sound in general still felt more muted, dry and matte than I'm used to. Markus however was perfectly happy. He confirmed it's what he aims for to avoid any chances at aggression.


On another critique—Ella's plinth has no spikes to account for uneven flooring—Markus has an easier solution. He offers an optional €260/pr black steel sub plinth with micro spikes and an identically shaped foot print that's just a tad smaller. It thus recedes invisibly underneath the main plinth and to the eye floats the speaker*.


Markus admitted that this was a first for him to retune Ella's upper tonal balance. He stressed that in narrower rooms he routinely sets her up directly against the side wall and without any issue. His entire design approach revolves around controlled directivity for very linear broad off-axis response. His installations are to make good consistent sound for multiple listeners. It's the opposite of Jim Smith's 'why to be sour on a wide sweet spot' credo. Markus explained that even though Ella 3 might look like mostly an omni, she behaves less so than you'd think.

* There's also the €300 (+ cables) A-Link, a dual audio line driver interface from PSI "to adapt RCA-out preamps with typically low output voltage to our amp modules whose balanced inputs need rather high voltage for saturation."

At that 'not omni' tech talk juncture he lost me completely. Never mind. His surgical adjustments had successfully compensated for my room's livelier midband reflections but more absorptive treble. It restored suppressed energy in the upper bands. If you're curious about more of Klangwerk's design goals, I point you at Gradient's Helsinki 1.5. It's a speaker Markus appreciates a lot. Its propaganda reads "tailored to produce the most beautiful sound in all environments from home to lounge bars and hotel lobbies... easy to position, delivers a very open wide soundstage everywhere to the room." Having measured the Gradient, Markus grants that on consistent off-axis behavior its legendary Finnish designer Jorma Salmi exceeds what he himself has managed for the Ella 3. But he finds its dipole bass more room-dependant to prefer his own active down-ported solution.


Now we encounter how consistent applies itself to the Ella 3. I wager a guess. In it she probably mirrors other active speakers with pro-audio genes like PMC's compact AML2 below whose progenitor I reviewed. That's to say, the marriage of precision-tuned amps and drivers with active filters makes fundamental seasoning to taste rather less effective.



Aside from the LF adjustments to eliminate/minimize room issues, this sound is mostly locked in. Preamp and source swaps telegraph less. This reiterates in no uncertain terms that the sonically most critical interface really is the amp/speaker junction. Here that's etched in stone. A Nagra Jazz tube versus Esoteric C-03 transistor preamp or Metrum Hex vs. AURALiC Vega DAC have less flavor effects than they do with passive speakers.

Unlike Ella 3, this PMC monitor and various Genelec variants add the HF tilt control for easy user adjustments which I think the Ella 3 ought to have too but doesn't.
The obvious plus is greater predictability. If you liked Ella 3 in a shop or show provided the demonstrator set her up properly, you're bound to like her at home as she'll sound more or less the same. The minus is less shift potential. The usual audiophile game of nip'ntuck with component substitutions is rather less effective. To my thinking and admittedly limited exposure this would seem part of the active speaker credo. To a more defined extent than the far more moving target of passive boxes, the sound is what it is.

Which now demands a description of what this sound is.