The Ultimate Preamplifier?
Before loading up for bear, it’s probably best to start simply, don’t you think? Yeah. So, here’s a simple question: what is a preamplifier?
Put at its simplest, a preamp is a volume knob with a switch. That’s really it, too. The job of the preamp is to attenuate the signal — turn down the volume — coming from the source. If you have more than one source, the preamp very handily allows you to select the one you want to listen to without having to move any wires. Other features, like a remote control, are
for lazy pansies creature comforts and not necessarily central to the purpose.
I take it that that’s pretty much unobjectionable.
With that, the Lords of Argument will allow us to move to the question: “What is the ultimate preamplifier?” I submit that how you answer that question says a lot about you as an audiophile. Of course, this is like saying “tell me about school vouchers” or “describe the proper role of government”. If you’re not “in the know”, the answers you get are likely to be bewildering — and if you are in the know, the question itself is polarizing.
I’m going to punt on most of that. Sorry.
Here’s the issue at hand — there are many folks that feel that the only good preamplifier is an absent preamplifier. But what does that mean? Taken one way, there’s a camp of audiophiles that believe that the only contribution a preamplifier can make is bad one. The argument goes like this: if the job of the source is to create a pristine signal and the job of the speaker is reproduce that signal pristinely, and the job of the amplifier is to let the speaker “be all it can be”, then the job of the ideal preamp is fairly straightforward: to get the hell out of the way. That is, it should add nothing. The corollary is that should also remove nothing — except volume. And that’s it. It makes a certain kind of sense.
There’s another camp of audiophiles that feel that a preamp can make the system sound better. That it can — and should — add “something” to the sound, especially if that something is coloration provided by blessed beautification of thermionic joy, aka, vacuum tubes. We’ll get to this later. Let’s go back to that first group for now.
For some of that first group, you can take that notion of “absence” a whole lot farther — right up to “not there”. And no, I don’t mean “sonically invisible”, I mean, “get rid of the preamp entirely”.
There is real movement in this direction these days and I know quite a few audiophiles are ditching their preamps and selling them on AudiogoN, or simply not buying them in the first place, and simply running “DAC-direct”. Yes, this approach is limited to the digital-only crowd, but that is the fastest growing segment in the community (said I based on no research whatsoever), so I feel confident in saying that this is an approach worth keeping an eye on. They note that the DAC itself can implement a volume control, and do so quite capably, therefore — why do I need a preamp at all? Of course, this is the same crowd that looks at the “vinyl resurgence” with something rather more than mere disdain, but put the whole digital vs analog question aside for a second. If you can eliminate an entire device in the playback chain, the argument goes, won’t this help quality? Better — can this do anything but help improve the quality of the playback chain?
This also makes a certain kind of sense, no?
Unfortunately, nothing is ever straightforward. In this case, it assumes that the DAC’s implementation of a volume control is “bit-perfect” — by which I only mean “non-destructive” to the signal being converted (a huge assumption, by the way). In point of fact, most volume controls implemented digitally are destructive of the signal to some extent. As a quick example, take a 16 bit file. To attenuate the volume digitally, DACs will perform math that effectively “tosses bits”. The short form goes like this. Want to drop the volume 6dB? Toss a bit. Want to drop another 6dB? Toss another. And so on. Got it?
Here’s the problem. Most volume controls — and digital ones are no exception — start at full. That is, they attenuate — the baseline is “maximum”. Most of us don’t tend to listen with the volume knob buried all the way over to the right, so this means we’ll be attenuating — and attenuating quite a bit. In fact, most of us will listen at normal volumes, somewhere below -24dB from unity (max). That requires 4bits to be excised. I personally tend to do quite a bit of day-to-day listening around -36dB. That’s a total of 6bits being tossed from the audio table like semi-gnawed bones. If I had only 16 to start with, I’m now down to 10 bits! What happens when you toss 6 bits? Well, if you’re starting with only 16 to begin with, the answer is pretty simple (for an audiophile): bad things! You can lose dynamics, detail, and soundstage — the end result tends to sound boring.
Some argue that tossing any more than 2 bits is seriously damaging to sound quality, if true, this leaves us in something of a pickle. 2 bits? That’s only 12dB down — which, on most systems, would be deafening. Which is why it’s lucky that most DACs that tend to implement volume controls also tend to be 24 bit. Those extra 8 bits are there, among other things, to allow your CD-quality (16-bit) files plenty of room to walk bits overboard willy nilly and still have plenty o’ bits left over to reproduce your file bit-perfectly. And this works superlatively … right up until you feed it a 24 bit file. Oops. Sorry, HD Tracks!
Aside: the argument that says “I wouldn’t/couldn’t notice it anyway, so what’s the big deal” is interesting, but I’ll punt the psycho-acoustic debate, too. For now, I’ll simply take it as read that it’s at least possible to perceive such degradations and leave the evidence collection and analysis to some future all-too-clever neuro-psych grad student. Does the question grip you? Take it up with The Future.
So, non-destructive attenuation is one problem a good preamp needs to solve. Noise suppression is another, I would think. It simply wouldn’t do for a preamp, or any device in a playback chain, to introduce noise to the chain. Of course, this is easier to say than do, and my guess is that this is exactly where preamp designs go from functional to expensive. All preamp makers need to address noise in a fundamental and non-trivial way, so let’s punt on this too.
Okay, so what else does a good preamp do, aside from switching sources? I don’t want to minimize this, as this is the overriding reason as to why a preamplifier is important to me — I’m not getting rid of my analog front-end — but after switching, it gets a little fuzzy. Let me suggest a rather important, and generally ignored bit — the preamp has to actually drive the amp.
Sounds simple! But again, nothing is quite straightforward, innit? What I’m waving my hands at by saying this is, quite generally, the analog output stage. If you want it to be transparent, you’re gonna have to do something here. Wise Men Have Said (always a good indicator of Absolute Truth whenever They get invoked) that the analog output stage is at least as important as the “everything else” in any device that’s doing to the work of a preamp. Honestly, this is precisely why teeny-tiny DACs will never sound as good as a “full fledged, fully fleshed out” preamp will — the tiny analog section is … anemic … at best.
I’m going to keep waving my hands here and offer that the power supply is also a huge factor, but more serving the output than the input — whatever happens upstream, the amplifier downstream must be fed good, clean signal. Or else.
Last but not least, there’s the issue of gain. Most preamps actually do amplify signal — that’s what makes them “active” — and for good reason. A modest gain stage can do a lot to normalize the signal going to the amp and (attempt to) correct some of the flaws introduced by the attenuation circuit, and still provide enough juice to make the amp sit up and bark like a dog. If it stays modest, it can also be relatively low-noise. Whether or not this is necessary is another issue, but whatever.
Anyway, this is the job of the preamplifier. In a nutshell. All good? Okay.
Let me introduce a rather special one: the Genoa.
I ran across Sonus Veritas at RMAF back in 2011 when they were showing their Venice phono preamp, what I think what was a pre-release Genoa linestage, and a not-yet-completed Modena DAC (all three of which are now orderable). The Modena made a debut at last year’s RMAF and earned a very favorable impression of Mal Kenny, our new “West-Coast Correspondent”. I think he said something like: “this is the total opposite of suck”, which is about as high as praise can get.
Sonus Veritas showcases vacuum tube technology — the Genoa sports a quartet of 6N30P tubes. Mmm mmm! Tubes are good. I like tubes! The cases themselves are rather unassuming, until you climb up on top of them; the finish is brushed-aluminum in a lustrous black, with chunky knobs and buttons, and softly glowing LEDs.
As for the price, the Modena and Genoa pieces retail for $16,999 each and the Venice starts at $21,999. Pricey, yes, but the lure … the lure! I’m a sucker for those Sirens. Yeah, it’s a character flaw.
So I reached out to Joe Rosovitz, the president of Sonus Veritas, shortly after RMAF and requested a review sample of … something. Honestly, I was curious enough at this point to take whatever they would send, so I told Joe what I was curious about, generally: a reference preamp and something to one-up my reference DAC, a Berkeley Audio Designs Alpha Series 2. If he felt so inclined, he could take his pick. I was hoping he’d say “preamp”, because at that point, I didn’t really have one. I’d recently sold my Plinius Tautoro, but that was okay, as I still had an overachieving “activated passive” from Wyred4Sound and a superlative true passive from Luminous Audio. What I didn’t have was a truly active reference preamp. As luck would have it — and luck has whatever it damn well pleases, I don’t care what you say — a Genoa preamp found its way to my audio rack in short order.
The Genoa is, properly, a linestage — there is no option for an included phono section or DAC (those are separate products, remember?), so all the Genoa does is switch sources and change the volume. Nice and classic, just like I wanted. The unit supports both balanced and single-ended connections, with an interesting twist: little toggle switches on the latter to lift the ground — by input! My old Plinius only let me do that for the system as a whole. Double-dual outputs — two sets of XLR and two sets of RCA — round out the back, and all outputs are active. The volume control is motorized for use with the included, matching and quite hefty black brushed aluminum remote.
Overall, the look is a bit Volvo — boxy, but good — from the front and top, and from a distance, all the distinctive bits fades into a soft black blur. The fascia is simple and borders on plain. The stenciling is all in gold, which shows off rather subtly against the matte black of the casework, and is hard to see from any distance or in challenging lighting. The LEDs, your only indication that things are actually happening, can be dimmed remotely. You know. In case you’re easily distracted.
If you were expecting a glittering jewel of audio technology for money that could just as easily purchased you a Honda Civic, you’re going to be disappointed. It’s not on the menu here, and Kevin straight up told me that looks aren’t the point. He cares about what it sounds like, and bling in the casework simply adds to cost. The cost, while considerable, comes from somewhere else — in this case, it’s the parts. We’ll come back to this.
A couple of interesting fiddly-bits — from the remote (and the front panel), you can switch the polarity (good to know) and from stereo to mono. That latter one is fun, especially when undoing some early stereo recordings where each instrument is nailed to a given speaker with rebar and railroad spikes. Okay, maybe it was just me, but I think it’s a cool option. There’s also a Mute button and a Power button; the latter initiates a 30 second warm-up delay. Last is Gain button. The Genoa has 6dB of gain normally; hitting this button adds another 6dB lift. Very nice when you’re working with a low-gain amp, but have a care as it does quadruple the output impedance to (a still-moderate) 400 ohms.
The unit is deceptively large, especially for a preamp. While this gives plenty of room and isolation for the internals, it will eat up some space on the rack. It’s also a “healthy” 45lbs in weight. Clean living and a robust diet?
The first, and perhaps most central, tech that the Genoa leverages is a shunt volume control. Heard of that one? It was new to me. Here’s why it’s interesting. As all you learned audiophiles no doubt know, a series volume control is, in essence, an operation on a relationship between input and ground. As you move the dial, say, on a potentiometer, the relationship between input and ground changes. At any point on the dial, R1 is value between the selected point and input/unity (max volume) and R2 is the value between the selected point and ground (no volume). Turning the pot’s dial changes R1 and R2, and this is equivalent to changing a pair of resistors for ones with different values at each step on the attenuator.
In a shunt control, R1 is constant, and R2 is the only thing that varies. In a balanced shunt attenuator, like the one used here in the Genoa, there are two R1 values (two phases!) and still only one R2, with no ground. The two R1 are out of phase with each other, so feeding some of one into the other cancels the overall output in relationship to the amount of R2. Slick, eh?
A shunt volume control still uses a resistive divider to reduce the signal level like the more common series volume control, but fixes the resistors in the signal path and switches a resistor that shunts a portion of the signal to the other signal phase, thereby cancelling the shunted portion reducing the volume by that amount. Given the choice of the differential amplifier topology we use in the Genoa, this type of volume control is a natural choice as it obviates the need for closely matched and tracking dual controls for each channel to maintain a high degree of common mode rejection as would be needed for series volume control implementation.
You ever notice how some answers simply open up new questions?
Q: The preamp is called a “differential amplifier design” — can you explain what that means and why its different and important?
A differential amplifier has two input wires and it amplifies the electrical difference between the signals on these wires. The circuitry itself can be vacuum tube, solid state or a combination thereof, as ours is. For an exceedingly common example, virtually all opamps have differential input stages. The rejection of noise that is common to both input wires and the minimization of distortion are the two important attributes of differential amplifiers that are the most significant for audio applications. Our implementation takes the technique off in a direction that is not common outside of the professional audio business. That technique is the use of coupling transformers for the input and output of the differential amplifier.
The noise rejection property of differential amplifiers is properly referred to as common mode rejection. This technical term simply means that noise present equally on the two input terminals is not amplified (actually it’s rejected, hence the term common mode rejection), but a signal present across the input terminals is amplified. To ensure that the signal presented at the input terminals is faithfully amplified, the vacuum tube differential amplifiers used are supported by extensive power supply decoupling and regulation. The differential amplifiers themselves “float” relative to power supply ground, decoupled from the power supplies by very high impedance cascode current source/sinks. Simple, but sonically effective shunt regulation is used to stabilize the power supply voltage at the differential amplifier loading transformer’s center tap. The output of the differential amplifier, being transformer-coupled, is also independent of the local ground, providing a “natural” balanced output.
Q: You use a lot of transformers in your design — and that’s a lot of wire. Some might say that “shortest path” is best. Can you talk a bit about why you chose this design? Are there trade offs you had to make for the gains?
I first started experimenting with audio coupling transformers back in the mid-1990s and compared the sonic results to the coupling techniques used more commonly for interfacing tube circuits. My decision to use audio coupling transformers in this and the other Sonus Veritas designs was entirely motivated by the sonic results. The amount of wire in the signal path might be an issue if considered on its own merit, but listening convinced me that the result was the better for it when it came in the form of a transformer. Several years ago I had a local mastering engineer visit our facility to talk about his monitoring system. He brought one of his latest projects on CD to give me some sense of what he was doing in his studio. After auditioning the music he asked me about the reproduction chain. The tubes were obvious, but he was floored when I told him that there were five coupling transformers in the signal path. He allowed that when he was in the sound engineering program (at a prestigious university) they were taught that coupling transformers were nearly obsolete, because of cost, high distortion, and poor bandwidth. He was forced to reevaluate the performance part of that teaching based on what he heard that day. High quality transformers can provide a level of fidelity that rivals the fidelity provided by other techniques and offer some unique advantages. You’ve had a chance to evaluate the sonic results yourself.
Of course the added weight and cost of implementing the best transformers is not a matter to take lightly (pun intended!), but the results speak for themselves (another bad pun). Technically speaking, a good input transformer has even better common mode signal rejection than the amplifier itself and allows for the mixed use of unbalanced and balanced inputs. The output transformer allows us to reduce the output impedance of the amplifier without use of large numbers of paralleled tubes or negative feedback. The other important advantage provided by coupling transformers is that of providing galvanic isolation between audio components, which avoids the problem of ground loops and the noise that can result.
Q: While I have — can you talk about the power supply? Why solid state regulation?
The raw power supply in the Genoa is very simple. After solid state rectification, there is a single stage choke input filter. The choke input filter avoids much of the noise generated by less expensive capacitor input power supply filters. After the raw power supply, each differential amplifier stage is “coddled” by a “blanket” of current and shunt voltage regulation to set the tube operating conditions precisely for consistent sound over the life of the tubes. We use a special circuit that indirectly monitors tube emission and lights an internal red LED (that can be seen looking through the top panel ventilation slots) in each channel when the tubes have slipped below a predetermined threshold for performance. Listening tests convinced me that the sort of solid state support circuitry we use provides the best sound compared to a variety of other approaches. I won’t claim that the same techniques would be the best with other types of amplifier stages, but they work in an advantageous way with our differential stages.
And there you go. But how does it sound?
What is the sound of one hand clapping?
This is a frustrating little Zen puzzle. In short, there is no sound to one hand — and that’s the point. Sometimes, no sound is the most profound sound of all. In the domain of a audio’s high-end, this sums up the preamplifier pretty well, don’t you think?
Great! … Which is a problem.
Some components just scream out their successes and deficiencies with bold or italicized declamations. These are easy products to write about! The main challenge, for the reviewer, is to ensure that no molehills are granted ‘mountain’ status. Given that the entire reviewing enterprise seems an exercise in hyperbolic nitpickery, this is harder than it sounds.
That said, I’d love to say that the Genoa is shockingly, embarrassingly, obviously, awesome along any metric you care to measure. The problem is that to do this, it requires reference to a objective standard … that doesn’t exist. More properly, you have to compare with nothing. How do you compare something with nothing? If the goal of a preamplifier is to “do nothing”, how do you know when you’ve gotten there?
Before the Genoa, I hadn’t really considered the problem. Hadn’t happened, really. Each preamp I’ve tried has deviated from perfection in some predictable or obvious ways. The Genoa … did not. In fact, the Genoa in many ways resembled the objective so closely that it was hard to really tell if it was doing something. Hell, it was sometimes hard to tell if the damn thing was on. I mean, there was something … but ….
How to capture that, though? I don’t really have a good analogy, which makes this problematic. I’ve likened similar effects to the difference between F-stops in photography. Moving from an f/2 to a f/8 usually adds a significant depth-of-field, rendering more of the landscape “in focus”. This isn’t necessarily a good thing in photography — the goal is to bring the subject into focus, and at least in portraiture, if the background can vanish into a pleasant blur, that’s perfectly fine. In audio, of course, this is reversed — the audiophile wants (preternatural) focus all the way into the stereo image, even if this does distract from the action going on.
This isn’t like that. That’s (almost) quantitative — with that analogy, I’m talking about detail retrieval approaching “real life” (and sometimes going beyond that). Here, with the Genoa, I’m not seeing any deeper into a sound stage, per se. The depth is whatever it is — the linestage seems fairly faithful to the quality of the recording being handled. No, here it’s more about depth. Not the fact of it. The character of it.
I think the most common way this gets talked about is to say that my references sound flat. They don’t, of course is, I’m just talking about by comparison.
Let me try it this way. I’ve talked before about going from “regular” TV to a 3D TV. With the 3D TV, you now get some interesting cues about space. But the images in 3D TV can feel something like cardboard cut-outs floating out into distinct layers — it’s all very interesting, but it’s not real. With the Genoa, the sound stage I was hearing was like comparing a really good 3D TV image to watching a play and seeing, with my own eyes, the actual actors move through real space. That’s the kind of depth I mean. It’s not a terribly subtle effect, to be sure, this difference in a representation and the thing being represented, but it still wasn’t exactly obvious at first listen. If I were to exaggerate for the purpose of making a distinction, I’d say that the with the Genoa, what I was looking at wasn’t a movie screen, it was a box seat — I was watching the real thing. An exaggeration, but an interesting one.
Okay, so failing abjectly to precisely capture the overall character, let’s slice and dice by doing some comparisons.
I started with a couple of passives, including a “true” passive, a Luminous Audio Axiom II “Walker Mod”, as well as an “activated passive”, the award-winning Wyred4Sound STP SE. The $399 Axiom is a true over-performer, and for the price, is one of the most effective bangs for the audio buck I’ve found in the high-end. It’s simple, compact and affordable. I’m on record as being a fan. If it had a fault, it was that a 23 step attenuator was a bit coarse. Interestingly, Team Luminous will be offering a 47 step attenuator shortly, and that may eliminate my objections entirely.
By contrast, the Wyred4Sound STP SE, sounds more dynamic than the Axiom — an extra grand-and-a-half does buy you something! The sound stage is deeper and wider and the bass is more controlled. Overall, I found that the STP SE has better structural support, especially at the frequency extremes, but does present a bit more opaquely than the Axiom through the mids. Not much, but some.
The Genoa here split the difference — clarity through the mids and frequency support at the farthest reaches. Shazam! Interestingly, the Genoa also seemed to “do something” with and to the sound I was getting. By comparison. Said one way, I felt that the Genoa brought a fuller palette of color to my Magic Bus painting party. Everything was just subtly more real, more textured, more nuanced. Vocal resonance? More organic. Said another way, it was like walking into a cliche about tubes vs solid state — even if that “tube sound” was modern (i.e., linear) — everything just felt a bit less about speed and a bit more about flow. At no point did I get any sense that there was anything remotely like roll-off going on and it wasn’t just the mid range that seemed to benefit from the Genoa — things just sounded right. Like I said, a total cliche, but a very happy and livable one.
A note — 6dB isn’t a lot of gain these days, and you may well find yourself cranking the volume control farther than you’ve traditionally been comfortable doing. As for me, I was routinely hitting and pushing to (and occasionally past) the 3-o’clock when I wanted to “go there”. When was the last time you did that on your preamp? Kevin Carter seemed to shrug when we chatted, saying that I shouldn’t be afraid to turn it up. There is all that space on the dial — seems silly not to use it all — and unlike other approaches, his volume control was made to be useable all across the band — at all positions. And no, I didn’t run out of gas before my ears did ….
The question about value is something that’s a bit harder to justify. $17k is a lot of money — for anything. So, to do battle with the Evil Demons of Audio Economics, due diligence requires we ask — what am I getting for all that cash?
“Quality bits” might be a good place to start. Kevin tells me his parts-cost alone puts the tab north of $4,000 per box. And then there’s assembly, marketing, dealer margins … at the end of the math, the dollar figure is $16,999. Yes, it’s a lot, Kevin says. But in all fairness, he wasn’t designing to a price point — this is what he felt would be a great way to build a preamp — so he did it and it costs what it costs. “Affordable” is what he does over at K & K. Sonus Veritas is another thing entire.
Speaking of which, I asked Kevin why he didn’t choose to implement any silver transformers, an offering that his favored brand, Lundahl, recently brought to the market. It was here that the cost-no-object mentality got tossed out the window — Kevin seemed to audibly wince over the phone while chatting. Those new silver transformers do in fact sound better, he says, but they’re also far more expensive — as in over three times the price of their copper siblings. An all-silver version of the Genoa would be thousands of dollars more. By way of an only tangentially relevant example, Purity Audio is another preamplifier maker that heavily leverages transformers throughout their designs, and they do offer an all-silver version — for well north of $30k. Yikes. So, while in an absolute sense, the price of the Genoa is certainly high, it may not be outrageous when compared with competing approaches.
Getting back to the question of value ….
One can argue that the difference between a good preamplifier and a great piece of audio gear is usually measured with extra zeroes on the price tag, fair or not. If that bothers you, there’s a lot of very good work being done down-market — check out the STP SE and the K & K stuff for handy examples — but moving up from good requires some serious dance moves. As the Wise Men Have Said, the relationship between quality and cost is not linear — and with the Genoa, it’s necessary to realize that you are fully around that curve heading toward crazy-cash for incremental gains.
This is a congested area of the competitive band, as Art Dudley has noted — there are quite a few designers hitting out there with big, gold-plated, baseball bats. Take, for example, Pass Labs’ XP-30 in solid state and the BorderPatrol Control Unit EXT-2 for tubes — both are stunningly good examples of what’s possible. A bit more cash (what’s a couple a bucks between friends?) and we’re now talking SOTA gear from Constellation and others …. All in all, this is rarefied air, to be sure. Can the Genoa hang in such company?
Oh my, yes.
Here’s my only hangup — when you shell out that kind of dough, it’s no longer all about the sound … and if I was being honest, I think the aesthetic should match the lofty price. If that matters not at all to you, you’re in good company, but I’m a shallow, vain, fool of a man and I have my expectations.
Okay, time for the bottom line.
The Genoa is a fantastic preamplifier. The sound quality is impeccable — it renders the music as transparently as a passive and brings all the life you need or would want from an active. You could say that it steps out of the way … and I think this would be entirely fair and high praise at the same time … and perhaps the best compliment you can pay to a preamplifier.
Nice work — recommended.
Final note: a Most Wanted award is pending. I need more time on the box and more time with some of its competitors before making any final moves here. In the meantime? Get yourself to RMAF or call Sonus Veritas about how you can arrange your own demo.