Elac Adante AF-61 loudspeaker Measurements

Sidebar 3: Measurements

I used DRA Labs' MLSSA system and a calibrated DPA 4006 microphone to measure the Elac Adante AF-61's frequency response in the farfield, and an Earthworks QTC-40 for the nearfield responses. (I didn't use the grilles.) My estimate of the Adante AF-61's voltage sensitivity was 86.1dB/2.83V/m, slightly below the specified 87dB/2.83V/m. The specified nominal impedance is 6 ohms. Fig.1 shows that the impedance magnitude remains above 6 ohms for much of the audioband, with minimum values of 4.75 ohms at 92Hz, 5.25 ohms at 240Hz, and 5.1 ohms at 9.1k ohms. However, there is a current-hungry combination of 5.3 ohms magnitude and electrical phase angle of –43° at 75Hz (fig.1). This speaker would work best with an amplifier that is comfortable driving 4 ohm loads.

118Elacfig1.jpg

Fig.1 Elac Adante AF-61, electrical impedance (solid) and phase (dashed) (2 ohms/vertical div.).

The traces in fig.1 are free from the small wrinkles that would imply the presence of cabinet-wall resonances. However, I did find a mode at 333Hz on the sidewalls level with the woofers (fig.2), and this was also present on the rear and top panels. This is high enough in Q, and sufficiently low in level, that it shouldn't give rise to any coloration.

118Elacfig2.jpg

Fig.2 Elac Adante AF-61, cumulative spectral-decay plot calculated from output of accelerometer fastened to center of sidewall level with middle passive radiator (MLS driving voltage to speaker, 7.55V; measurement bandwidth, 2kHz).

There are several impedance peaks in the bass in fig.1, these due to the AF-61's unusual low-frequency alignment, in which each of three internally mounted woofers fires into a ported cavity coupled to the outside world with a passive radiator (footnote 1). The three radiators behave identically, and the sum of their nearfield responses is shown as the red trace in fig.3. This is crossed over to the midrange unit, whose nearfield response is shown as the blue trace, at the specified 200Hz with what appear to be symmetrical fourth-order, 24dB/octave slopes. The lower-frequency rolloff is also fourth-order, with an approximate –6dB frequency of 40Hz—higher than I would have expected for such a large loudspeaker.

118Elacfig3.jpg

Fig.3 Elac Adante AF-61, anechoic response on tweeter axis at 50", averaged across 30° horizontal window and corrected for microphone response, with nearfield responses of passive radiators (red), midrange unit (blue), and their complex sum, respectively plotted below 375Hz, 875Hz, and 300Hz.

The black trace above 300Hz in fig.3 shows the farfield output of the Adante AF-61's coaxial midrange/treble unit on the tweeter axis. Other than a slight lack of energy in the mid-treble and a small peak centered on 10kHz, the Elac's response is superbly flat and even. The plot of the Adante AF-61's lateral dispersion, normalized to the tweeter-axis response (fig.4), reveals that these on-axis features disappear to the speaker's sides, meaning that, as Andrew Jones told TJN, the perceived treble balance in a normal-size room will be neutral. As anticipated, the tweeter begins to become directional above 12kHz, which might make the AF-61 sound a little airless in large or overdamped rooms, as TJN found in his room. In the vertical plane (fig.5), the Adante's even tweeter-axis balance is maintained across a wide window—just as well, considering that the tweeter is a high 46" above the floor.

118Elacfig4.jpg

Fig.4 Elac Adante AF-61, lateral response family at 50", normalized to response on tweeter axis, from back to front: differences in response 90–5° off axis, reference response, differences in response 5–90° off axis.

118Elacfig5.jpg

Fig.5 Elac Adante AF-61, vertical response family at 50", normalized to response on tweeter axis, from back to front: differences in response 15–5° above axis, reference response, differences in response 5–15° below axis.

In the time domain, the Adante AF-61's step response on the tweeter axis (fig.6) indicates that all of its drive-units are connected in positive acoustic polarity, with the tweeter's output arriving at the microphone before the midrange unit's, which in turn arrives before that of the passive radiators. The output of each unit smoothly blends with that of the next lower in frequency, suggesting optimal crossover design. The cumulative spectral-decay plot on the tweeter axis (fig.7) is generally clean.

118Elacfig6.jpg

Fig.6 Elac Adante AF-61, step response on tweeter axis at 50" (5ms time window, 30kHz bandwidth).

1118Elacfig7.jpg

Fig.7 Elac Adante AF-61, cumulative spectral-decay plot on tweeter axis at 50" (0.15ms risetime).

The Elac Adante AF-61's measured performance reveals excellent speaker engineering, even if, as Tom Norton found, the speaker doesn't extend as low in the bass as you might expect.—John Atkinson



Footnote 1: See my video interview with the AF-61's designer, Andrew Jones here.
COMPANY INFO
Elac Americas
11145 Knott Avenue, Suites E & F
Cypress, CA 90630
(888) 541-0996
ARTICLE CONTENTS

COMMENTS
Bogolu Haranath's picture

May be TJN or JA or HR could review the new Elac Navis self-powered loud speakers?:-) .........

Moshe Ufnik's picture

"The company's current flagship line, Adante ..."

You haven't been doing your homework.
What about Elac's FS409 or the much higher-priced 500 series?
And did you forget their actual flagships - the Concerto and its little sibling Concerto M?

prerich45's picture

I believe Mr Norton was referring to Elac's new direction, as the 500/400 series and the Concerto lines were not designed by Andrew Jones. I've always wondered, as Andrew puts his stamp on the company, would he gradually design speakers to take the place of their old flagship lines?

Bogolu Haranath's picture

The Elac Flag-ship Concentro speakers (almost) look like KEF Blade loudspeakers .......... May be JA could review those Elac Concentro speakers? :-) .................

audio1321's picture

Hello, great article! May I suggest when using the OmniMic System to perform the bass measurements by placing the OmniMic on the floor and splicing the results with the high frequency measurements. This is very accurate considering access to a Anechoic Chamber is not feasible. This is because the reader will not be listening to the system in your space, we don't care to see the room's reverberation in the low end. Thanks :)

Ortofan's picture

... KEF R11 speakers - also priced at $5K/pr.
Or, for just $3,600/pr., one can now get the discontinued R900.

Bogolu Haranath's picture

If someone is gonna use Elac sub-woofer(s) anyway, they may be better off buying the Adante AS-61 bookshelf speakers and save some money ............ AS-61 are priced at $2000 - $2500/pair ......... AS-61 was favorably reviewed by TJN for Sound & Vision :-) ..........

jazzman1040's picture

Seeing that the Adante's have been out for the better part of a year, I wouldn't hold my breath waiting for a review of the new active ELAC's. How a so-called leading audio magazine can afford to be a year late when reviewing a highly anticipated product is beyond me. Btw, Stereophile, you should ditch CAPTCH immediately. It's a ridiculous verification system that took me about 50 clicks to verify I'm not a "robot".

Bogolu Haranath's picture

I got through .......... and I'm a robot ....... Hasta la vista, baby :-) ............

"I Robot" ................ The Alan Parsons Project :-) .............

John Atkinson's picture
jazzman1040 wrote:
How a so-called leading audio magazine can afford to be a year late when reviewing a highly anticipated product is beyond me.

We depend on companies being willing to loan us review samples, and it is getting increasingly difficult to to get samples from some companies. Although I doubt that this was the case with Elac, I suspect that this is often due to the fact that our reviews always include measurements. Companies appear to want a slew of positive non-measurement reviews elsewhere before they will send us samples for review.

John Atkinson

Editor, Stereophile

Bogolu Haranath's picture

Thank you JA for the measurements .......... JA is doing great service to the audiophile community by doing the measurements ........... Defend ton territoire (Stand your ground) :-) ..............

jazzman1040's picture

John, appreciate the thoughtful reply. Sounds like a tough predicament. A suggestion, for products like these that have more of an audience than some of the higher end stuff you review that is bought by scant few folks, maybe you could do a series where you visit the actual buyer's home and conduct the review and measurements with them, bypassing the company altogether. Now that would be innovation and certainly be more authentic than getting a pair from the company!

Mrsnikoph78's picture

In that case, please let the manufacturers know that if they restrict my access to quality information about their products I'm much, much less likely to ever purchase one. I depend on getting a full view of such luxury products to save hassle and shipping costs prior to purchase. Given how impossible it is to demo every great product that gets released, one is forced to shop wherever they can go to actually hear something. That eliminates 99% of the products I discover only on the Internet for a guy like me.

BTW measurements have a relationship to the sound quality of a product but they don't totally define it (they can't, because not everything is or can be measured let alone interpreted). Why fear a glowing review coupled to measurements that might not be "perfect"? Kinda ridiculous.

Bogolu Haranath's picture

"I Won't Back Down" ............. Tom Petty :-) ............

Juhazi's picture

TN wrote "Since bass frequencies are radiated from nowhere else on the AF-61, such as a port, these measurements indicates the bass capability of the speaker itself, with no help from so-called room gain."

This is still bad - there will be very little help from room gain when output drops that fast. Room help is most beneficial for closed box woofers that eg. Magico uses in floorstanders.

I don't understand at all why they use 6th order bandpass with a floorstander. Cost of copper is not that high...

(Message EDITED because of a misconception that JA noticed)

John Atkinson's picture
Juhazi wrote:
TN wrote "Since bass frequencies are radiated from nowhere else on the AF-61, such as a port, these measurements indicates the bass capability of the speaker itself, with no help from so-called room gain."

This is wrong - there will always be room gain indoors...

Not with a nearfield measurement, which is what Tom Norton was discussing.

John Atkinson
Editor, Stereophile

Juhazi's picture

Oops you're right! I was reading too fast.

I EDITED ORIGINAL POST

motberg's picture

The measured response seems a perfect match for my small dedicated room.. but I was wondering about the effect of using these more near/mid-field, like around 2 meters or so.

I currently have monitors but I think the feeling of overall image size may be better with floor-standers.

Bogolu Haranath's picture

If you already have monitors, and if you are happy with their midrange and treble sound, it may be a good idea to spend the money on powered subwoofer(s) ......... Separate subwoofer(s) allow the flexibility of room placement .......... There are several powered subwoofers available today, which are not very expensive and come with built-in crossovers and output levels .......... Chances are the same company which makes your monitors, also makes matching subwoofers :-) ............

Christian Thomas's picture

Bandpass loading is a superb idea, and this is a particularly interesting embodiment of it. Its advantage over direct radiators is that the acoustic low pass filter created - at 200Hz here - removes a significant proportion of the distortion components generated in the voice coil and magnet assembly. Down at the bass end is where they really start to get going and in some speakers you will see some truly horrific figures. These can't be filtered out by the crossover because they happen after it. If distortion is generated in the cores of the inductors, this too will be filtered. There can also be a small pickup in efficiency, which means that the speakers need a little less excursion, and so again improves distortion.

In this example the ports also vent into the bandpass chamber - which is a totally brilliant idea - so any port noise generated is also subject to this filter and, since that happens significantly above 200Hz, it will get attenuated almost entirely. This also applies to unwanted signal from the rear enclosure. There really is almost no downside to using a bandpass enclosure except that it can be difficult to get it to work above about 250Hz, which can limit your choice of crossover frequency.

Yes, there is a delay in the system, but we know essentially what it is. It's a phase shift of somewhere around 90 degrees at 200Hz (in this instance) and it makes up part of the crossover. It is essentially what you would have got by doing it electrically. What is even prettier is that there is a similar complementary mechanical high pass filter in the loading of the midrange unit, so you can have both the mechanical and electrical parts of each unit's filter be symmetrical. All in all it's a tremendous advance in helping get rid of distortion.

hb72's picture

Thanks for very interesting and detailed explanation of the system. Is 90deg phase angle really the overall phase angle of passive membrane excursion vs voltage input?

Christian Thomas's picture

The 90 degrees is an educated guess at what the target might be for the acoustic/mechanical part of the overall filter. It's exactly the phase shift one gets at the -3dB frequency with a 2nd Order Butterworth. That phase shift varies from zero at 0Hz to 180 degrees at infinity - in a smooth, reversed, s-shape. If you choose a soggier filter shape like a Bessel, then you will get a little less phase shift.

It's probably a pretty good guess, given the measurements above which show near perfect 24dB/octave rolloffs. So from that I would guess that the target is a 4th Order Linkwitz-Riley on both sides, and this is borne out by the fact that they cross over at 1/2 output, or 6dB down. A 4th Order Linkwitz-Riley is made up of two cascaded 2nd Order Butterworths so I would expect both the mechanical and electrical parts (of both drivers) to each have a Q of 0.707, which is the definition of a Butterworth. What is important about this alignment is that the phase responses of bass and mid overlay each other, so the delay at every frequency is exactly the same in both units.

As a caveat it is possible that either practical considerations, or a desire for greater efficiency in the drivers (which keeps them cooler), could mean that the acoustic Qs are higher and that this is compensated for in the electrical part, resulting in very nearly identical responses. Still, 90 degrees is a pretty good estimate, if possibly a near approximation.

Bogolu Haranath's picture

Andrew Jones (Elac) is one of the very talented designers around .......... He is very well known for designs with great sound for not a lot of money ........... His various designs have been praised for their great sound over many years by numerous audio magazines/websites including Stereophile .......... When he comes up with a new design, it is worth paying attention .............

lozsgo's picture

How do these compare to similarly priced GoldenEar Triton Towers?

John Atkinson's picture
lozsgo wrote:
How do these compare to similarly priced GoldenEar Triton Towers?

Both are excellent speakers. The GoldenEars will offer more bass, the Elacs a more refined midrange and treble.

John Atkinson
Editor, Stereophile

Bogolu Haranath's picture

There is a good chance the Elac Navis floor-standing powered speakers will match the GoldenEar Triton Reference in the bass department (almost) ......Triton reference probably will go lower in bass .............. We are waiting for a review in Stereophile :-) ..........

Elac Navis powered speakers are priced at $2,000/pair for bookshelf, and $4,000/pair for floor-standers ....... and they don't need any external power amps and speaker wires ............. Interested? :-) ..............

Bogolu Haranath's picture

Who is better? .......... KEF or Elac? ............. JA could be the judge :-) .............

Relayer's picture

I can't help but think the loss of low frequency extension could be solved by getting rid of the 6.5' drivers and just using 3 8' divers for the bass. It would simplify box design and probably lose weight to boot. Footprint, of course, remains the same.

Bogolu Haranath's picture

Looks like the AF-61 has somewhat 'screwed up' over-engineering in the bass department :-) ...........

Another good example why measurements are important ............. Thank you JA and TJN ........
Many other audio magazines/websites don't do these kind of measurements ........

The new Elac Navis powered speakers appear to be the very common and very popular bass-reflex design ............

Bogolu Haranath's picture

Google search and Wikipedia have info about band pass filter, passive radiators, bass-reflex (ported), sealed box (acoustic suspension) and transmission line designs :-) ...........

Also, look for 'loudspeaker enclosures' in Wikipedia :-) .............

hb72's picture

Looking at the step response I see a considerably longer delay until the passive membrane produces the peak (2.5ms ca after tweeter), a delay that appears consistent with the high order (6th) of the bass unit to fade out on frequency extremes. Note other speakers do 1ms or less (2-ways are usually particularly good, but also the tidal audio akira 3-way + passive membranes!).
The AF-61 has actually never been reported of lagging behind in bass (on the contrary – to my surprise!); still it seems a safe assumption that a driver driving a passive membrane via reflex system might need some semi-cycles to excite the passive membrane to full amplitude (and also to stop it). And I understand that the ear & brain parts are not so sensitive to phase shifts (think of a piano: the perceived sound of a chord involving many keys is not dependent on the exact phases the piano keys are played).
But then I wonder, whether speakers which are coincidental or near-coincidental down to bass (e.g. active digitals ones with FIR) are perceived notably different also e.g. w.r.t. reproduction of bass drums and possibly so for their ability to better convey PRAT etc. than e.g. the AF-61 does. Or are such differences in step response not really audible? Or audible more w.r.t 3d stage rather than PRAT & snap.

@ John Atkinson: any thoughts?

Many thanks in advance

rom661's picture

This feels like a blast from the past. I still own an Amp 5 and was a dealer in the day. It's... good. Surround processor, old demo disks. I don't like to evaluate speakers for two channel setups with surround processors but maybe that's just me. I've never heard a processor as good as a quality two channel preamp. My point for writing though, is simply how much I felt like I'd slipped back into the 90's while reading it. Oh, and the comparison with KEF's, no doubt the 104/2's. Wow.

X