YG Acoustics Sonja 1.3 loudspeaker Page 2

One thing that impressed me about the Sonja 1.3 was that, once each module was in place, two long, threaded, horizontal rods were inserted from the rear, each screwing into three tapped feet on top of the lower module that mate with three threaded cones on the base of the upper module. Once these rods are in place, the modules are locked together, with no external signs of the join other than the seam. That the alignment of the threaded holes can be maintained over the length of the rod is a testament to the precision YGA gets from its CNC machines.

The Sonjas were not quite toed-in to the listening seat, and the 1.1 modules were tilted down toward my listening position. However, sitting in my usual seat, which placed my ears 36" above the floor, I found the sound a little lacking in top-octave life. Pink noise sounded smooth when my ears were between the two midrange-unit axes, so I replaced my usual chair with an office chair. This placed my ears 42" from the floor, which, at the 9' listening distance, put my ears on the tweeter axes.

Diamond and St. James had brought with them a complete set of Kubala-Sosna Elation! cables, including balanced interconnects, AES/EBU digital cables, biwire speaker-cable sets with jumpers to link the 1.2 and 1.3 woofer modules, and AC cables: total retail price, almost $40,000. Usually, I won't change the cables in my system when I'm setting up a new component for testing. However, in this instance YGA strongly felt that customers for the Sonja 1.3 would use K-S cables, and that I should take the contribution of the cables into consideration. I agreed, though reluctantly—not because I have anything against K-S cables, but because they would introduce another variable for which I would have to allow.

Next up was the choice of amplifier. I had to hand three pairs of monoblocks: Classé CT-M600s ($13,000/pair, 600Wpc into 8 ohms), Lamm M1.2 References ($23,890/pair, 110Wpc into 8 ohms), and MBL Reference 9007s ($42,800/pair, 440Wpc into 8 ohms). The dynamics were great with the Classés, but the sound was lean overall, the amplifiers keeping too tight a control of the YGAs' woofers. The Lamms usefully fattened up the lows, but the sound was too forward, the balance a little bright, and the dynamic range somewhat restricted compared to the other two amplifiers. The MBLs worked best, the treble balance proving optimally refined and the lows fully fleshed out. All of my following comments refer to the sound of the Sonja 1.3s driven by the MBL amplifiers and connected with Kubala-Sosna cables.

As with the Sonja 1.3's predecessor, the Anat Reference II Professional, I noticed no overt coloration. Pink noise sounded very smooth and even, though there was a slight mid-treble emphasis to the balance. Jacqueline Du Pré's cello in Elgar's Cello Concerto, with Sir John Barbirolli and the London Symphony (24-bit/96kHz Apple Lossless files from HDtracks, transcoded from FLACs), acquired a narrow band of brightness. However, as many other recordings didn't sound bright, it was difficult to decide: Was this a characteristic of the speaker, or was the Sonja 1.3 merely revealing a problem with some recordings?

It may well have been the latter, as there was no trace of brightness when I played the opening chorus, "Kommt, ihr Töchter, helft mir klagen," of one of my 2009 "Records to Die For": John Butt and the Dunedin Consort's performance of J.S. Bach's St. Matthew Passion (24/88.2 Apple Lossless files transcoded from FLACs, Studio Master download, Linn CKD 313P). This work opens with an ominously dark minor-key figure on violins doubled by oboes, over insistent ostinato, almost double-dotted tonic E-flats in the bass (footnote 2), which resolves with ascending cycle-of-fourth scales. When the voices enter, counterpoint is piled on counterpoint until, within the growing tension and complexity rather than on top of it, the sopranos float a slow-moving, pure-sounding cantus firmus based on a verse of the hymn "O Lamm Gottes, unschuldig." This presumably represents the innocent Lamb of God being led to the slaughter at Calgary, and through the Sonjas, the clarity of the sound and the stability and superb definition of the stereo imaging allowed this vocal line to emerge from the scoring without being exaggerated or buried.

In this respect, as massive and tall as the Sonja 1.3s are, their imaging was more like that of a superb pair of minimonitors—such as the KEF LS50s I reviewed last December. Dual-mono pink noise was reproduced as it should be, with a very narrow central image at all frequencies and no splashing to the sides. Playing "Corpus Christi Carol," from Jeff Buckley's Grace (CD, Columbia CK 57528), the speakers "disappeared," leaving Buckley's otherworldly alto voice hanging there in space, undefined by the speaker positions. But unlike good minimonitors, this ability to float an accurate, stable soundstage was matched by a full-range frequency balance. The bass guitar doubling the lead guitar in the widely spaced arpeggios following the diminished chord at the start of the chorus of "Lover, You Should've Come Over," also from Grace, had the appropriate body to its tone and excellent definition.


Listening to the warble tones on Editor's Choice (CD, Stereophile STPH016-2), I could hear the tones reproduced at full level down to 25Hz, with the 20Hz tone audible without doubling. The speaker's output was undoubtedly given a helping hand by the 32Hz mode in my room; while the low bass sounded a little disconnected in absolute terms, when there was bass, I heard it. In "Friends to Burn," from Jimmy Webb's 1993 album, Suspending Disbelief (CD, Elektra 61506-2), superbly recorded and mixed by George Massenburg, Art Dudley's favorite bass guitarist, Leland Sklar, drops down to a low D-flat (35.6Hz) in the chorus. This note was reproduced with superb clarity, there being no false bass boom to obscure definition. In "Fever," from the Tierney Sutton Band's Desire (CD, Telarc Distribution 83685), two double basses spin a rhythmic and harmonic web under Sutton's breathy singing, which the Sonja 1.3s kept distinctly defined even when the kick drum entered.

The replacement of the Anat Reference II Professional's powered woofers with passive bass modules in the Sonja was, I feel, an important step forward. But it was in the midrange where the YGA speakers blossomed. The half-step–spaced low-frequency tonebursts on Editor's Choice sounded generally clean, though with some faint ghosting/aliasing apparent in the upper midrange. Even so, female voices sounded maximally different from one another, whether it was Tierney Sutton in "Fever"; Ulla Meinecke singing "Die Tänzerin," from Wenn Schon Nicht für Immer, dann Wenigstens für Ewig (24/192 needle drop from LP, German RCA 426124); the title track of Shelby Lynne's Just a Little Lovin' (CD, Lost Highway B0009789-2); or even Adele's grossly overcompressed pipes on 21 (CD, XL Recordings 8697-444699-2). These speakers loved voices—perhaps, I conjecture, because their width, hence dispersion, approximates the size of the human head. With Richard Lehnert's speaking voice, in the channel-identification and speaker-phasing tracks on Editor's Choice, Richard was in the house, thanks to the Sonja 1.3s.

However, voices acquired a hard edge when the SPL at my listening seat rose above 100dB, measured with fast C-weighting with the Studio Six app on my iPhone 3GS. I was playing Vaughan Williams's Fantasia on Christmas Carols, with the Choir of Guildford Cathedral conducted by Barry Rose (CD, EMI Classics CDM 5 67427-2). The speakers "disappeared" at the start of the work, as expected, leaving the solo cello hanging in space, but it quickly became apparent that I'd set the playback level too high, breaking the 100dB barrier when the choir came in at full level. (See the "Measurements" sidebar for the probable explanation for this barrier.)

The YGA speakers left my room way too soon, to take up residence in another reviewer's system. All I'm left with is the lingering memory of the insistent baión rhythm of the very last track I played before Dick Diamond and Kerry St. James came to pick up the speakers: "Campo de Encino," from Jimmy Webb's 1972 album, Letters (CD, Reprise 10305). At his request, Webb soared away from my listening room, courtesy of the Sonja 1.3s.

Summing Up
YGA's Sonja 1.3 is that rare beast: a true full-range loudspeaker capable of playing at realistic sound-pressure levels with very low coloration and superbly stable, accurate soundstaging. Yes, it is very expensive—the most expensive product I have reviewed in my 37-year career of writing about audio—but its immaculate build quality and equally immaculate sound quality justify that price, at least for those with pockets deep enough to not only purchase a pair but to match them with appropriately high-quality amplification, cables, and source components. And the fact that the would-be owner can begin with a pair of Sonja 1.1s, add Sonja 1.2 woofer modules, and finally the 1.3 woofer modules, will take away some of the financial pain. Class A, all the way.

Footnote 2: E-flat minor? Six flats? That's a cruel and unusual key, even for the Well-Tempered Bach. The published score is in E minor, suggesting that the Dunedin Consort is using instruments tuned to historically correct rather than modern concert pitch.
YG Acoustics LLC
4941 Allison Street, Unit 10
Arvada, CO 80002
(801) 726-3887
Share | |
carlosgallardo's picture

I'd like to read a comparison between the Sonja and the Magicos Q5 and Q7, or Willson Alexandria


Bulsara's picture

It's a bargain ... stupid !

Lofty's picture

after the YGA guys left  - why didn't you install your usual cabling? It surprises me you didn't revert to your constant , then install the K-S cables and report the difference?

As you might say, "color me puzzled?"

John Atkinson's picture

To respond to these points, with speakers as large as the YG Sonja 1.3, Wilson Alexandria XLF, and Magico Q7 or Q5, the logistical problems doing side-by-side comparisons are immense. For example, I have a basement listening room and I don't see how I could get either the Wilsons or either of the Magicos in that room? If I remember correctly, each Magico Q7 weighs 700 lbs. (The YGs each split into three modules, which made the task manageable - just.)

But you can find our reviews of the XLF and Q5 at http://www.stereophile.com/content/wilson-audio-specialties-alexandria-xlf-loudspeaker and http://www.stereophile.com/floorloudspeakers/magico_q5_loudspeaker/index.html .

Regarding the lack of cable comparisons in my Sonja review, I ran out of time. However, I still have the Kubala-Sosna cables and will publish some comparisons, using different speaker, later in the year.

Regarding the sarcastic comment that the YG speaker is a "bargain," I feel that while the Sonja 1.3 is very expensive, it is indeed fairly priced given the bill of materials,  the cost of skilled labor, and the use of capital-intensive equipment (the cost of which has to be amortized over a relatively limited production run), to make it. Remember also that with a speaker this bulky and heavy, even the cost of shipping a pair to the dealer is expensive and that cost has to be built into the retail price.

I am reminded of a story told me by YG's Yoav Geva. He was contacted by a would-be customer who complained about a YG speaker's price and who estimated that there was around $200 worth of raw aluminum stock in that speaker.

"Send me a check for $200 and I'll send you the raw aluminum so you can build your own speaker!" responded Mr. Geva :-)

John Atkinson

Editor, Stereophile

GeorgeHolland's picture

Bulsara, yes it is stupid to say it's a bargain.

Lofty, Mr Atkinson doesn't test cables just allows subjective reviews.

Mr Atkinson, "I feel that while the Sonja 1.3 is very expensive, it is indeed fairly priced given the bill of materials,  the cost of skilled labor, and the use of capital-intensive equipment (the cost of which has to be amortized over a relatively limited production run), to make it"................"Having seen YGA's capital-intensive operation, I am surprised not only that the Sonja 1.3 doesn't cost more than it does, but also that it costs slightly less than the company's earlier flagship model."

Well I'll keep that in mind next time I have a complete house built from scratch. I think I'd get much more for the money than just a pair of speakers. Good luck to them selling enough of these to recoup their investment plus make a profit. Not many very wealthy people are going to buy these.

For $108,000  you can buy all the electronic, audio, computer, wood working equipment to build your own speakers and have plenty left over for a good sized down payment on that house. Plus you have the satisfaction of learning how to make those speakers and can make more and more for your family and friends. People do this all the time. Check out the DIY audio and speaker forums. Plenty of already tried and true speaker plans on there.

John Atkinson's picture

GeorgeHolland wrote:
For $108,000  you can buy all the electronic, audio, computer, wood working equipment to build your own speakers and have plenty left over for a good sized down payment on that house.

Not if you want to emulate the YG speakers. The 5-axis CNC machines used to mill all the speaker parts cost, if I remember correctly, several hundred thousand dollars each. And then there is the positive air pressure machine room, the swarf and scrap collection machines, the salaries of the skilled operators, and the cost of the milling tools, themselves. Let alone rent, cost of capital, worker health insurance and benefits, etc. And while aluminum is relatively easy to machine, the mineral-loaded phenolic resin material used by Wilson wears out several tools, at a couple of hundred dollars a time, for each enclosure.

It is all too easy for an armchair critic to poke scorn at companies like YG and Wilson. But the enterpreneurs behind these companies have done something the armchair critic has never done, which is to put at risk their ability to feed their families in order to realize their dreams.

No-one forces you to buy speakers like thiese, Mr. Holland, but you have to respect what their manufacturers are trying to achieve.

John Atkinson

Editor, Stereophile

GeorgeHolland's picture

Who wants to emulate the YG? There are plenty of designs out there just as good or better.  Look at Mr Linkowitz  http://www.linkwitzlab.com/   or John K's http://www.musicanddesign.com/naomain.html  others are also out there.

You don't have to spend obscene amounts like on a YG or Wilson to have fantastic sound.

Those impedance and phase numbers for the YG aren't exactly astounding by any means, you would think for the money that they would have tidied up the problems there and also the panel vibrations are not worthy of the price paid.

"But the enterpreneurs behind these companies have done something the armchair critic has never done, which is to put at risk their ability to feed their families in order to realize their dreams"

Oh stop it I might cry. Poor little Wilson has it rough I can tell.

"No-one forces you to buy speakers like thiese, Mr. Holland, but you have to respect what their manufacturers are trying to achieve."

Their achievement is to charge as much as possible and hope the rich buy it.

ChrisS's picture

Your links don't work.

JohnnyR's picture

  Too lazy to copy and paste the links themselves into their browser




[Edit by John Atkinson]

[Additional edit by John Atkinson]

Regadude's picture

The speakers in those links don't look impressive. But, I will reserve judgement because I HAVE NEVER HEARD THEM. 

A person would be silly to denigrate a product they have never seen and heard in person...

GeorgeHolland's picture

Oh that's right you can't.wink

Both of those speakers are very well known and admired in the audio community of course they aren't silly expenisve so they don't have the snob appeal like Wilson or YG does,

JohnnyR's picture

will be able to pick out what is and what isn't good about a product soley based upon the measurements, layout and other appearances. [flame deleted by John Atkinson]

corrective_unconscious's picture

Hope those manufacturers are better with audio design than they are with graphic design or even information graphics, because those are some primitive repositories. Irksome.

(Just for the record I'm quite sure it's possible to obtain superlative speakers for under 10k or 15k...even semi mass produced ones via brick and mortar audio shops.)

GeorgeHolland's picture

getting across the information. They are more of an info site than trying to sell something. If you want glitz then look elsewhere.

xsipower's picture

First just let me say WOW!! Nice speakers (no I am not being sarcastic). I would love to have a pair of these in my living room.

I will have to agree with Mr. Atkinson on his assessment of why these speakers cost what they do. Given that these speakers are not really considered “mass produced”, but made in small quantity I can understand the cost. Machining is very expensive and not a cheap way of making metal parts. It is a slow process and requires expensive machines that must be maintained and staffed by well-paid labor. The benefit of machining is its flexibility and precision. As they say – time is money.

 If they were truly made in high quantity, most manufactures would shift to casting or other forming methods which can save a lot of money because it is fast and does not require as expensive machines. That kind of work is usually farmed out to “third world” or “developing countries”.  Of course some finish machining would be required, but nowhere near as much as with billet machining.

I know that there is a lot of animosity toward the high end audio industry for their prices. Some of it I would say has foundation, but not all of it. People need to understand that many of these companies are small businesses that sell only in the hundreds, some if that. We get used to the mass produced “Made with slave labor” prices from Best Buy and the like. The business model for the smaller companies is totally different. They need to have on staff not only the founders and lower waged assembly people, but engineers and technicians who are expensive. Plus pay for insurance, rent, taxes, utilities, etc. Then there are the materials and subcontractors and expensive capital equipment. They must also spend their own money on trade shows and advertisements. They are not a non-profit organization.

I am an engineer myself; and have spent some 25 years in research and development. Developing products is expensive. While developing a product, you are making no money on it. It eats away at your cash like no tomorrow. By the time let say a high end amplifier goes from concept to production, a year can be spent.  It first needs to be designed on paper and simulated on computers. Then it must be bread boarded, then prototyped – maybe two or three iterations. Then it must be designed for manufacturing. That can add up to hundreds of thousands of dollars if not more. That investment must be recouped with hopefully some profit. Some can make stuff in a garage and be cheap and good, but once you move out from a garage your expenses balloon.

Yes we can look at it from an armchair DIY point of view, but that is way too simplistic and disrespects the effort it takes to make a well-designed product that can be reliably reproduced in quantity. We need to be careful when making judgments on high end items.

ChrisS's picture


Are you amongst the many un-wealthy who are also not going to buy these speakers?

carlosgallardo's picture

Thank you for making it clear for all of us.

I understand your logistical problems, my intention was not to complain but having more information from a reviewer like you that got the opportunity to listen to all of them, and at the same time have an respectable oppinion.

MVBC's picture

"YGA's Sonja 1.3 is that rare beast: a true full-range loudspeaker capable of playing at realistic sound-pressure levels"


"Sensitivity: 88dB/2.83V/m"

Good one!laugh

GeorgeHolland's picture

Mr Atkinson's words: "The Sonja's voltage sensitivity is specified as 88dB/2.83V/m. However, my estimate was significantly less than that, at 85dB(B)/2.83V/m"

So takes even twice more the power for the same loudness. If 1 watt = 85 dB then 32 watts = 100 dB and above 100dB it made the amp clip? Yep really worth every penny.

"This speaker needs to be used with an amplifier capable of delivering both amps and volts in quantity—I suspect that the hardness I noted at sustained SPLs above 100dB was simply due to the amplifiers clipping into the demanding impedance."

So lets say you spend another $50,000 on an amp capable of driving these. Nice.

ChrisS's picture

If Georgie can't find it at Wal Mart, Georgie ain't buyin'....

MVBC's picture

Nope, that is not his point: it means it'll take me 130 watts to get to 120dB on my JBL pro drivers system while you'll need 2,000 watts with these $100k fancy speakers driving mickey mouse drivers with very expensive diaphragms... To call these being able to reproduce realistic levels is a farce.

harishcs's picture

Even assuming that YG is not being unreasonable about the selling price, the question is whether all that elaborate machining,etc is necessary for achieving state of the art sound ?  





JohnnyR's picture

The answer is NO

[Off-topic text deleted by John Atkinson] Magico uses aluminum structures in their $20,000 speakers but is it worth the cost?  You can throw endless amounts of money into making things appear "better" and look impressive but I'd rather have a set of speakers that come with an actual frequency response graph included and it's not that hard to do so yet who does? It would prove that quality control is constant with not just the sample loaned to Atkinson but also the pair you yourself bought.

ChrisS's picture

Having truth and fact defined for you by Uncle Joseph for so many years, Comrade JRusskie, no wonder you trust a piece of paper with a graph over your own hearing.

JohnnyR's picture

Yet you seemed to have overlooked that fact. Atkinson takes frequency graphs, do you trust him? Floyd Toole trusts graphs, that's how the speaker industry found out what consumers liked the best. A speaker that has flat frequency rersponse. Too bad some designers overlook the importance of a benign impedance that doesn't tax the abilities of the amp.Do you trust anything other than your willingness to open mouth and insert foot?

ChrisS's picture

And they all trust their hearing too!

John Atkinson's picture

JohnnyR wrote:
I'd rather have a set of speakers that come with an actual frequency response graph included and it's not that hard to do so yet who does?It would prove that quality control is constant with not just the sample loaned to Atkinson but also the pair you yourself bought.

For production QA on finished speakers, YG uses MLSSA with the speaker suspended high in the air to give an anechoic time window of 13ms. But as I said in the review but you seem to have overlooked, YG did send me the response of what I understood to be the finalized protoype sample of the Sonja 1.3, measured in the NRC's anechoic chamber in Ottawa. And as I also wrote in the review, my measurement of the farfield response of my sample of the Sonja was almost identical in the midrange and above to YG's NRC measurement of the prototype. YG therefore does have very effective quality control and your insinuation is incorrect.

John Atkinson

Editor, Stereophile

JohnnyR's picture

Isn't the same as the production run in many cases plus you got a cherry picked set of speakers or do you believe that it was just a randon set the shipping clerk chose to send you? You seemed to have overlooked that I said  "but also the pair you yourself bought." so my point is still valid. Will YG send out frequency specs for every speaker?

John Atkinson's picture

JohnnyR wrote:
Isn't the same as the production run in many cases...

Again you misunderstand what you read. The sample measured in the NRC chamber was the final protoype, ie, the reference sample of the design. All Sonja 1.3s are measured as the final step of the production process and their response compared to that of the reference measured under identical conditions. The close tolerance of that QA process is confirmed by the fact that my measured response of a production sample was essentially the same as that of the reference sample, even given the fact the former was measured with a swept sinewave in the NRC chamber and I used a quasi-anechoic MLS technique.

I am sure there are companies that have poor or inconsistent QA. YG Acoustics is not one of them. It's time for you to stop beating this dead horse.

John Atkinson

Editor, Stereophile

GeorgeHolland's picture

So in other words, they don't supply a print out of the frequency response of the actual speakers you buy.


Enter your Stereophile.com username.
Enter the password that accompanies your username.