LAB12 integre4 Mk2 Toroid – homecoming to a powerful KT150 tube amplifier

Nostos

A word in Homeric Greek that English has never quite managed to carry across. Nostos. It is usually translated as homecoming, which is accurate and almost entirely useless, because it leaves out the only part that matters.

Nostos is not simply arriving back where you started. It is arriving back as someone else. Odysseus is twenty years getting home to Ithaca, and when he finally stands on that beach he is unrecognizable to his household, to his dog, very nearly to himself. The whole weight of the poem sits in that gap. The island did not change. He did. The homecoming means something precisely because of everything that happened in between, and a man who had simply stayed at home would have nothing at all to come back to.

Our hobby is full of small nostalgia. You return to a record you have not played in eight years, and it has become a different record. And occasionally something arrives at your door that sends you all the way back to the beginning of your own story, then makes you look at it again with everything you have learned since. That is what happened here, and it took me a while to understand what I was actually listening to.

TL; DR — The Short Conclusion

For those of you who read the conclusion first, and I know there are many of you, here it is, distilled.

The LAB12 integre4 Mk2 Toroid Version is a 65-watt-per-channel push-pull integrated amplifier built around a quad of KT150 power tubes, and it is not the amplifier its category leads you to expect. This is not a romantic valve amplifier. There is no golden haze, no soft focus, no bass line that turns into a suggestion when the music gets difficult. What it does instead is considerably harder: it keeps the tonal density and liquid, continuous midrange that make people love valves in the first place, then adds dynamics, grip, and micro-detail that would not embarrass a good transistor amplifier at the same price.

Its greatest achievement is control. True dual mono, with two separate toroidal power transformers and fully independent supply stages, gives it a noise floor I did not expect from a tube integrated and an authority that never sags when both channels are working hard. It drove everything without reaching, and it simply got louder when asked.

Its second achievement is one you will use every week rather than admire once. The built-in bias system, with direct anode current on the OLED and four trimmers under the screen, turns the worst chore in valve ownership into a pleasure. The KT150 window runs from 38 to 45 mA, and LAB12 leaves the choice to you, which means you can genuinely voice this amplifier to your own room in an evening with a screwdriver. The compatibility list runs from EL34 through 6550, KT88, and KT120, so the tube rolling is real, not theoretical.

It is hand-built in Athens, component by component and joint by joint, with its own tubes matched and numbered to its own sockets, and it carries a five-year warranty.

If I am being demanding, the very last degree of micro-dynamic shading is a step behind reference tube amplification, which costs several times as much. At €6,799, that observation is close to irrelevant.

It leaves with our Highly Impressive Award, and it leaves reluctantly.

April 2023

My first review for Soundnews was published on April 4, 2023. It was an integrated amplifier, and I wrote about it for seven and a half thousand words, most of which were about a Mars rover called Opportunity and a Greek goddess called Ēṓs, and only some of which were about the amplifier.

The machine was the Trafomatic Audio EOS, from Saša Čokić in Serbia. Back then, it cost eight and a half thousand euros, probably a lot more today. A push-pull integrated amplifier named after the goddess of the dawn, the one who opens the gates of heaven every morning so the sun can climb through. I ended that review with Billie Holiday singing I’ll Be Seeing You, the song NASA played across the solar system on the last day they tried to wake Oppy up. I have not been able to hear that recording since, without something happening in my chest. The amplifier is still in production, and my review is still online, so you can check it here. But a smaller story is buried in that review, and it is the one that matters today.

The EOS came to me configured for KT120 power tubes. I wanted a spare quad, but I couldn’t get them anywhere, and in the middle of that frustration I discovered that the newer, considerably more powerful KT150 was, absurdly, the easier tube to find. I swapped a few messages with Saša, who said it would be fine, and I fitted them.

Of everything I tried in that amplifier, and I tried a great deal, that was the single biggest change I could produce. I wrote at the time that the KT150 was faster, more articulate, extending better at both ends of the spectrum, and that the difference was not small. I also wrote, with the slight nervousness of a man walking on ground he does not yet know well, please correct me if I’m wrong; I know I am walking in unknown territory.

Nobody corrected me. Three and a half years and a great many amplifiers later, I would write the same sentence again, with rather more confidence and no less enthusiasm.

The Russian tube that quietly took over

It is worth pausing to consider what the KT150 actually is. It is a beam-power tetrode built in Saratov at the Reflektor plant and sold under the Tung-Sol name. It looks nothing like its ancestors. Where a KT88 or a KT120 is a conventional bottle, the KT150 is a fat glass ellipsoid, almost egg-shaped, and the shape is not stylish. The envelope geometry is part of how it handles heat and plate dissipation, which lets it deliver what it delivers.

It delivers power, headroom, and grip in a tube that will drop into many amplifiers originally designed around the KT88 family. Over the past decade, it has become the default answer whenever a designer wants genuine muscle from a valve output stage without moving to something exotic or unobtainable. If you have heard a modern tube amplifier that made you say, “I didn’t know it could do that,” there is a decent chance four of these were glowing quietly in the back.

And here is the part that took me three and a half years to appreciate. The KT150 is not the most powerful. It is accessible. The sound I first stumbled into by accident, because a shop was out of stock, is no longer the preserve of 10.000 euro amplifiers. Which brings us to Athens.

A Storage Room at Number 12

LAB12 was not founded. That is the wrong verb, and the company is honest enough to say so themselves.

Stratos Vichos was a child whose toys were loudspeakers, cables and circuits, a sentence that will make a certain kind of reader smile with recognition. He studied electronics, then spent years in professional audio, in sales, system design and installation, learning the whole chain rather than one link of it.

Then, in 2011, a serious accident stopped everything. He was confined for months, removed from his profession, with a great deal of time and nothing obvious to do with it. So he started building. Simple devices, for himself, during his recovery. He chose valves deliberately, not out of nostalgia, but because he had concluded that tubes could carry something specifications could not describe. Those first pieces went to friends, then to clients, and the response answered the question of what he would do next.

The workshop was a storage space in a house. The street number was twelve. The name followed, without a branding meeting. The company existed officially in 2012, and in 2013, one year old, took a room at High End Munich, which is either brave or slightly mad. Larger premises in 2018, doubled again in 2025. Today LAB12 exports to more than thirty-five countries through roughly two hundred and fifty selling points, and it is still Stratos and a small team in Metamorfosi, north of Athens.

And here is the part that I think matters most, because it has survived everything else. LAB12 still builds every single device by hand. Not assembled by hand from pre-populated boards, not hand-finished after a machine has done the work. Hand-built, component by component, soldered joint by soldered joint, by the same handful of people in the same building in Metamorfosi.

I find that story genuinely moving, and I think it explains the amplifier. A company that began as therapy tends to build things that sound like somebody cared. Their own summary is four words long: emotion through sound. I would normally roll my eyes at that. Having lived with the integre4 for several weeks, I am not rolling anything.

What the Toroid Version Actually Changes

The integre4 has been LAB12’s integrated for years, and everyone has liked it. What arrived at my house is the Mk2 Toroid Version, and the name tells you where the work went. The headline is that this is now a true dual mono design, with two separate high-current, low-noise toroidal power transformers and fully rebuilt dual power supply stages, one per channel. Not a shared supply with some separation downstream. Two of everything, from the wall inwards.

That matters more in a valve amplifier than people expect. A tube output stage is hungry, and its demands aren’t polite; they fluctuate with the music. When two channels draw from one supply, the bass line in the left channel can and does modulate what is available to the right. Separating them removes an entire class of interaction that most listeners have never sought out. The second change is the output transformers, which have been upgraded, and this is where a tube integrated amplifier lives or dies. You can compromise almost anything else in a valve amplifier and recover. Compromise the output transformers, and nothing downstream will save you.

Third, the chassis itself has been reworked for noise reduction, and the whole topology redesigned around the assumption that there will be KT150s in the sockets rather than something gentler. This is the crucial point. Plenty of amplifiers will accept a KT150. This one was rebuilt so that it could stop apologizing for one.

The result is sixty-five watts per channel, which does not sound like a large number until you remember what kind of watts they are, and a noise floor that LAB12 describes as unbelievable for a tube integrated. Marketing departments say that sort of thing routinely. In this case, I went looking for noise with my ear against the tweeter of a many-thousand-euro loudspeaker and heard nothing, not a trace of a hiss, not a hint of a whine.

The numbers

  • Output power: 65 W per channel
  • Topology: true dual mono, no global feedback, push-pull
  • Power transformers: two toroidal, one per channel
  • Tube complement: 4x KT150 power pentodes, 4x 6N1P dual triodes
  • Compatible output tubes: EL34, 6L6, 6550, KT88, KT120, KT150 from the factory
  • Frequency response: 15 Hz to 70 kHz (-1 dB)
  • Inputs: five unbalanced RCA pairs, one XLR pair (shared with input 1)
  • Outputs: speaker binding posts, 6.3 mm headphone jack fed from the output transformers
  • Proprietary: SRSG and Fine Symmetry implementations
  • Dimensions and weight: 420 x 220 x 380 mm, 22 kg
  • Finishes: Matt Black, Avus Silver
  • Warranty: five years, ninety days on the valves
  • Price: €6,799

Eight Tubes and What They Do

The tube complement is worth a paragraph each way, because LAB12 has made two decisions here that are more interesting than they look.

Up front: four 6N1P dual triodes. Another Soviet-era design, another tube that has quietly outlived the empire that produced it, and a choice that says something about priorities. The 6N1P is not the fashionable option. It is robust, linear, cheap to replace, and in good NOS examples it has a directness and a lack of editorializing that more glamorous small-signal tubes sometimes lack. LAB12 ships them matched and numbered, marked to the socket position they were calibrated in. At the back, four KT150s, also matched and numbered. The amplifier was tuned around them, and it shows.

But the compatibility list is the real statement of philosophy, and it deserves careful reading: EL34, 6L6, 6550, KT88, KT120, KT150. That is not a footnote. That is an invitation.

This is the same purist logic Trafomatic followed with the EOS, and I found it so appealing. You are not locked to a single tube by a fixed bias circuit that was set at the factory and sealed, or, even worse sometimes (from the sound perspective), an auto-adjusting bias with every swing of the music. You are given a proper adjustable bias, a chart of safe values, and the assumption that you are an adult who might want to hear what an EL34 does to your loudspeakers on a Sunday afternoon. It treats tube rolling not as a risk to be engineered away, but as one of the reasons you bought a valve amplifier in the first place.

The difference between the EOS and the integre4 is what LAB12 has built around that freedom, and it is the best part of this review.

Twenty-Two Kilograms of Greek Restraint

It is a big machine and does not pretend otherwise. Forty-two centimeters wide, thirty-eight deep, and twenty-two kilograms on the scale. You will want a shelf that means it, and you will want to think about where the heat goes.

You won’t find any drama in the styling. The front is a five-millimeter aluminum plate with a glass-blasted anodized finish, in matt black or silver, and the only things on it are a power switch, a navigation knob, an OLED window, four small trimmer holes, and a volume control. That is the entire interface. There is no chrome, no illuminated logo, no attempt to look like anything other than a serious piece of equipment made by people who did not want to distract you.

I like it a great deal, and I know I am supposed to have opinions about industrial design. The restraint reads as confidence. The build quality matches the weight. The chassis is genuinely solid, the finish is even and well done, the binding posts are serious, the sockets are tight, and the whole thing has the slightly industrial density of equipment built by a small team that assembles it themselves and knows they will be the ones fixing it if it breaks. LAB12 backs it with a five-year warranty, which tells you what they think of their own soldering.

One detail I appreciated: every tube arrives numbered, matched to the socket it was calibrated in during factory burn-in. A bias screwdriver comes in the box, along with the remote and a manual that is genuinely useful rather than legally defensive, which is rarer than it should be.

Built by hand, and I mean that literally

I want to spend a moment here, because handcrafted is one of the most abused words in this industry, and LAB12 is among the few companies still entitled to use it without a lawyer present.

Metamorfosi has no production line. There is no pick-and-place machine populating boards overnight while everybody sleeps. Every LAB12 device is assembled by hand, component by component, and every joint is soldered by a person. The company’s own literature puts it plainly: each amplifier is totally handcrafted with matched parts, and every component in it was chosen after hours of listening and testing rather than selected from a catalog on price and tolerance. The consequences of that are visible if you know where to look, and audible in ways that are harder to prove but easy to believe.

Start with the tubes. Your integre4 did not receive a generic matched octet from a bin. It received its own set, matched and burned in inside your chassis during calibration, then numbered to the exact socket each valve was calibrated in. That is why the manual tells you to fit V1 into V1 and not simply anywhere convenient. Somebody sat with your amplifier while it warmed up and wrote those numbers down.

Then the quality control. One person, Charis, inspects every stage of every device’s production, and the company describes his hearing as that of an analyzer, which is a joke with a serious point behind it. The final arbiter at LAB12 is a pair of ears attached to somebody who will be personally embarrassed if the unit is wrong.

Now the uncomfortable half, because I am not here to write a brochure. Hand building is slower, more expensive per unit, and completely dependent on the people doing it. It does not scale. It cannot be rushed when demand spikes, which is precisely why LAB12 has doubled its facilities twice rather than outsourcing a single stage of it. And it puts an enormous amount of faith in individuals having a good day.

The counter-argument is the one I find persuasive, and it is written into the warranty. Five years. Not two, not three. Five years on an amplifier full of high voltages and hand-soldered joints, from a company small enough that a bad month of workmanship would be an existential problem rather than a line item. You don’t offer that unless you know exactly who made each unit and can go ask them about it.

There is one more consequence, and it is the least discussed. A device built entirely by hand can be repaired entirely by hand, indefinitely, by the people who built it. There is no proprietary populated module to become unobtainable in twelve years. In a hobby increasingly full of equipment that becomes landfill the moment a surface-mount board fails, that is worth considerably more than it costs.

The Best Half Hour I Spent With It

Here is where the integre4 stops being a very good amplifier and becomes an interesting one. Setting up is undramatic. Solid surface, air around it, speakers connected before you power up, which is not optional advice with a valve amplifier and is printed in bold in the manual for good reason. Turn the main switch, and a soft-start circuit takes about a minute to bring the tubes up gently. The OLED wakes into the menu.

That menu has five entries, and every one of them earns its place. Input selection. Display contrast. Valve hours, a running total of how long your tubes have actually worked, which you reset when you change them and which removes the guesswork from the single most tedious question in valve ownership. Input naming. And then the one that matters.

Valve bias. Turn the knob, and the display walks you through each of the four power tubes in turn, showing the anode current directly, in milliamps, live. Below the screen sit four trimmers. You take the supplied screwdriver, you watch the number, you set it. No multimeter, no test points, no removing the bottom plate, no crouching on the floor with probes in one hand and a torch in your teeth. But still very sensitive, so any movement will have a dramatic effect; great care shall be taken.

LAB12: printed chart from the manual: 6L6 — 11 to 14 mA, EL34 — 13 to 18 mA, 6550 — 23 to 29 mA, KT88 — 27 to 28 mA, KT120 — 32 to 39 mA, and KT150 — 38 to 45 mA.

Read that last line again, because it is the interesting one. On the KT150, you are given a seven-milliamperage window, and LAB12 does not tell you where in it to sit. They tell you the boundaries, and then they get out of your way. The manual’s only firm instructions are sensible ones: always turn the trimmers fully counterclockwise to cold bias before fitting new tubes, disconnect your inputs while adjusting, and repeat the procedure at least three times during the first five hours as everything settles. I did it twice; the third time wasn’t necessary, as the bias didn’t change anymore.

So I spent an evening living inside that window, and I recommend you do the same.

At 38 mA, the bottom of the range, the amplifier is leaner, faster on its feet, and slightly more forward. Transients arrive with a snap. It runs cooler, and it feels, if anything, more like a very good solid-state amplifier having a good day. On bright recordings, or into a loudspeaker that is already lively in the presence region, this is not necessarily where you want to be.

At 45 mA, the top end relaxes and fills out. The midrange gains body and moisture, the bass gains a little weight and bloom, and images become rounder and slightly larger. It also runs hotter and, I assume, shortens tube life somewhat, though I haven’t owned it long enough to say that with authority.

I settled at 42 to 43 mA, and I settled there quickly, which surprised me. That is where the integre4 stopped sounding like it was leaning one way or the other and simply sounded correct, with the dynamics of the low setting and most of the body of the high one. Your loudspeakers, your room, and your taste will move that number, and that is exactly the point. This is not a tone control. It is a genuine voicing adjustment, in your hands, with a screwdriver, taking about ninety seconds.

I have set the bias with a multimeter on the carpet in the dark. I am not romantic about it. What LAB12 has done here is take the one genuinely off-putting chore of valve ownership and turn it into the most enjoyable half hour of the review.

The Rest of the Machine

A brisk tour of what else is in there, because several of these are not obvious from a spec sheet.

No feedback

There is no global negative feedback. The amplifier is not correcting itself against its own output, which preserves dynamic freedom and an unforced quality, and which places enormous demands on the linearity of everything in the signal path because nothing downstream will tidy it up. It also means, as on the EOS, that changing tubes changes the voicing more than in a heavily fed-back design. That is a feature, not a bug, and it is why the bias window matters as much as it does.

SRSG and Fine Symmetry

Two registered LAB12 implementations that appear on every relevant product and about which the company is not especially forthcoming in public. What I can tell you is what the amplifier does rather than how: channel-to-channel matching is exceptionally tight, and the balance holds together at both very low and very high listening levels in a way that is not typical of the breed. 

Protection and monitoring

A powerful processor sits behind the OLED, watching the bias continuously rather than only when you ask. This is the design’s quiet luxury. A valve amplifier with a tube going soft can do expensive damage before you notice anything is wrong, and a system that is paying attention all the time is worth more than any single component upgrade you could make for the same money.

The volume control

A motorized Blue Velvet ALPS potentiometer, which is the good one, driven from the remote. It tracks properly at low levels, which is where cheap potentiometers embarrass themselves and where most of us actually listen most of the time. The remote is simple and does what it should.

Connectivity

Five RCA inputs and one XLR pair, with input 1 shared between the two socket types. Plenty for a real system. There is no phono stage and no DAC, which I regard as correct at this price; you will get a better result from separates anyway.

The headphone output

Not an afterthought op-amp bolted to the side. The 6.3 mm socket is fed through the amplifier’s own output transformers, so what comes out of it is genuinely the same tube amplifier, at a scale that will not wake the house. I ran the Susvara through it out of curiosity, and it was better than it had any right to be. That deserves its own review.

Burn-in

LAB12 asks for 150 hours before final judgments, and they are not being coy. It arrives sounding good, reaches close to its potential around a hundred hours, and settles properly after that. Be patient, and re-check the bias as it goes.

Only the Music

Let me start with the sentence I kept circling back to over several weeks, because it’s what you most need to know.

This is not a romantic amplifier.

I want to be clear about that, because a large part of the audience for a valve integrated is looking for exactly the opposite. If your mental image of a tube amplifier is warmth, softness, a golden haze, and a bass line that is more suggestion than statement, the integre4 will confuse you for about ten minutes and then win the argument.

What it does instead is harder to engineer. It takes the genuine virtues of valves, the tonal density, the liquidity through the midrange, the way a voice occupies space rather than merely appearing in it, and it keeps them. Then it adds dynamics, bass control, and micro-detail that wouldn’t embarrass a modern transistor amplifier at the same price.

It is a modern tube amplifier in the most literal sense. It sounds like it was designed by someone who loves valves and refuses to use that love as an excuse.

Power, and the control that comes with it

Sixty-five watts per channel from a valve output stage is a great deal more authority than the number implies, and the dual mono supplies are audible in exactly the way theory predicts. The amplifier does not sag. When the music demands current in both channels at once, which is most of the time and all of the interesting time, nothing thins out, and nothing hardens.

Isao Suzuki Trio — Black Orpheus (Three Blind Mice, analog, 45 rpm reissue)
I go here first with anything claiming grip, because Suzuki’s cello and bass work lives in the shadowy lower registers where most amplifiers offer an approximation and hope you are not paying attention. Through the integre4, the instrument had a body, an actual physical size, wood and rosin and string tension rather than a generalized low-frequency event. More importantly, it stopped when he stopped. The decay was his, not the amplifier’s.

That control is the headline. A great many valve amplifiers produce impressive bass and rather less impressive bass definition, and the gap between those two things is where the romance charge usually sticks. No gap here is worth discussing.

Dynamics, and a modern test

Tool — Fear Inoculum (2019, digital, 24/96)
This is the record I use when I want to know whether an amplifier is actually in charge, and it is a deeply unfair thing to play through valves. Danny Carey’s drum work is fast, complex, and layered across a mix dense enough to punish anything with a soft grip. The title track will expose compression, congestion, and a lack of headroom within two minutes.

The integre4 held it. Every one of it. The kit stayed spatially separated instead of collapsing into a rhythmic smear; the low toms had weight without bloat, and the long crescendos landed as crescendos rather than gradual increases in volume. I never felt the amplifier reaching its limit. I then did what any reasonable person would do and turned it up well past the volume reasonable people listen to. It simply got louder. That is the whole report.

The midrange, and the word liquid

Etta Cameron & Nikolaj Hess — Etta (2009, digital)
I have written about this record before, and I will keep doing it. Etta Cameron’s voice is one of a small handful that translates directly into emotion with nothing in between, and this album was recorded a year before she died, which you can hear whether or not you know it.

Liquid is a dangerous word in a review because it usually means the writer enjoyed something soft. What I mean here is specific: the midrange has no discontinuities. The voice moves from chest to head, from a whisper to full projection, without any point at which the presentation changes character, hardens, or draws attention to a transition. That continuity is what valves do at their best, and it is fully present.

The absence of thickening is notable: no extra syrup, no artificial glow around the vocals, and no romantic soft-focus effect. The voice exists independently, with its unique texture and weight. I listened to both Summertime and Motherless Child without jotting down any notes, which is the only review that counts.

Detail, and where the noise floor pays for itself

The Dave Brubeck Quartet — Time Out (Analogue Productions 45 rpm, analogue)
Familiarity is the point here. Take Five has been played so often that unfortunately it has become furniture, and a system that makes it interesting again is telling you something real. Through the integre4, Joe Morello’s snare had skin, stick, and air around it; Paul Desmond’s alto floated with that dry-martini elegance that must never harden; and the 5/4 pulse locked in hard enough to make my foot move without permission.

The micro-detail is genuinely impressive and directly connected to the noise floor. With no signal playing and my ear at the tweeter, there was essentially nothing. What that buys you is not quiet as an end in itself. It is everything that was previously sitting just underneath the noise, which now has room to be heard. This is where the integre4 separates itself from the tube stereotype most decisively: fine detail, the small breathing gradations of a performance, is normally the first casualty of a warm-sounding design, and here it survives intact.

Effortlessness

Marcus Miller — Laid Black (2018, digital)
Miller is my favorite contemporary electric jazz artist, and Laid Black is a superb test of a specific quality that is hard to describe but immediately obvious when you hear it. When a system is genuinely capable, this record does not get louder. It gets easier.

That is what happened. The bass texture went deep and stayed articulate, the horn arrangements had bite without glare, and the whole thing had a relaxed forward motion that comes only from an amplifier with reserves it is not using. I played the album twice and started it a third time before noticing.

If I am being demanding, and at this money I should be, the very last degree of micro-dynamic gradation, the smallest changes in pressure and intent, is just a step behind what my reference chain does with amplification costing several times as much. That is an entirely honest observation and an entirely unsurprising one. At this price, it’s also close to irrelevant.

Loudspeaker matching

LAB12 recommends four to eight ohms on the specification sheet, and the manual is more generous at four to sixteen. I used it with the Wilson WATT/Puppy 50th Anniversary, which isn’t a difficult loudspeaker on paper but requires a lot of current to reach the level of transparency and control that it is famous for, and it drove them with plenty of room to spare and no sense of strain at any level I tried. Based on these weeks, I would be comfortable putting this amplifier in front of the vast majority of loudspeakers a reader is likely to own, including many conventionally considered to need transistors. The company claims it can be the ideal driver for even the most demanding speakers out there. That is a strong claim, and I didn’t find it inaccurate.

Market Value

Let us speak plainly about money, as we always do. The LAB12 integre4 Mk2 Toroid Version costs €6,799. Please confirm current pricing with your dealer, because this category moves and taxes vary.

Now let me put that number where it belongs. The Trafomatic EOS, the amplifier that started this story and the one that taught me what a KT150 could do, was €8,500 when I reviewed it in 2023. Both are push-pull integrated amps built around a quad of the same tube; both take a purist approach to bias and tube compatibility; both are made by small European teams who assemble every unit themselves.

I can not tell you the LAB12 is better than the Trafomatic or the other way around. They are different amplifiers with different characters, and I loved the EOS enough to give it our Gold Award. The EOS had a mythic tonal generosity that suited its name. The integre4 is the more modern of the two, tighter and more explicitly controlled, and it does more with dynamics and low-level detail.

What I will tell you is this. Three and a half years ago, the sound I am describing here was something you reached for at eight and a half thousand euros and up. It is now available for one thousand seven hundred euros less, with five years of warranty, a genuinely useful bias system, dual mono supplies, a headphone output fed from the output transformers, and a tube compatibility list that treats you as an adult. That is what progress in this hobby is supposed to look like, and it happens far less often than the marketing suggests. And remember what you are actually buying at that number: an amplifier hand-soldered in Athens by people whose names are on the company website, with its own tubes matched to its own sockets, backed for five years.

Set against the wider market, the case is straightforward. At around seven thousand euros, you are choosing between a well-built solid-state integrated from a large manufacturer, which will give you more raw watts, more features and less soul; a smaller tube integrated of perhaps thirty or forty watts, which may give you a more delicate midrange and considerably less authority; or this. If you want valves and you want to drive real loudspeakers to real levels and you want the modern virtues of detail and control, the number of genuine alternatives at this money is small enough to count on one hand.

The Return

I began with Odysseus on a beach, unrecognizable after twenty years, and I want to close there.

The amplifier that taught me what a KT150 was capable of arrived at my house by accident, because a shop was out of KT120s, at a time when I had written exactly zero reviews for this magazine and was nervous enough about it to write- please correct me if I’m wrong– my first published paragraph about tubes. Three and a half years later, four of the same bottles arrived from Athens in an amplifier costing considerably less, built by a man who started making audio equipment while recovering from an accident in a storage room at number twelve.

The island did not change. The tubes are the same tubes, made in the same plant in Saratov, doing the same physics they have always done. What changed is everything around them. The supplies are doubled. The transformers are better. A processor watches the bias every second the amplifier is on, an OLED counts your valve hours, a screwdriver comes in the box, and a seven-milliamp window lets you voice the thing to your own room in an evening. The noise floor has gone somewhere I did not expect a tube integrated to reach.

And what also changed, if I am honest, is me. I know what to listen for now in a way I did not in 2023. That is what makes this homecoming worth writing about, because the integre4 is not a nostalgic object and never asked me to be nostalgic about it. It asked me to pay attention, and then rewarded the attention for weeks without once making me wish for something else.

It is dynamic and powerful, with real grip and real detail, a midrange that is liquid without being sweet, and a bass line you can follow note by note. It will drive nearly anything you own, sit quietly in your rack for five warranted years, and let you change its character with a screwdriver on a wet Sunday if you feel like it. If that isn’t one of the most complete arguments for a valve amplifier currently being made anywhere near this price, I haven’t heard one that beats it.

The “Highly Impressive” Award

Some components impress with ambition; others impress with execution. The rarest ones impress by proportion, by delivering so much more at their price than the market has taught us to expect, that they quietly redraw the map for everybody else.

Based on everything documented above, the true dual mono architecture with two toroidal transformers, the sixty-five genuinely capable watts, the exceptional noise floor, the bias system that turns the worst chore in valve ownership into a pleasure, the wide tube compatibility that respects the owner’s intelligence, the five year warranty, and above all a sound that combines the tonal truth of valves with the dynamics and control of modern hardware, I am delighted to present the LAB12 integre4 Mk2 Toroid Version with our Highly Impressive Award.

Highly impressive is exactly what it is. Stratos and the team at Metamorfosi should be extremely proud of this one.

Thank You — HiFi Expert

Every experience of this quality begins with somebody generous enough to make it possible. My warmest thanks go to the team at HiFi Expert (hifiexpert.ro), who entrusted me with this amplifier for all this time, with patience and genuine enthusiasm.

They are the sort of dealer this industry needs more of: people who actually listen, actually care, and actually believe in what they represent. If you are in Romania and any of this has caught your interest, go and hear it in their room. That is what listening rooms are for, and theirs is a good one.

For everyone else, LAB12 can be found at lab12.gr, where you will also find a dealer near you.

Thank you all, and as usual, please feel free to leave a comment in the section below. Until next time, I wish you to hear only the best!

Associated Equipment

Loudspeakers: Wilson Audio WATT/Puppy 50th Anniversary. DAC: LampizatOr Pacific. Streamer: Rockna Wavelight Server. Roon Server: Gustard S26. Turntable: Acoustic Signature Montana NEO with TA-5000 NEO tonearm. Cartridges: Ortofon MC A95, Hana Umami Red. Phono stage: Aesthetix Rhea Eclipse. Cabling: Albedo Monolith Reference speaker cables and XLR interconnects, Crystal Cable Van Gogh phono, AudioQuest Dragon power. Power distribution: Tsakiridis Super Athena. Network: optical separation with Roboli S80 ExpertLine. Reference amplification: Aesthetix Calypso Eclipse and Atlas Eclipse. Under review: LAB12 integre4 Mk2 Toroid Version.

Catalin Cristescu

I’m a Graphic/UI designer, user experience specialist, a tech addict, an enthusiastic entrepreneur and last but not least – a passionate music lover. I love art in any shape or form, transforming my music collection into an essential part of my entire life. I’m not your typical hardcode audiophile, but it’s still the dominant hobby that drives me forward.

Related Articles

Back to top button