Like a lot of adults who attended too many rock concerts in their reckless youth, my hearing is not what it used to be. On more than one occasion, I remember stumbling out of Winterland in San Francisco after seeing high-watt bands like Hot Tuna or Pink Floyd, putting the key in the car’s ignition, giving it a turn, and then having no idea whatsoever if the engine had roared to life.
That’s what four or five hours of standing in front of a wall of speakers pumping music at more than 100 decibels will do to a person’s hearing—for the following 30 or 40 minutes, the world sounds soft and muffled, as if the air is thick with invisible clouds of cotton balls. At the time, it didn’t occur to me that I was doing lasting damage to the cochlea in my inner ear, and all these years later, I don’t necessarily wince at every sound that happens to be loud. But I do have trouble hearing someone speaking to me in a crowded restaurant, and certain sounds with the wrong acoustic profile (for me, anyway) will make my ears ring for hours.
“It doesn’t matter if it’s tubes, transistors, or a hamster on wheel. All that matters is that you got lost in the music.”
Most painful—emotionally and literally—is the mediocre fidelity of my home stereo system, which teases the listener with the occasional splash of treble or thump of bass, but mostly delivers cacophonous mush. It’s the opposite of what most people would describe as “warm,” which is a narrow, technical term of art among audiophiles. For the rest of us, warm suggests rich and rounded tones, notes and chords of such depth the listener can almost imagine he’s in the presence of the singer or musician performing without the aid of microphones or amplifiers. Warm is intimate, warm is clean and pure, warm doesn’t make my ears ring.
Uniquely, tube amplifiers, which use vacuum tubes to amplify electrical signals, are said to deliver this sublime auditory experience more reliably than their solid-state counterparts, which use transistors to do the same thing. (Digital devices run on integrated circuits, and use software to achieve their sound, so they are not considered here.) In particular, most tube amps are regarded as being less likely to create harmonic distortions at higher frequencies than all but the best and most expensive solid-state amps. They are generally worse at the lower frequencies, but our ears don’t hear most of those lower frequency distortions very well, which makes them “sonically benign.” Distortion at high frequencies, however, is easily heard and contributes to listening fatigue (i.e., ringing ears), which may be one of the reasons why tube amps are said to sound warm.
I don’t know much about harmonics, but I can vouch for the sound quality of tube amplifiers. I grew up listening to music played through my parents’ Fisher 500-C stereo receiver, a tube amp from the early 1960s that did wondrous things to albums like “Let It Bleed” by the Rolling Stones when played at very high volumes. My ears never rang after listening to that, no matter how loud. The same could not be said for the Kenwood KR-7200 solid-state receiver I took with me to college—if memory serves, I sold it to a wide-eyed freshman early in my sophomore year.
Lately, the desire to replicate the warm auditory memories of my youth has become a musical preoccupation of mine, since I’m secretly—if only aspirationally—in the market for a new stereo. Sure, a tube amplifier won’t help me hear someone talking to me in a noisy restaurant, but it does promise relief from the worst sonic indignity of all—not being able to listen to the music I love at a respectable volume without destroying what’s left of my hearing. If tubes could do that, it would be nothing short of a miracle.
So, I went shopping. Not for equipment yet, but for knowledge. Is there something about the way in which tubes, or “valves” as they are known in the U.K., amplify sound that changes how we experience it once it finds its way through our ear canals and into our brains? And although I know what the word “warm” means to me, what does it mean to the audiophiles and the people who make tube amplifiers and other types of hi-fi stereo equipment for a living? To get answers to these and other questions, I spoke to some of the leading authorities and manufacturers of tube and solid-state amplifiers in the United States. And, the icing on the cake, I got to listen to best stereo system I’ve ever heard.
To begin, I consulted the highly regarded “Sounds Like?” audio glossary, written by the late, great J. Gordon Holt, who founded “Stereophile” magazine in 1962. According to Holt, warm describes sound that is “the same as dark, but less tilted.” In case you’re curious, “dark” refers to “the audible effect of a frequency response which is clockwise-tilted across the entire range, so that output diminishes with increasing frequency,” while “tilted” indicates an “across-the-board rotation of an otherwise flat frequency response, so that the device’s output increases or decreases at a uniform rate with increasing frequency.”
This is not at all what I was expecting. Turns out, my definition of warm (“intimate,” “clean and pure,” “doesn’t make my ears ring”) is too “euphonic” for Holt, which is a word he dismisses in the “E” section of his glossary as “pleasing to the ear” but having “a connotation of exaggerated richness rather than literal accuracy.”
Audiophiles, it seems, have no use for emotional words like “warm,” but isn’t “pleasing to the ear” what I want? In the world of high-end hi-fi, which is where you need to go if you want to learn anything meaningful about tubes and tube amplifiers, the answer to that seemingly simple question is just about always “no.”
Much to my surprise, the way in which tubes work is actually one of the reasons why just about every audio expert since Holt has been hesitant to ascribe emotional characteristics like warmth to sound, as I learned from a number of hi-fi authorities I spoke to, including Kevin Hayes, who runs Valve Amplification Company, or VAC, in Florida. Hayes has been making tube amplifiers for about 25 years. In the mid-1990s, when tube amps enjoyed their first mainstream post-transistor comeback, VAC made thousands of replicas of the Model 7 preamplifiers, Model 8B stereo amplifiers, and Model 9 monoblocks for Marantz, which along with McIntosh, Scott, Harmon Kardon, and Fisher was one of the leading home-electronics manufacturers of the late 1950s and early 1960s, the golden age of tubes. These days, VAC makes labor-intensive, obsessively perfectionist craft pieces, usually no more than 100 or 150 per year.
All those years of working with tubes have made Hayes more than a casual observer of 12AX7s, 300Bs, 845s, and other types of triode tubes. “The triode is the most common type of tube in audio amplifiers,” Hayes explains to me over the phone. An hour or so later, I had learned that a triode vacuum tube has three main elements, arranged inside its glass housing, or envelope, in a concentric fashion. The inner element is called the cathode, which is surrounded by a coil-like length of electrically charged wire referred to as the grid, beyond which is a ring of sheet metal called the anode, or plate, which is shaped, Hayes says, kind of like a small can with no top or bottom. When the cathode is heated to around 2,000 degrees Fahrenheit, either directly or indirectly, the electrons on its oxidized surface are released into the vacuum, where the charge of the grid propels them toward the anode.
A little bit of heat here, the release of some electrons there, and just like that, a small electrical signal has been converted into a mighty jolt of AC voltage. And that, in a grossly oversimplified nutshell, is the definition of amplification.
Tubes, though, are more than just elegant voltage amplifiers—they are excellent linear voltage amplifiers, which means that when you increase the input going into the tube, you end up with an almost perfectly proportional increase on the other end. That makes tubes easier to work with than solid-state parts, as Charley Hansen of acclaimed solid-state audio-equipment manufacturer Ayre Acoustics knows. “If you know how to solder and the basics of electronics,” he says, “you can make a tube amplifier that’s going to sound really, really good, almost guaranteed.”
In contrast, you can’t just buy a solid-state version of a tube and swap it out it, one for one. It takes a small handful of solid-state parts to replicate what a tube does. Each of those parts can introduce an acoustic artifact that must be corrected, forcing solid-state circuit designers to employ tricks like negative feedback, in which the output of the amplifier is re-injected into the input to achieve the inherent linearity produced by tubes. Negative feedback is actually widely used in almost all amplifiers, be they tube or solid-state, although Hansen does not resort to it in Ayre products, and thinks it’s a detriment to either technology.
“The best triode tube is very close to perfectly linear,” continues Hayes. “The best transistor is more likely to have a sharp bend somewhere along the line, which must be corrected. Transistor-amplifier designers have a complicated job of transforming these bends into a facsimile of linear at output. To their credit, they’ve gotten very good at doing that, but you have to use more parts, and the more parts you have, the more secondary effects there are.”
Charley Hansen knows these parts as intimately as Hayes knows tubes. Hansen’s company in Boulder, Colorado, is highly regarded for the magic it works with capacitors, rectifiers, field-effect transistors, and other solid-state goodies—without resorting to feedback to achieve linearity. Hansen’s wizardry with solid-state audio circuitry is so renowned, he was selected by Neil Young himself to produce the audio circuit for the rock-star’s portable music player, the Pono, which is sort of like having Babe Ruth choose you as the guy to make his bats. And when Hansen really gets it right, as in the solid-state circuit he designed for Ayre’s KX-R Twenty preamp, the result is said to replicate the sound of—wait for it—tubes.
Like me, Hansen grew up listening to music played through tube amplifiers. Unlike me, Hansen got his hands dirty building them before he was 10 years old. “When I was 3 or 4,” he recalls, “my dad, who was an electrical engineer, bought a Dynakit tube amplifier, which he put together. Prior to that, he had bought a cabinet, made a mono Eico EF-12 tube amplifier from a kit, bought a turntable and a speaker, and put it all in the cabinet. When he upgraded to the Dynakit, I put the Eico in a case, the speaker in a cabinet, and the turntable on a stand. That was my system in the 1960s, when you could still buy tubes.”
In the late 1960s and early 1970s, electronics manufacturers started bailing on tubes because solid-state components were all the rage. In part, it was because the electronics industry wanted something new to sell (sound familiar?), but transistors represented a major improvement over tubes, even if it was a hassle to replicate their linearity, for the simple reason that unlike tubes, solid-state components did not wear out.
“Tubes are it, tubes are the ultimate, but tubes are completely messed up,” Hansen says. “They wear out the minute you turn on a piece of equipment.”
For Hansen and millions of other hi-fi-equipment consumers, this is a deal-breaker, even though many of these same people are untroubled by the planned obsolescence of the portable music devices they carry around in their pockets.
Buying a cool new music player, though, at least rewards a consumer with some new bells and whistles to play with. Wondering when a particular tube in your stereo is going to die, and remembering to plan for this absolute certainty, is more like getting excited about a trip to the dentist.
Besides, most of us can’t remember to change the batteries in our smoke alarms twice a year, and you can buy batteries in any grocery store. In contrast, the smallest power amp VAC makes, the Phi 200, has eight tubes—four are KT88s (power tubes that last anywhere from 6,000 to 8,000 hours before failing, and which cost between $40 and $400 each from specialized online retailers) and four 6SN7s (dual triodes, which last from 10,000 to 15,000 hours and range in price from $20 to $90 each).
It gets worse: Power amps are just the devices the speakers are hooked up to, and some people swear by the sound of what are called monoblocks, which are speaker-specific power amps, one for the left channel, one for the right. In front of the power amp, or pair of them if you go the monoblock route, is the preamp, which is what you plug your CD player, tape deck, and other components into. Some audiophiles also purchase what are called phono stages, which are specialized preamps that have been designed expressly to correct the imperfect equalization that is deliberately designed into all vinyl records (they have more high frequencies than low). All of this means there may be another dozen or more tubes, per component, that a music lover must remember to replace.
How often does this replacing occur? Well, taking the worst-case scenario of 6,000 hours, that’s roughly the equivalent of every live concert by the Grateful Dead during its 30-year performing history, including its 37 shows in 1974 that were amplified by 48 McIntosh MC2300 amps, producing 28,800 watts of power for the band’s famous Wall of Sound. If you listened to just one of the shows from the Dead’s catalog per day, it would take you about five-and-a-half years to experience the whole thing. At that point, you’d know it was time to change your tubes—or some of them, anyway, since tubes wear out at different rates—but by then your wife would probably have filed for divorce, so you’d likely have other things on your mind.
“Generally speaking, a tube will last anywhere from 2 to 20 years,” says Hayes, getting as specific as it’s apparently possible to get, “depending on how much someone listens. But I have tubes that were made in the 1930s that still sound glorious.”
By this point, I was more than ready to hear “glorious,” to experience for myself the sonic beauty of “warm,” but playing the role of devil’s advocate, Hayes wanted to make sure I knew what I was getting into. Those tubes he has from the 1930s, for example, are the exceptions rather than the rule. Most tubes go bad much sooner, and in all sorts of entertaining ways.
The most common failure scenario for a tube is for the oxide on the surface of the cathode to essentially run out of electrons to release when it’s heated. That means there’s basically nothing to send to the anode to produce amplification. The problem with this type of failure, though, is that if you have an amp that doesn’t seem to be delivering what it used to, how do you know which tube is bad? New McIntosh tube amps actually glow green when everything is a-okay and red when the tube has failed, while the VAC Statement 450 iQ delivers an alert, but what about the vast majority of tube amps out there that just begin to sound crappy?
In the 1950s and first half of the ’60s, checking the quality of a tube was simple, since just about every drugstore sold tubes and offered customers a machine for testing the tubes in their radios, televisions, and home stereos. Today, some enterprising online tube retailers make up for this lack of basic information by selling packages that allow you to upgrade, say, all 11 of the tubes in your older McIntosh MC275 at once, which is undeniably convenient but both wasteful and costly (expect to pay about $500 for this type of routine maintenance).
Some failures are simple to spot, such as a crack in the tube’s glass envelope, which causes the silver-black coating on the top-inside surface of a tube to turn white. Other tube malfunctions, though, are downright insidious.
“If you’re hearing sputtering, kind of like the sound of a thunderstorm coming in over an AM radio,” Hayes says, “that’s typically the result of some of the oxide on the cathode condensing after the tube cools down, either on the grid or the disc that holds the tube elements in place. The condensed coating forms conductive pathways, allowing a bit of current to conduct between elements that shouldn’t. Those pathways break and re-form, resulting in this random sputtering noise. Technically, nothing has worn out, but cross-contamination has occurred inside the tube.” In other words, the tube may still have plenty of life left in it, but you wouldn’t want to use it.
Lest it go without saying, solid-state circuits don’t do this sort of thing.
Tubes are also microphonic, says Hansen. “If you tap on a tube, especially in a phono stage, you will hear very loud noises coming through your speakers. I’m fairly convinced,” he adds, “that much of the ‘depth’ from tubed gear is literally just the sound of the speakers vibrating the tubes and creating an acoustic feedback loop, much like reverb in a guitar amplifier. Pleasant? Yes. Accurate? Probably not.”
Nor are all new tubes created equal, which has always been the case but seems more pronounced now that practically all new tubes are made in China and former Soviet Bloc countries. The last major U.S. tube manufacturer was a company called MPD, which made tubes in the old General Electric plant in Owensboro, Kentucky, until the mid-1990s. Today, the vast majority of the world’s vacuum tubes, most of which are produced for guitar amplifiers, are made in just a few factories in China, Russia, the Czech Republic, the Slovak Republic, and Canada.
As a curious side note, the tube industry in the former Soviet Bloc is actually a product of the Soviet military’s paranoia. Vacuum tubes are more immune to electromagnetic pulses, one of the many effects of a nuclear blast, than solid-state circuits. Therefore, the Soviet military kept a fully operational, tube-based communication system going long after solid-state circuits had become the norm in commercial applications. That way, Soviet generals on the ground would be able to communicate with crews aboard their nuclear-armed MiG fighters that might still be in the air, even as nuclear warheads are exploding below. Tubes would provide the Soviets with a critical communications lifeline, although that’s probably the wrong word to describe the state of humanity post-World War III.
Today, opinions about the quality of the tubes being made in the former Soviet Bloc and China vary wildly. Ken Kessler, who literally wrote the book on McIntosh—his coffee-table tome is called “For the Love of Music”—and has been an audio-equipment reviewer for more than 30 years, believes consumers can get very good tubes today. That, he says, has not always been the case.
“The one thing true today that 25 years ago no one would’ve believed is that the quality of tubes is no longer an issue,” Kessler says. “There are still some pretty crappy tubes out there, but you get what you pay for. I use tubes from EAT,” he adds, “which is based in the Czech Republic. Good tubes cost more than $15 or $20. In some cases, you could add a zero.”
Kessler is not exaggerating. Remember VAC’s Phi 200 amplifier, the one whose KT88 tubes range in price from $40 to $400 each? The EAT tube is the one at the high end of that spectrum.
Most tubes are somewhere in between. “A receiver tube is a consumer-grade tube, so it’s got to be cheap,” Charley Hansen says of tubes made for home electronics. “Receiver tubes have indirectly heated cathodes, which means they have much shorter lifespans than industrial tubes with tungsten filaments.”
During the heyday of tube-based home electronics, drugstores sold tubes like they do batteries today. “They were designed to last a few years,” Hansen continues, “basically until the warranty was up. In the ’50s, if you bought a TV set, it might have a two or three-year warranty, which was about the life of the tubes. If it failed after that, it wasn’t the manufacturer’s problem. You’d just go down to the drugstore with a handful of tubes, plug them into a tube tester, and buy some new ones. They were designed to be disposable.”
That ethos, Hansen cautions, is alive and well today. “Do you think those guys in Russia or China give a shit how long their tubes last? Hell, no. They want them to wear out so you’ll buy another one.”
Though hardly a tube-industry cheerleader, Hayes has a more nuanced view of the tubes he selects for new VAC amps. “There are some really fine-sounding tubes being made today,” he says. “In fact, surprisingly, some of the tubes coming out of China are measurably more linear than ones that came before. For example, the original spec for one of the most commonly used tube, the 12AX7, had a certain amount of non-linearity to it. The Chinese manufacturers have actually reduced that, so some interesting things are going on.”
Finally, and of almost no concern to any of the experts I spoke to, there’s the most surprising characteristic of tubes for people who have grown up with iPods or solid-state gear—their heat. Heating a tube’s cathode to 2,000 degrees Fahrenheit is what makes it glow, which means tubes not only make music sound warm, they throw off a fair amount of heat. “I typically listen to tube units in the winter,” jokes Steve Rowell of high-end audio retailer, reseller, and repair shop Audio Classics in upstate New York. “That way I can use their heat to warm the house.” Generating so much heat, though, requires a lot of energy, far more than is needed by their solid-state cousins, which means tube amps, though sonically appealing and undeniably cool to look at when they are glowing amid the rest of your hi-fi components, are also agents of climate change.
Given their high cost, unpredictable failure profile, variable quality, and carbon footprint, why does anyone bother with tubes at all? Well, it comes back to their sound, which by now I was positively itching to hear.
For Luke Manley, owner of Vacuum Tube Logic, better known as VTL, the sound of music played through tube amplifiers is as close as electronic audio is ever going to get to live, unamplified music—be it the sound of the human voice or an instrument like an acoustic guitar or grand piano—which is the reference point for all of VTL’s products, the bar against which everything else is measured. Manley finds this reference point with the help of his wife, Bea Lam, who is VTL’s operations manager and arguably the company’s most important pair of ears.
Here’s how they do it: “As soon as the engineer has a prototype, we go into the listening room,” Lam says. “Usually, I play a chamber-music piece first. That allows me to listen to the individual tonal color, or timbre, of the violin, viola, and cello. Then I might move on to a larger orchestral piece, to see if the prototype can reproduce the broader orchestral sound.” In some cases, Lam is very familiar with the pieces she’s listening to because she’s played them before on the piano. I would probably describe Lam as a pianist, but she won’t go that far, calling herself “a lifetime student” of the instrument. “Piano music is especially important to me,” she says.
Lam speaks of music as if it’s three-dimensional, a form with mass, shape, and weight, not merely a combination of sounds that occur, are heard, and then fade into the void. “Sometimes I will say, ‘Luke, I’m hearing the individual instruments but not the whole.’ That sense of the each instrument coming together to create a whole is very important. It’s not enough to reproduce each instrument by itself. The palpability of the music is very important.”
“When you play a large-scale symphonic piece, that’s where the rubber meets the road.”
Communicating what Lam is hearing to VTL’s engineers is Manley’s job. “I try to resolve the measurements we’ve seen on the bench with what Bea says she’s hearing,” he says. “What we look at depends on her description of the sonic characteristic. If she says it sounds dark or rolled off, then we look at frequency response. If it sounds like everything’s there but it’s just not that engaging, if it’s sort of boring sounding, then we check to see if we’ve got too much feedback somewhere. If it’s got muddy bass, we look at the bass response. These are things that you can’t measure easily with test equipment, so we listen to go further.”
After a product goes back to the engineers for redesign and improvement, it’s returned to the listening room for another round with Lam, and the process continues until Manley and Lam feel like they’ve gotten it right. “Critical listening is an iterative process,” Lam says. “I know it should be objective in that I’m trying to be the ears of all listeners,” she adds, “but in actuality it’s quite subjective because I’m in the equation.”
According to Hayes, that’s always been perfectly fine with audiophiles and the people who make gear for them. “There was an engineer named Daniel von Recklinghausen, who worked for H.H. Scott back in the 1950s and ’60s. He always said, ‘If it measures good and sounds bad, it is bad. If it measures bad and sounds good, you’ve measured the wrong thing.’ The sounds we can measure easily aren’t the same sorts of things the ear and brain naturally key in on. It’s a very different process than a bench test.”
John Marks of “Stereophile” could not agree more. “We have a very elegant hardwired system when it comes to sound,” he says of the connection between the ear and the brain. The involuntary startle reaction, he says, is a good example of the system’s sophistication. “When you hear a sharp sound off to the left, your eyes immediately begin moving in that direction, totally unconsciously. Your eyes know which way to go because of the head-related transfer function of the sound hitting one eardrum first and then, a certain amount of time later, wrapping around the head and hitting the other eardrum. The head-related transfer function gives you 180 degrees of horizontal localization, a survival mechanism hardwired from evolution.”
Ears also pick up sounds that bench tests miss. “The ear can hear very sharp, brief, transients, such as the start of notes, or the little scratchy sounds that the microscopic hairs on a violin bow makes,” Marks says. “All of these teeny, tiny sounds are not represented in bench tests,” he says, and there is no standard laboratory test for time distortions, which is the phenomenon of perceiving the passage of time differently depending on, in some cases, the pitch of what we are listening to. “Tubes handle the time domain slightly differently than solid-state devices,” he says, “and so, just on that listening basis alone, many people find that they prefer the sound of tubes.”
Charley Hansen probably wouldn’t agree with that last statement, but he’s definitely not a fan of bench testing to determine sound quality, although he has plenty of equipment designed to do just that. “We’ve probably got $100,000 worth of state-of-the-art test equipment at Ayre, but I have not found any single measurement that correlates with what you hear, not one. If you were going to measure something, I would say there’s only one thing to measure that makes any sense, and that’s the brain activity of the listener.” To learn how music affects our brains, Hansen recommends the PBS documentary “The Music Instinct,” which you can buy here.
In a way, Manley’s and Lam’s search for unamplified reference points through critical listening and product iteration is an attempt to get closer to what stimulates brain activity, and makes people like me think the sounds I’m hearing are warm. Once those reference points are established, Manley and Lam can listen to anything they like, stimulating their brains with jazz, techno, or whatever, secure in the knowledge that what they’re hearing is the real deal.
In other words, if they know what a soprano like Elly Ameling is supposed to sound like when accompanied in an unmicrophoned concert hall by a pianist like Dalton Baldwin, and if the recording they have of Ameling and Baldwin performing Schubert duets sounds accurate to Lam’s ears when played on an SG1.1 turntable by Spiral Groove (equipped with a Lyra Etna cartridge) that’s plugged into a VTL TP-6.5 Signature Phono Stage, which is connected to a VTL preamp called the TL-7.5 Reference Linestage before being sent to two VTL Siegfried Series II Reference monoblock power amplifiers, each of which is wired to its own Wilson Alexia loudspeaker, then it’s probably okay to be blown away by the sounds of Daft Punk pumping through the same gear.
As I was when I spent the better part of a morning with Manley and Lam at their home in Northern California, coincidentally close to where I happen to live. I had been talking over the phone with Manley about harmonics and with Lam about those listening sessions she subjects herself to in order to improve the circuitry in VTL products, but it was time to move beyond the rhetoric of sound and experience it. Graciously, Manley invited me to the couple’s home to hear what a good tube system could do to sound, and he even suggested I bring along a few of my own musical reference points so I could compare apples to apples.
At Manley and Lam’s, we listened to music from a variety of sources—vinyl records, CDs, high-quality digital files—and from a range of genre, including jazz, classical, rock, Americana, and electronica. Among other things, I brought along my vinyl copy of “Speaking in Tongues” by Talking Heads, which I had purchased in 1983 in part for its hit, “Burning Down the House,” but also for its Robert Rauschenberg-designed packaging. I also came armed with a Hot Tuna CD, a reference point from the days when I was happy to have guitarist Jorma Kaukonen and bassist Jack Casady do their best to destroy my hearing at Winterland. And since we were going to be listening to analog gear, I brought some music by Mr. Analog himself, Neil Young, albeit a digital CD of his “Living with War” album, whose opening track, “After the Garden,” is introduced by a handsomely distorted, unaccompanied, three-chord electric-guitar riff that’s crowned by a textbook-perfect drum fill and punctuated by an emphatic cymbal crash.
I knew what this stuff was supposed to sound like, or at least I thought I did, until I got to Manley and Lam’s. There, we listened to an engaging Natalie Merchant tune off “Leave Your Sleep”; a haunting song by singer Rhiannon Giddens of the Carolina Chocolate Drops from “Lost on the River: The New Basement Tapes”; Cécile McLorin Salvant having a whole lot of fun with a jazz number called “Nobody”; an unreleased digital track by The Swell Season, the duo that won an Academy Award for a song from the musical “Once”; a portion of a Sibelius symphony; the aforementioned Ameling/Baldwin duet; pianist Grigory Sokolov playing Chopin; and violinist Anne-Sophie Mutter delivering a completely unironic rendition of that classical-music cliché, “Carmen Fantasy, Op.25 – 1. Moderato,” backed by James Levine conducting the Vienna Philharmonic. Plus, of course, Talking Heads, Hot Tuna, Neil Young, and Daft Punk (for what it’s worth, we heard “Giorgio by Moroder” rather than the insufferable “Get Lucky”).
For me, Neil Young’s guitar had never sounded crunchier, Salvant’s intoxicating voice made me want to drop everything and start listening to jazz again, and I even momentarily forgave Daft Punk for those idiotic costumes they insist on wearing everywhere. This, I thought, is what music should sound like—bold, intoxicating, and alive, the audio equivalent of a crowd-drenching cannonball in a cold mountain stream on a hot summer’s day.
“It was flat,” Manley says, when I ask him a few days later to describe the sounds I had heard produced by his hardware and those Wilson Alexia speakers. “By flat,” he explains, “we mean flat frequency response, responding fully to the audio, whatever its frequency. The reason I stay away from words like ‘warm,’” he adds, “is mainly because when somebody says they’re looking for a warm-sounding amplifier, it suggests that the person is really looking for something that doesn’t do very much at high frequencies. Warm implies sound that’s rolled off at the top, which would be similar to saying it’s ‘sweet.’ It’s an overemphasis on the midrange, which is not what we’re after. Both warm and sweet immediately make someone think that tubes are necessarily retro or euphonic-sounding. I don’t think they have to be.”
But I like euphonic, I want warm, and I want a stereo that sounds like the one I heard at Manley and Lam’s, even though I’m pretty sure the almost 200-grand price tag for all that gear (“Oh, a budget system,” remarks Ken Kessler, when I tell him about my listening session) will eternally be beyond my means.
You don’t want that, says Charley Hansen.
“When you went over to Luke and Bea’s place,” he says, “I’m sure they had a killer-sounding stereo, probably a lot better sounding than your stereo, right? What you have to do is live with it for a few weeks, if not months, so that the thrill, that initial dazzle, wears off. Once the thrill is gone, you will have achieved the baseline for whatever level the system is. Now you are ready to really listen to it. That’s when you should put on your favorite song.”
The same sort of thing happens all the time, Hansen says, when people go shopping for stereo gear. “In a 5- or 30-minute demonstration in the sales room, the guy goes, ‘Wow, I want to buy that!’ But then he lives with it, and six months later it’s for sale on Audiogon.”
The problem, Hansen believes, is that people are too fixated on the gear and not enough on the experience the gear is supposed to provide. “Don’t be thinking about how the treble or bass sounds, or this or that detail, the ‘soundstaging,’” he counsels. “None of that matters. After the song is over, ask yourself, ‘Was I completely sucked in? Did I forget about everything else?’ Or were you thinking about the bills you need to pay, the deadline on this article. If you can get completely lost in the music, then it’s a really good stereo system. It doesn’t matter if it’s tubes, transistors, or a hamster on wheel. All that matters is that you got lost in the music. If you’re listening to music and your system doesn’t do that for you, your system is broken.”
I don’t know, Charley: I think I do want that. I’m pretty sure matching pairs of VTL Siegfried Series II Reference monoblocks and Wilson Alexias would improve the quality of my life, and I’d be willing to bet serious coin that I’d also be more than satisfied by the sound of a hi-fi powered by the VAC Statement 450 iQ, which is filled with eight KT88s, six 6SN7s, and will let me know when any of those tubes are about to go south. No doubt I’d also enjoy living with any of the R series solid-state products made by Ayre Acoustics, and it would be an absolute blast to go to Audio Classics for a day to hang out with Steve Rowell and compare the sound of an original McIntosh MC275 amp from 1961, the Gordon Gow Commemorative version of the 275 that was released in 1993, and the 275 the company is still making today. And last but definitely not least, it would be an audio dream come true to replace the seven 12AX7s and 11 other tubes in an old Fisher 500-C to see if it sounded anything at all like what I think I remember back in high school.
A week to the day after my listening session at Manley and Lam’s my grandmother died. She was 12 days shy of her 100th birthday, so her death did not have the same sense of tragedy that the premature departure of someone cut down in their prime would have had, and she had not been herself for years, so in many ways, her passing was a relief. Still, I was taken aback by how it affected me, even though it had hardly been a surprise. When it came time for me to say a few words at her funeral, I could barely speak.
Music would play a major role in my state of mind over the next few days, beginning the night after she died, when I decided to keep a date I had to see Willie Nelson, who was performing nearby. Sitting under the stars in a hillside amphitheatre, listening to opener Alison Krauss trade vocals and solos with guitarist Dan Tyminski and the other members of her band, Union Station, I wasn’t thinking about the bills I needed to pay, the deadline for this article, or even my dead grandmother. I was happily, blissfully lost in the music.
And then something else happened that was different than being lost, beyond even warm. At one point in Union Station’s set, Tyminski and the rest of the band left the stage, leaving Alison Krauss to sing an a cappella rendition of “Down to the River to Pray,” a traditional American gospel tune made popular in the film “O Brother, Where Art Thou?” For some reason, every time the song came around to the chorus when Krauss sang the words “Let’s go down,” I got a chill, goosebumps. It was like someone had flipped a switch, and it happened for the rest of the song, for all half-dozen or so return trips to the chorus, but only on the first mention of the word “Let’s,” which is repeated several times in the chorus. Always the first mention, always a chill.
I don’t know if it was tubes, transistors, or a whole lot of hamsters powering the PA system at that hillside amphitheatre, but at that moment, the sound of Krauss’s voice was definitely what I would describe as warm. No doubt my brain was making a connection between this song about praying and the death less than 24 hours before of a person who had meant a lot to me, but my grandmother was born in Bombay, was a practicing Sephardic Jew, had roots in northern Iraq, and could speak Aramaic, for God’s sake, so it’s not like a Southern American spiritual like “Down to the River to Pray” reminded me of her. For that matter, it barely reminded me of me, but somehow in that amphitheatre, the amplified voice of Alison Krauss was warming my heart. The moment didn’t have a thing to do with tubes or transistors—it wasn’t even about harmonics or reference points. I was lost in the music, regardless of what I was hearing, what you might choose to call it, or how it had been produced.
After the show, were my ears ringing during the walk to the car? Honestly, I don’t remember.
Another great and informative article, Ben.
A good piece of writing and very factual. I am an audiophile with both tubes and solid state. I started in 1965 with a Fisher 500C also.
Thank you for that wonderful article. I liked your style. The points and counterpoints, opinions, and facts painted A picture of tube sound. Also, I found the article to be most constructive, rather than confrontational and destructive as most opinions and articles in this Hobie
I feel I’m so much in the same spot…. never a serious audiophile, but a music lover that’s damaged his hearing over the years going to concerts of all kinds, now looking for that, yes, warm sound. I understand exactly. I’ve been wondering for some time now if I’d just damaged my ears, or the PE tubes I now have are seriously changing the frequency response of my hearing. I think about a moment when I heard a fellow with an acoustic guitar in a Hoboken house sing to a little girl, with the wonderful acoustics of that room, how wonderful that sounded, warmth indeed.
Crossover distortion is more noticeable and annoying than harmonic distortion.
Transistorized amplifiers make much more crossover distortion, especially at moderate sound levels.
This is the main reason that tube amplifiers sound better.
– W8MIC Lee – retired electronics engineer near Cleveland OHio.
Well Done! I build amps a little bit and you have given me enough courage to try a real live Tube amp, Thanks for the memories,
It doesn’t make any difference whether it is a valve amp or a transistor amp if it is overdriven it will produce a square wave which is asking your ear drum to go from one extreme to the other instantaneously. (Think Jackhammer.)
Sad to say but American sound companies are renowned for overdriving and ringing ears.
Shout out to anyone who went to a Simple Minds concert in ’83 or ’84.
If the Monitor engineer (Me) knows his stuff the sound onstage will be crystal clear and all the bozo on the desk out front has to do is amplify it and get his sound effect cues right.
You were slammed by the sound but went out into the night listening to the crickets and could hear loads of other people talking.
Yet they always pay the Monitor guy less and consider him lower than a lighting designer or a Front of House sound guy which is why I’m a french farmer.
This blog contains a great deal of unverifiable presumptions and misinformation. There appears to be a pronounced display of ignorance to what knowledge has been attained since the introduction of solid-state circuitry. Adding to this, listening tests performed by Dr.Floyd Tool, and many others in the field, have identified that the human auditory system’s thresholds preclude that distortion below 1% and a variance in frequency response less than 3dB in music playback are undetectable, 1dB variances with tones. The claimed differences that people hear among audio products are predominantly the results of expectation and confirmation biases. This is present in all uncontrolled and sighted listening events.
Below, I will address but a few of the outstanding issues:
There is a convoluted statement that tubes are regarded as “being less likely to create harmonic distortions at higher frequencies”. It renders the reader unable to determine if the intention was “being less likely to create harmonic distortion of higher orders” or “at higher input signal frequencies”. In either case, the claim would be made without referenced qualifiers, and both would be false. Empirical testing has shown that the harmonics and inharmonious exhibited by properly designed solid-state amplifiers reside magnitudes below those of tubes amplifiers, and below the hearing threshold. This is exacerbated by masking effects. Furthermore, high-orders reside below the noise floor, where they become indistinguishable from shot and thermal noises. The only exceptions in these respects are poorly designed amplifiers.
A claim being made in the blog, is that tubes are excellent voltage amplifiers, and that an increase in “input going into the tube, you end up with an almost perfectly proportional increase on the other end”. Does this not also apply to aforementioned solid-state amplifiers? It certainly does, as evident through analysis with an Audio Precision One, Hewlett Packard 4192A, R&S UPV, and any qualified analyzer with a scope function. All amplifiers compress the audio signal, and solid-state’s magnitude of compression, below clipping, is a fraction of that present in tubes designs. To understand this further, one has to realize that harmonic distortion is the result of added frequencies to the fundamentals, or else it is generated by the compression of the output signal. When an amplifier clips, it produces harmonics because the crests of the waveforms are reduced in magnitude. Properly engineered solid-state amplifier’s demonstrate harmonics as much a seventy decibels lower than the claimed high-end tube designs in this blog, and will only produce appreciable compression when they begin to clip.
Another detractor of any credibility from the sources named in this blog is the following statement: “It takes a small handful of solid-state parts to replicate what a tube does. Each of those parts can introduce an acoustic artifact that must be corrected…” That is a serious error in judgement and evidence of a lack of scientific rigueur. The goal of solid-state has never been to replicate anything other than the source signal; the purpose has always been to recreate music as true as possible to the recording. That can only be achieved by diminishing artifacts to magnitudes below human auditory thresholds.
Continuing the previous sentence, “…forcing solid-state circuit designers to employ tricks like negative feedback, in which the output of the amplifier is re-injected into the input to achieve the inherent linearity produced by tubes.”. That line contains another misnomer; negative feedback is not being used to mimic tubes. It is used to improve bandwidth, and science has afforded engineers the good fortune that careful application of feedback results in improved harmonics, intermodulation, bandwidth, impulse response, square wave response, phase margin, output impedance and more.
On to the next: “Negative feedback is actually widely used in almost all amplifiers, be they tube or solid-state, although Hansen does not resort to it in Ayre products, and thinks it’s a detriment to either technology”. Any credible licensed engineer that works in signal processing (and not audio product design) will be aware of the fallacy in Hansen’s belief system. It is false and makes him appear quite ignorant and uninformed. I will add that transitory intermodulation distortion, TID, is not the result of negative feedback – it is directly induced through miller capacitance. A group of graduates looking to make a name for themselves in the 1970s renamed and rediscovered this distortion, then calling it transient intermodulation distortion (TIM). This distortion cannot and never did exist in properly designed amplifiers, because severe slew-rate limiting is a perquisite for its existence – although not particularly well conducted, Jan Didden’s thesis was among several that demonstrated what was already known for decades. NFB is off the hook.
“‘The ear can hear very sharp, brief, transients, such as the start of notes, or the little scratchy sounds that the microscopic hairs on a violin bow makes,” Marks says. “All of these teeny, tiny sounds are not represented in bench tests,” he says, and there is no standard laboratory test for time distortions, which is the phenomenon of perceiving the passage of time differently depending on, in some cases, the pitch of what we are listening to.'” Another utterly false series of statements which are the tell tales of ignorance. To the surprise of many, the highest frequencies of most instruments, including the violin, is 20kHz. This event occupies 50 microseconds, thus the rise is a corresponding 12.5uS. Therefore, those “tiny sounds” are not infinitely fast. The largest number of “zero feedback” marketed tube amplifiers do not extend linearly beyond 20kHz, while even a modestly built solid-state can achieve 100kHz; many solid-state amplifiers extend beyond several hundred kilohertz. This is easily quantified and demonstrated, leading on to question the intent of the original statement made by Mark. Here is a reference article for further reading: https://courses.physics.illinois.edu/phys193/Student_Reports/Fall03/Tammy_Linne_Andy_Schurman_Ivy_Thomas/Tammy_Linne_Andy_Schurman_Ivy_Thomas_Phys199pom_Final_Report.pdf
The extant claim being made by Hayes, “The best triode tube is very close to perfectly linear”, is a play on words. The best tube ‘would’ be very close to perfectly linear, but none are. Lastly, “Hansen’s company in Boulder, Colorado, is highly regarded for the magic it works with capacitors, rectifiers, field-effect transistors, and other solid-state goodies, without resorting to feedback to achieve linearity. Hansen’s wizardry with solid-state audio circuitry is…” blatant misinformation to confuse his market demographic. This should be science, not pulling white rabbits from black top hats. In his ignorance, Nelson Pass has made the same claim about his solid-state First Watt products. Each of them employ degenerative feedback on the Mosfets’ source to set its Q-point along the load line. The same applies to cathode biasing on triode amplifiers.
The article is beautiful and covers many topics in a simple to understand form i.e. excellent communicative skills. The AES nerdy response is typical of sand based critics.
Why is it that sand amplifiers measure perfectly but do not play music at all? A 2A3 based Class A1 amplifier using an amorphous core transformer will also give a 200 kilohertz output. You do not require any more than 15kHz for top class reproduction but the system does need the 200 kHz speed (dv/dt) otherwise it will sound slow.
If you adopt synthesized power supplies, it will sound faster than a MOSFET type amplifier and gives a wonderful tonality to boot.
That’s not an article. Article’s undergo peer reviews by appointed committees for academic authorization. That could have avoided the quotes in the so-called “article” being defaced on several forums.
Typists are responsible for what they’ve published.
A 2A3 doesn’t have a bandwidth of 200kHz without feedback. Anyone that thinks it does is showing their ignorance. An output transformer might have a wide range, but that doesn’t imply the active circuit is linear.
The required bandwith is determined by the maximum rate of change at full voltage swing, not an arbitrary number like 200kHz. It’s not a linear relationship. This really shows how little audiophiles know.
A 400W@8Ohms amp requires 10V/us to accurately reproduce 20kHz at full power while a 100W one requires 5Vus (Douglas Self; Audio Amplifier Design Handbook). A 2.5W 2A3 hardly needs a slew rate of 1V/us – vinyl has a maximum rate of change of half a volt per uS. Extending the bandwidth too high increases its susceptibility to RF interference.
AES’s comment is a more worthwhile read than the “article”.
To learn more about the Audio Industry’s blatant misinformation campaign and educate yourself, I suggest visiting the Hydrogenaudio forums. This is a forum for the factual minded. No subjective posts allowed.
Class-A solid state amplifiers weigh as much and make almost as much heat as a similar-power vacuum state amp, and sounds the same kind of good. SS amps survive moves better and warm up quicker. Vintage gear has vintage styling: horses for courses.
Nice article. Thanks.
A fine ending to your article, but I believe that your experience is why we have music in our lives. Thank you for this Brother Man…
An amplifier, tube or solid state, is part of an audio system. What reaches a listeners ear is the result of many, many factors: source material, playback equipment, amplification, signal and power cabling, of course the speakers, and even the room. Instability of the wall current can also impact the performance of the system. So matching system components to each other and to the room is part of the art of creating a satisfying musical experience. I have heard very expensive ($300K) systems that, for all of their purported “accuracy”, do not move me and I’ve heard systems that cost 50-60 times less that let you merge with the musical performance. I’ve heard Ayre and VAC and, with the right setup, both can offer an extraordinary experience.
By the way, this business about being above or below the threshhold of hearing is another fallacy. Sub-woofers go below that threshhold and super tweeters go above. We hear not only with our ears, but our bodies receive vibrations above and below and our brain integrates that information with what the ear does register. Enigma Acoustics makes a super tweeter that goes to 100,000 Hz. Plug it in to a highly resolving system and you can easily tell the difference in attack and decay.
Disclaimer: I own a VAC integrated amp.
Tube amps can make your ears ring just the same as SS. Why even go there?
I read your article with great interest, I am crazy about my Fisher tube amps and receivers.
I am an avid builder of Altec speakers and love their Horn sound, I have a pair I built in my bedroom with a Fisher x-101-b amp.
I have an Audiophile friend that comes to my house from time to time for listening and discussions, he always wants to go to the bedroom to listen to that system.
He always says it has a liquid sound, I really don’t know where that ranks or what it means, but it does have a sweet, life like, none fatiguing sound that you can listen to until you fall asleep.
As I read the article I came to the end about Alison Krauss and the song down to the river to pray, I just happen to be listening to that very cd and always love it when that song comes up.
This time I listened differently, she is an amazing young woman.
Great article and + info! I have a Tubed Marantz with BW speakers, love how it roars!
it’s funny, i purchased an old (1952) tube AM clock radio to re-cap and restore so i can listen to the Red Sox play by play next season. my grandmother had the same model clock radio, and used to listen to the games, and talk radio. i am also looking for “warm” sound, and a link back in time.
in reading about your audio adventure and describing the types of music you brought to listen to, my first thought was “i wonder how Alison Krauss would sound on that system, probably blow his freakin mind…” she has one of the most beautiful voices i’ve ever heard. so it’s funny that you end the article with her.
bottom line? SS or Vacum Tube? I am so confuse. please help me make good smart decition. I am listening to hi res music only from my computer .
Hello, thank you very much for this article. I grew up listening to quality symphonic and popular music on my Dad’s Fisher tube receiver, from the mid 60s. Right now, the unit is in the garage. I am afraid to handle it, turn it on, plug it in, connect it up…..will the tubes explose? Please advise. It does NOT have a covering wood case.
A Very informative article! I actually believe that adding ‘Old School’ amplification does make a difference. There seems to be more of a ‘Deepness’ to the music that comes out.
having read this article before and now revisiting the comments, i cannot but agree with the theory and facts in the comments section. No i have to admit that i’ never go back to SS or even Class D, unless it is a Class A implementation in the case of the former – such that impresses me, or a very careful implementation in the case of the latter, which has proven to impress me easier for much less money spent. I will remain in the valve camp, accepting and enjoying hybrid designs, which deploy Mosfets in the output and this is because of the following key reasons :
1. Mostly all (non class A) SS amps give me a ear fatigue after a while and i want the volume down, even if the volume was not much to start with !
2. I can count on one hand (and have fingers in reserve) the times that the soundstage components impressed me in SS – usually i am missing depth and hight mostly and sometime width too.
3. As in 2 above – very much the same applies to notes decay, and air around instruments/notes – in the SS amps i’ve heard, the decay is notably shorter than a good, and often cheaper valve design, and unlike how instruments do sound in real life
This is it.. These 3 above made me move away from SS and have great moments and hours with valve – Jadis, Leak, Ming Da, Copland – to mention but a few ..
Finally, it may sound way too harsh, with a ping of dichotomy, yet, when i listen to a valve sound, i’m hearing the music – outside and inside the envelop, when i listen to SS, i hear the technology reproducing music, almost as good as the valves
” …but transistors represented a major improvement over tubes, even if it was a hassle to replicate their linearity, for the simple reason that unlike tubes, solid-state components did not wear out.”
Um, from what I recall from my various electronics classes, and a lifetime of reading on the subject, and the comments of my grandfather, whose company’s products you referenced several times in the article, one of the biggest attractions of going solid state, once the manufacturers got the production processes sorted out, was that it was *easier* to achieve linearity across the audio passband with transistors than it was with tubes. Transistors were also longer-lived, more predictable, more reliable, and less expensive power-wise to operate. It wasn’t just the “desire to sell something new”, it was the desire to sell something *better* that drove manufacturers away from the firebottles to silicon.
I might also point out that MC-2300s you referenced in your comments on tube lifetime and tube replacements and the Grateful Dead are solid state amps.