Sonus Faber Lumina III loudspeaker

Time for some towers. In recent months, a succession of standmount speakers has passed through my listening room: GoldenEar BRXes, Bowers & Wilkins 705 Signatures, Marten Oscar Duos, original KEF LS50s, and the new LS50 Metas. All these loudspeakers sounded excellent, though different from one another. I felt that a floorstanding loudspeaker would make for an interesting change.

Italian company Sonus Faber is primarily known for its high-performance, relatively high-priced designs, like the $130,000/pair Aida that Michael Fremer reviewed in October 2018 or the $15,900/pair Guarneri Tradition, which I reviewed in June 2018. In the fall of 2020, Sonus Faber announced an affordable series, the Luminas. I asked for a pair of the Lumina III towers, which cost $2199/pair, for review.

The Lumina III
Sonus Faber explains that the word "Lumina"—Latin for "to brighten" or "to illuminate"—is a contraction formed from "Luxury Sound Experience" (LU), "Minimalist Design" (MIN), and "Natural Sound" (NA). The "Natural Sound" will have to wait for the results of my auditioning, but with respect to "Luxury Sound Experience," the Lumina III is a slim, elegant tower standing 38" tall on its spikes. The spikes are necessary to provide clearance for the reflex port, which fires downward from the speaker's base. (Metal discs are provided for those who don't want the spikes to damage their floors.) The enclosure is clad in a black leather wrap, while the front baffle is a 10mm-thick sheet of plywood finished in dark brown, called wengè, with inlaid horizontal maple stripes. A cloth-over-plastic-frame grille is supplied.

321sonus.1

The four drive-units are mounted vertically in-line on the baffle, each outlined with chrome. At the top is Sonus Faber's Damped Apex Dome (DAD) tweeter, the same unit featured in the company's Sonetto series. The point of a tiny damping cone is held against the center of the hand-coated silk dome by a vertical bracket, which is said to optimize the driver's dispersion.

The 5" midrange unit, which operates between 350Hz and 3.5kHz, is also the same as that used in the Sonetto speakers. It uses a cone formed from cellulose pulp blended with other natural fibers and a diecast basket described by the manufacturer with the phrase "ultra-free compression." The two 5" woofers, which are new in the Lumina speakers, feature paper-pulp cones, rubber-roll surrounds, and those "ultra-free compression" baskets.

"Minimalist Design" refers to the speaker's physical design, to its claimed ease of setup, and to the crossover, which uses Sonus Faber's "Paracross topology," described in my Guarneri Tradition review. The crossover features Sonus Faber–branded, "high-quality" capacitors. Electrical connection is via two pairs of nickel-plated binding posts inset on the rear panel, allowing for biwiring.

321sonus.close

Setup
I initially placed the Sonus Fabers where standmounted speakers have worked best in my slightly asymmetrical room, but the midbass was lighter in weight than I expected, and the upper bass was somewhat exaggerated. After some experimentation, I ended up with the Lumina IIIs 7" closer to the wall behind them, which brought the upper- and low bass into better balance; the two steps up to a vestibule behind the right-hand speaker didn't allow me to move the Sonus Fabers any closer. Each speaker's front baffle was 70" from the wall behind it. The woofers of the left-hand Lumina III were 31" from the LPs that line the nearest sidewall; the right-hand speaker's woofers were 40" from the bookshelves that line its sidewall. I didn't use the grilles.

Each speaker was supported on its spikes, which meant that the opening of the downward-firing port was about three-quarters of an inch from the carpet beneath. My ears were just above the Lumina IIIs' tweeters, which were 35" from the floor.

321sonus.bac

The music source was mainly the Grimm MU1 streamer that I reviewed in the March 2021 issue, sending upsampled AES/EBU data to an MBL N31 DAC (although I also used the N31 to play CDs and dual-layer SACDs). Amplification was the Pass Labs XP-32 preamplifier, which I also reviewed in the March issue, and a pair of Parasound Halo JC 1+ monoblocks.

321sonus.edchoicecd

Listening
Once I had finalized the speakers' positions, I started my serious listening. As always, I started my critical auditioning of the Lumina IIIs with the test-tone files I created for my Editor's Choice CD (Stereophile STPH016-2). The Lumina IIIs were initially toed in to the listening position. With the dual-mono pink noise track, the treble balance was reminiscent of the Bowers & Wilkins 705 Signature's: a little accentuated in the top two octaves, especially when compared with the KEFs. I increased the toe-in a little so that the speaker axes crossed just in front of me; now the high frequencies sounded in better balance with the midrange, though still a little high in level, especially if I sat upright so I could see the tops of the speakers. While the pink noise had a touch of "character" in the lower midrange, the image of the pink noise was narrow and stable, with no "splashing" to the sides at any frequency.

COMPANY INFO
Sonus Faber S.p.A.
US distributor: Sumiko
655 Wedgwood Rd. North, Suite 115
Maple Grove, MN 55311
(510) 843-4500
ARTICLE CONTENTS

COMMENTS
georgehifi's picture

Very affordable at $2199/pair but! you'll need a good brute amp.

"The EPDR (footnote 1) is 1.5–1.6 ohms between 96 and 120Hz and 1.6 ohms between 660Hz and 810Hz, with a minimum value of 1.25 ohms at 111Hz."

This to me says no tubes allowed, and even some mosfets.
Looks to me like a good complimentary BJT amp with solid power supply to get the best out of these girls EG: Gryphon, JC Halo's, Krells etc etc.

Cheers George

Ortofan's picture

... not be required to, as you say, "get the best out of these girls."

Note that, according to JA1's measurement, the Sonus Faber Lumina III has a relatively high sensitivity of 91dB(B)/2.83V/m.

The $599 Parasound NewClassic 2125 v.2 two-channel power amp has a rated output power of 225 watts/ch into 2Ω, when the load switch is set to the 2-3Ω position.
https://www.parasound.com/2125-v2.php

If you want more power, then the $999 Parasound NewClassic 2250 v.2 two-channel power amp has a rated output power of 400 watts/ch into 2Ω, when the load switch is set to the 2-3Ω position.
https://www.parasound.com/2250-v2.php

https://www.audioadvisor.com/products.asp?dept=8#/filter:brand:Parasound

Neither amp is quite as capable as the Halo JC1+, with its 1300 watt into 2Ω rating, but keep in mind that the manufacturer's suggested power rating for these speakers is 50-250 watts.
https://www.sonusfaber.com/en/products/lumina-iii/

LTig's picture

... I'd suggest to first try a decent class D power amp. They can deliver lots of power for small money and new models are really quite good. Might be all one needs to drive this affordable speaker.

tonykaz's picture

Once again Mr.JA's analysis shows how & why he's been my Audiophile of the Year & Reviewer of the Year since the 1980s.

Mr.HR's & Mr.Micallef's prose float them up to my top levels as does the Audiophiliac's production level and insight.

Tyll would have to be in this group if he wasn't retired.

These four individuals are the clearest & most transparent portal into DIY Home music reproduction we've ever had.

These 4 are NOT shameless gear promoters.

Thank You

Tony in Venice Florida

Anon1's picture

“ This graph suggests that the optimal blend of all three outputs will occur just below the tweeter axis.”

“ My ears were just above the Lumina IIIs' tweeters, which were 35" from the floor.”

Ummmm....

“ —but the Lumina IIIs have the highest high-treble output.”

Ummmm....below 16kHz. There are some of us that can still hear above that level. The last time I checked 17-20 kHz is higher than 15kHz.

tonykaz's picture

You may be onto something.

I don't "frequent" stores, do I ? What does "frequent" suggest ? Are we discussing frequency of visits or frequency of ear sensitivity ???

I can't seem to resolve the enigma you present, I might be sorry but probably not.

I'm about 8 db down at 17-20 kHz. in both ears, I don't feel deficient. Audiologists tell me that everyone's hearing droops like mine once they hit a certain age ( in the 60s ), no exceptions .

Tony in Venice Florida

ps. when was the "last time you checked" ? are you being sarcastic ? ( which is disguised anger )

ps.2.). This transducer system has an affordable cost!, which is the point behind all this discussion, isn't it ?

jimtavegia's picture

At 3" taller than a yard stick I was surprised at how big they seemed until I looked at the measurements. Nicely finished speaker system.

Jack L's picture

...........cardioid mikes arranged in an ORTF" quoted J.A.

For chamber music stereo recording, have you tried the Blumlein 2xfigue-8 "7-mic" technique which reportedly produces pretty pure stereo imaging of the assembly with minimum reverberation !

Jack L

John Atkinson's picture
Jack L wrote:
For chamber music stereo recording, have you tried the Blumlein 2xfigue-8 "7-mic" technique which reportedly produces pretty pure stereo imaging of the assembly with minimum reverberation !

I used to be a Blumlein purist but found that finding the optimal site for the microphones' position was close to impossible, especially under the pressure of live performance. Crossed, coincident figure-8 miking also lacks lower-frequency "bloom." I use an ORTF pair of DPA or Neumann cardioids flanked with outrigger DPA omnis. This allows me to optimize the balance and imaging in postproduction

John Atkinson
Technical Editor, Stereophile

Jack L's picture

Hi John

The issue just came across me when I listened to BBC Prom in the Royal Albert Hall performing Richard Strauss Also sprach Zarathustra around 15 years ago streaming on YouTube this afternoon (my day off).

The kettle drums beating immediately after the first starting low low organ bass notes, sounded soft & blurred ! Same music performances on YouTube, like one conducted by Zubin Mehta in another concert hall in June 11, 2014, delivered sharp & punchy kettle drums beat. Big contrary !

The best kettle drumming effect of this music piece so far I auditioned, is delivered by my DGG Resonance label LP (2535 209): clean, sharp, punchy & powerful. It was performed by Boston Symphony Orchestra, conducted by William Steinberg in 1971. I got 5 LP of the same title, only this LP really satisfied my very picky ears. Surprising !!

The performance started with some low low crawling organ bass notes followed immediately with sharp & powerful kettle drum beats.

This short music piece is what I use to test the transient response of a HIFi system.

My question is: why this BBC Prom performance in the Royal Albert Hall, London sounded so lousy as said above. Was it due to its miking system there ??

Hopefully you could enlighten me on this with your profound live recording experience.

Jack L

John Atkinson's picture
Jack L wrote:
My question is: why this BBC Prom performance in the Royal Albert Hall, London sounded so lousy as said above. Was it due to its miking system there ??

Most likely the miking, or perhaps the sound was mixed for the television broadcast. Will check out the YouTube video of the performance.

John Atkinson
Technical Editor, Stereophile

Jack L's picture

Hi John

Also sprach Zarathustra performed by Gustav Mahler Youth Orchestra.

Jack

Jack L's picture

....... for upper-midrange issues. " quoted J.A.

There are way way so many mic. arrangements involving quite a few mics to record a grand piano performance right. All depends on the recording engineers' choice.

I would use a few piano recordings to average it out !

Jack L

Jack L's picture

....... I don't recommend this speaker for use with source components that are themselves too forward in the treble," quoted J.A.

Have you given a FAIR audition to this Lumina III which provides bi-wiring connection at the back ???? It is there for one single reason: to improve the sound as it costs the loudspeaker maker more to design/build a bi-wireable X-over network !!!

FYI, my KEF 2-way standspeakers got only standard 2-terminal connection like most most brandname loudspeakers, irrespective of pricing. Knowing the technical merit of BI-WIRING, I converted it to bi-wiring many years back.

I replaced its lousy micky-mouse like factory-built X-over board with my customer design/built BI-WIRING X-over board with high-power polypropylene caps & wire-wound resistors. Also replaced ALL cheapie skinny wires inside with AWG#12 thick cables. No joke!

It sounds so so much better in term of high-low frequency balance, transient attack with clean smooth bass & silky soothing highs. Like a new reborn !!

Listening is believing

Jack L

Charles E Flynn's picture

Anyone searching for information about this topic should also search for "biwiring".

https://www.qacoustics.co.uk/blog/2016/06/08/bi-wiring-speakers-exploration-benefits/

Jack L's picture

Hi

The technical merit of bi-wiring is easy to understand.

Being an audio handyman, I decided to prove it the very hard way - to convert my KEF 2-way loudspeakers to bi-wiring. I've succeeded to improve its sound substantially for many years now !!!

I would challenge any nayers who use their mouths more than their ears.

Listening is believing

Jack L

Charles E Flynn's picture

Thanks for your reply.

The naysayers include a few manufacturers.

I have always been puzzled about why more attention is not paid to this issue. I wonder if KEF has made an experimental sample of the LS50 Meta with biwirable inputs, just to see if they can measure any difference.

Glotz's picture

I agree with you, but you make a challenge without explaining your position first.

Can you give readers reasons please?

hb72's picture

see q-acoustics link above.

I have gone early to bi-wiring, went back to single wiring briefly and was puzzled by the loss of quality (less clear, and various levels of sibilance.. speakers, cables, amps will influence the results, no doubt).
Sound issues, and particularly any tendency to pronounced sibilance might be related to sub-optimal connection: if the single speaker lead is connected "across" the terminals, i.e. plus to treble and minus to bass, both, treble and bass see one branch of jumpers which eliminates sibilance allegedly, see also related info at audiquest.com.

anyway - I see 2 reasons for improved sound assuming a 6db crossover 2-way speaker:
1) separating bass from treble signals into two separate cables does not change the situation of electric field within the cable (both treble and bass see the complete frequency range of the electric field created by the voltage signal), but in terms of magnetic field (caused by the instantaneous current) there is quite a difference: the relatively large low frequency current driving the speaker, and current resulting from EM-back force are not present in the high frequency side within the cable, which would otherwise meddle with the wires and introduce noise & distortion to some degree (note better cables are often very rigid, often hyperlitz, thus reducing the possibility for relative motion between conductors).
2) once EM-back force (created by a deflected bass speaker membrane, then pushed back to neutral position by the spring) induces the emf voltage via coil system, this voltage is practically shorted by the amp output terminals (the lower the output impedance the better). Ideally. Within the frequency overlap between treble and bass in the crossover, such back emf may find its way from bass to treble and cause distortion in case of single wiring, while in the bi-wiring case this way is blocked. The only way this may happen is via speaker cables, however, before back emf voltage reaches the treble stage, it is simply shortened at the node where treble and bass branches are connected, ie at the amplifier's output stage, and only a very very small residual component may affect the treble unit.

hope this was not too confusing.

Glotz's picture

Thanks man. I also like networked-cable systems like MIT, but mine are long in the tooth vs. the newer single wire cabling I am using... I may buy another run and rewire the speakers.

Jack L's picture

Hi

Yes, my KEF 2-way standspeakers got 6dB 1st order Butterworth X'over. It's a simple network with the unique advantage of minimum impedance variation to the driving amp & less demanding on the tolerance of the network parts values vs 2nd or 3rd orders.

Many years back I upgraded a pair of vintage KLH 3-way standspeakers which sounded flat & monotonous driven by a Quad 303 power amp for a audio fan by modifying it's complex 12dB 2nd order single wired network to tri-wiring. With same Quad power amp driving, the KLH sounded much livelier & musically appealing. Saved my fan changing amp & loudspeakers !!.

Bi-wiring & tr-wiring, if done PROPERLY, can be a musical game changer !!!!

Listening is believing

Jack L

Jack L's picture

Hi

I made no challenge to nobody except pointing out the fact that Lumina III already come with bi-wiring feature which would improve its sound for good technical reason. Or else the loudspeaker maker would not have spent more to make it bi-wireable.

I've also established my "position" loud & clear that I am a die-hard bi-wiring advocate basing my hands-on & ears-on experience in bi-wiring my KEF standspeakers.

I don't hearsay & talk like those naysayers. I have done it right !!

Listening is believing

Jack L

PS: bi-wiring & tri-wirng get their technical & sonically merits. Please Google it to know more about it. In fact, quite a few brand-name loudspeakers come with bi-wireable terminals, including bookselvers as shown in Sterophile product reviews.

Glotz's picture

The Magneplanar 1.6qr's came with bi-wiring provisions as well as past Mirage 7si's I owned. MIT bi-wiring cable was used as well.

This isn't a challenge to you. You did state "I would challenge any nayers who use their mouths more than their ears". I was just asking you to share why, in a positive way.

The other poster in response gave some great answers either way.

Jack L's picture

You like it better ?

Jack

Jack L's picture

You like it better ?

Jack

Glotz's picture

I had it originally bi-wired but then went single for a year or two..

I put it back today and oh yeah! Hard to ignore that discreetly-rendered sound! It was a good reminder. Keeping it that way!

I just wish I could do the 1.7's now... But they use a series X-over. Oh well. Thanks!

Jack L's picture

Now You've upgraded yr music for no money !

Congratulations !

Listening is relieving

Jack

hb72's picture

it happened a couple of times to me, following some tiny changes I made to the system, resulting e.g. in a slightly better balanced frequency response overall or similar, that some weeks later I noticed that I then was not listening to music that often anymore, and even felt that something was missing in live, one source of joy actually. Music wasn't that much fun or, more importantly, *moving* anymore. Then I undid the last change, slightly more uneven balance was back, but so was musical expression, touch and life in music, and more reasons to spend time listening.

Glotz's picture

and I do use that as a guidepost as well.

Charles E Flynn's picture

From https://www.vandersteen.com/support/faqs :

What is bi-wiring and what are the advantages?

[excerpts from long answer to the question]

Bi-wiring uses two separate sets of speaker cables to connect a single pair of loudspeakers to an amplifier. Coupled with a crossover designed specifically for bi-wiring, it offers many of the advantages of bi-amplifying the speakers with two separate amplifiers without the cost and complexity of two amplifiers.

[later]

The benefits of bi-wiring are most obvious in the midrange and treble. The low-current signal to the midrange and tweeter drivers does not have to travel on the same wire as the high-current woofer signal. The field fluctuations and signal regeneration of the high-current low-frequencies are prevented from distorting or masking the low-current high-frequencies. The back EMF (Electro-Mechanical Force) from the large woofer cannot affect the small-signal upper frequencies since they do not share the same wires.

The effects of bi-wiring are not subtle. The improvements are large enough that a bi-wire set of moderately priced cable will usually sound better than a single run of more expensive cable.

All the cables in a bi-wire set must be the same. There is often great temptation to use a wire known for good bass response on the woofer inputs and a different wire known for good treble response on the midrange/tweeter inputs. This will cause the different sonic characteristics of the two wires in the middle frequencies to interfere with the proper blending of the woofer and midrange driver through the crossover point. The consistency of the sound will be severely affected as the different sounding woofer and midrange drivers conflict with each other in the frequency range where our ears are most sensitive to sonic anomalies. The disappointing result is a vague image, a lack of transparency through the midrange and lower treble and a loss of detail and clarity.

Some of the benefits of bi-wiring are from the physical separation of the high-current bass and low-current midrange/tweeter wires. So-called bi-wire cables that combine the wires in one sheath do not offer the full advantages of true bi-wiring although they may be an excellent choice for mono-wiring the speakers.

Glotz's picture

N/T

Jack L's picture

Hi Charles

Bingo !

First off, thanks for yr bringing up this very comprehensive info. for our readers. I can't agree more !!

Yes, All bi-wiring & tri-wiring cables MUST be the same structure & should be run SEPARATELY. Do not bundle them together for obvious technical reasons.

A simple analogy: traffic on a highway: one lane exclusively for heavy trucks & other lanes for lighter vehicles. Simple logic, no rocket science.

I only use low cost no-name AWG#12 256strand oxygen-free pure copper cable, one for the tweeter & one for the woofer of my KEF 2-way standspeakers. These two cables are of SAME length, running parallel to each other 3" apart, suspended from the fully-carpetted concrete floor by semi-rigid plastic foam blocks, each 6" apart.
So no airborne as well as structural vibration & EMI/RFI noises from the concrete flooring underneath.

One very critical feature I've installed since day one many years back years which has improved the sound substantially better. Please don't copy it unless you were a "Jack of all audio trades" like yours truly & yr expensive loudspeakers were already out of warranty.

OUTboard X-over network : I removed my design/built large-sized bi-wired PCB housed in a NON-metallic box out of the loudspeaker cabinet & relocated it right to the back of my tube power amp.
So the bi-wiring cables are only about a foot run hooked up to my tube power amp output terminals.

The idea is the ensure intermodulation free way for both high (over 1KHz) & low (below 1KHZ, X-over frequency of the X-over network) frequency signals to the loudspeaker driver units terminals.

Such purist out-board bi-wiring scheme was also done by some brand-name loudspeakers. I still recall B&W did market it's DM1 special version many years back. Yet I did not copy it as I outboards mine a few years earlier than B&W.

Great minds think alike ????

Listening is believing

Jack L

Ortofan's picture

... do not appear to be an advocate of bi-wiring.
https://www.dynaudio.com/dynaudio-academy/2017/august/how-to-set-up-your-speakers

Jack L's picture

Hi

Here is what I jotted down what Dyaudio stated there:

"..We'd rather put the effort into making our X'over as perfect as they can be.... Only need to buy one pair of speaker cables. "

I'll be very interested to know how "perfect" would be Dyaudio's X'over networks technically & physically. Make sure they are better than my home brews

Jack L

Charles E Flynn's picture

If the theory of bi-wiring is true, regardless of how excellent the crossover network is, the loudspeaker's output would be even better if it were bi-wired.

Jack L's picture

Hi

Bi-wiring will improve the sound of the "perfect" X'overs by eliminating intermodulation of high & low frequency music signals passing through the same cable.

Jack

xxthewhoxx's picture

I am jumping at the chance to move into the Sonus Faber lineup with the Lumina III. Any advice if the Naim Uniti Atom would be a good match for the towers? Considering also the Nova if given the right pre-owned opportunity. The goal is to keep it simple with an all-in-one amp/DAC/streamer as I have really enjoyed the practicality of owning the KEF LS50 wireless speakers in a second listening room.

X