You are reading the older HTML site

Positive Feedback ISSUE 23


Max Audio: HeadRoom's Desktop Millett Hybrid Amp and Desktop Power Supply, & Grado's 325i Headphones, with Some Reflections on Tubes vs. Solid State Amplification
by Max Dudious


HeadRoom's Desktop headphone amp has sound that is drop-dead gorgeous as I had hoped, and it is something I've been waiting for ...for a long time. I'll spend the rest of this little essay telling you how it sounds to me, why I think it sounds that way, and why I'm nominating it for my prestigious Max Dudious "Product of the Year" award. (Better late than never.) But first, some words about the 'phones I'm listening to most of all these days: Grado's model 325i.

Grado's 325i

When I asked John Grado what the i stood for regarding his new headphones, he got coy with me. "Guess!" he said. After I tried some silly i-words that came to mind (inimitable, invincible, irresistible, ineffable, insouciant, and incrementally innovative), I could hear his tongue practically pushing through his cheek. Finally, when I laughed that I'd had enough of the game, I coaxed him to tell me. "Improved," he said. If you've never heard the sound of one hand clapping, you should hear me slap my forehead sometime. Set myself up for that one. Duh.

And improved they are. I used to think they were "hot," in their previous incarnation, as model 325. Improved, they are not hot. I used to think they were too analytical with their rising high end. Improved, their high-end stays where I think "flat" is supposed to be, and they are not overly analytical. Improved, they seem neither bright nor brash, grumpy nor sneezy; nor doc: they seem without any common coloration, yet they still have the liquid midrange, the rock-solid bass, and (perhaps above all) what TAS's Dan Schwartz calls "hear-through-the-veils transparency" that characterizes most all of the Grado line. The improved 325i's tonal balance is much more like the Sennheiser 650s, yet with those qualities of presentation that have come to be known as "the Grado sound." I have noticed some of the magazines that cater to the recording studio sub-industry have touted the 325i as a must-have for recording engineers (good audio engineering). They are also tough enough to withstand some studio abuse, with their aluminum shielded ear cups (good product engineering).

In short, I feel the Grado model 325i headphones sound very detailed (without having a presence peak); more than most other quality headphones, they exhibit flatter frequency response and still retain their big bass, their voice-friendly midrange, and their silky highs. In addition, they have retained their relatively low (32 ohm) impedance and their relatively high (98dB) output efficiency for one milli-volt, so they are compatible with everything from the lowliest walkaround CD player to the highest priced tubed headphone amp. And, with swiveling ear cups, they travel well. If I were designing a pair of headphones, and I were trying to decide on my design goals, I think I would have chosen just those that John Grado and his gang have chosen. And that is why I have conducted all my recent listening tests through them. Due to their superior product-design, portability, and sonic performance, I hereby nominate the Grado 325i to Max Dudious's vastly prestigious "Product of the Year" list. (Actually, my 2006 list. I was always late with term papers.)

Some Prefatory Opinions

Interestingly, I still have around the lab the HeadRoom Micro-Stack that I recently reviewed, so I could A/B compare the transistorized Micro-Amp to the tubed Millett Amp. But before I get into that, I'd like to explore the notions usually cited as the defining characteristics of "tubed sound" and "transistorized sound." With regard to larger amplifiers, it is usually the case that tubed amps have a "damping factor" between single digits and, say, twenty to thirty; while transistorized amps can have "damping factor" computed in the hundreds. I think I remember one amp's advertising campaign claiming a "damping factor" of 1000. We might say, and not be too far off, that transistorized amps have damping factor nearly an order of magnitude stronger than tubed amps, and in some cases two orders. For our purposes damping factor might be defined as being analogous to braking horsepower, or the grip the amplifier has on the drivers, or a measure of how quickly the amplifier stops the drivers from "ringing" after the a pulse signal stops.

Tubes vs. Transistors, yet again

When coupled to a high-resolution set of headphones the differences between a tubed amp and a transistorized amp are readily apparent. Headphones demonstrate these differences in damping very aptly, better than, say, a pair of high class loudspeakers, in a good system, in a room with all its inherent echo and reverb, nodal points and cancellation problems. Moreover, loudspeakers typically have multiple drivers, crossover networks, and all the phase and gain problems therewith associated, while a headphone usually acts as a single-driver point-source, with no crossover, and with none of its attendant problems. HeadRoom's Tyll Hertsens says (and I'm paraphrasing a telephone conversation here), headphones, with only a few cc's of air between the driver and the ear, are the cutting edge of audio resolution, the best you can get from the industry, even though they obviously can't deliver the tactile zotz in the chest that a large scale system can deliver at full blast. I agree. If you want to hear what the music is doing, or what other pieces in your system are doing, or even what a pair of 18" interconnects are doing, you can't beat headphones. If you want to hear an approximation of an orchestra in a symphony hall, or a rock concert in a ballpark, you can't beat a big rig.

Acoustic vs. Electric surprise

Now, consider the differences between an acoustic bass and an electric bass. With a stand up, wooden, acoustic bass we hear the plucking sound (or attack), the sustained pitch with all the harmonic (woody) overtones that characterize such an instrument, and the decay to eventual silence. These three features – attack, sustain, and decay—exist in the sound of an electric, say, Fender bass as well. Without any of the various pedals that produce electronic artifacts, such as wah-wah, fuzz-tone, echo, or reverb, there is the attack that can be achieved with a finger or a plastic pick. The attack of an electric bass is distinctly different than plucking an acoustic bass. For one thing, the shorter vibrating string on a Fender bass doesn't have as long a sustain as that of a free string on an acoustic bass. The result is, the rise of the attack, and the diminuendo of the decay are both faster on an electric bass. These two physical characteristics make for noticeably different sonic characteristics when comparing two such instruments. With an amplified Fender bass you get a "thump" of a particular frequency, say an A at 55Hz. With an unamplified acoustic bass, you get a textured, woody A at the same 55Hz, but with audibly different attack, sustain, and decay ratios. These three defining sonic fingerprints are what differentiate a bass viol, a contra bassoon, a tuba, and a kettle drum (tympani) from each other when playing the same note, A at 55Hz.

The plank of solid wood into which an electronic pickup is usually imbedded doesn't resonate, so what we hear from an electric bass is largely the fundamental tone. In contrast, the typical acoustic bass has a deep, rich, and textured sound, and we hear the pluck, the fundamental tone, its harmonic overtones, its long sustain, and it's more gradual decay. So the basic sound and the sustain of each of these instruments differ. Finally, since everything about an acoustic bass is designed to resonate, and everything about most electric basses is designed so only the strings vibrate into the pickup, the decay of the two instruments is likewise different. In short, the electric bass was designed to make low frequencies, and it does that much differently than its cousin the acoustic bass. They are different, and they sound differently.

Damping, under and over

Furthermore, when we listen to music played back through a transistorized amplifier, the damping factor of the amplifier takes away from the sustain, so the decay is even quicker. Consequently, the transistorized amp (some with fierce damping) exaggerates the characteristics of the electric bass. Similarly, but to a lesser extent, it does the same with an acoustic bass. The tubed amp (with its relaxed damping) captures the characteristics of the acoustic bass better. Ironically, through all but the very best transistorized amps it becomes harder and harder to tell the difference between the two basses, acoustic or electric. The higher damping factor of transistorized amps also has this effect on tympani and bass drums. The decay just doesn't seem to be as long as it is when played through a tubed amp, and the textures through tubes become clearer making the tympani and bass drum more easily identifiable. Additionally, tubed circuits seem to have an easier time swinging a lot of volts (have greater headroom) on dynamic passages, while typical transistorized circuits don't do as well. The dynamics of the two instruments are reproduced similarly through average transistorized gear. Sadly, a handful of instruments often sound pretty similar through average quality transistorized amps at the same pitch and volume. If they did a better job differentiating the same note played by different instruments, AM-FM stereo receivers would be great value for money.

Then there is engineer induced echo and reverb. Sometimes, while listening to classical pieces such as Bach's highly recommended Coffee Cantata, (The Academy Of Ancient Music, Christopher Hogwood, conductor; L'Oiseau-Lyre 417 621-2), a piece calling for a soprano (Emma Kirkby) and a chamber group, you can tell that the performance venue was a church or a large recital room. Similarly, listening to small combo jazz such as Sarah Vaughan's quartet (How Long Has This Been Going On? Pablo, PACD 2310-821), it becomes pretty apparent that the spacious quality of the recording was a function of the record's engineering. Such classical recording sessions are often done with minimal microphones, and without any compression, or added echo or reverb. The natural acoustic of the hall and its decay (which is a function of its dimension) is what you get. In the equally valid rock and jazz world, most often the recording sessions are done in studios that have purposefully removed all identifying acoustics. The recording engineers want to control the reverb, echo, etc. in post-production mixdown. When this is done, echo and reverb units are patched into the mixer and the resulting sound approximates the sound of a singer in a larger room, say, a big nightclub, and it sounds like it.

When you play back such recordings through headphones, a transistorized amp will emphasize the sonics so much that I feel I can "hear" the engineers hand on the dials. It is as though the engineering is too "up" in the mix. Through a tubed amp the sound will emphasize the performer and the same mix seems more natural, further "down" in the mix. For me, that is an ideal to be striven for – the sound of music in its own environment with the engineering calling as little attention to itself as possible. That is, though I know the engineer has his hand on the dial, I think it ought to be as unobtrusive as possible. I want to hear the instruments, the voices, the skill of performance, not the skill of the engineer. The best engineer calls the least attention to his skill, but is in command. So sayeth Max the Dude.

[In audio there have come to be certain aphorisms that make sense only to other audio heads. Like, "The only good turntable is a dead turntable." "The best capacitor is no capacitor." "In theory, the best speaker cable would be infinitely short and infinitely thick." And so forth. "The best electric bass is a dead electric bass." You could make a Pooger's Almanac if you codified all of these. For sure, someone would embroider them onto canvas and sell them for framing through the back page ads of the hard copy magazines, and get rich. It does seem that I wrote something like, "The best engineer is no engineer." But, you know I don't mean that, exactly. I think I mean, "The best engineer is a shy engineer." Not too catchy. I've also heard a film editor say that, "The best film editor seems to be no film editor."]

Behind the curtain in Oz

I discovered forty years ago, when Saul Marantz allowed me to peek behind the curtain at a N.Y. Hi-Fi show, that in the demo room he was using a solid-state amp on a pair of KLH 12 woofers, and his tubed amps for the rest of the audio band played through electrostatic KLH 9s. I thought (being fresh out of college and possessed of bat-like hearing) that it was the best sound I'd heard. Sid Smith was there, a ruddy-faced guy in his late thirties; Saul Marantz seemed to be a gentleman in his late fifties. While Mr. Marantz was busy talking to folks, I got to talk with Sid Smith, who was Marantz's chief designer. I would become friendly with Sid twenty years later, when his daughter was a student at the Peabody Conservatory in Baltimore (my HQ), and I was doing modifications on my Marantz 7C preamp.

Sid Smith introduced me to the notion that one ought to have felicitously matched speakers and amplifiers. To his ear and his measurements, certain amps had certain rise times and decay times. Matching certain amps with certain types of speakers was an art. You had to be careful not to put certain types together that would result in too much damping and "dead" sound. You also might wind up with too little damping and "floppy" sound. So the game was to mix and match. Most woofers sounded better with transistorized amps, but a little woolly on tubes. Most electrostatics sounded better on tubed amps, but noticeably less shimmering on transistors. This was a rule of thumb.

During his last years Sid Smith listened to heavily modified Marantz tubed electronics through Quad 63 electrostatic speakers in his home. The importer of Lowther loudspeakers, Jon Ver Halen, recently told me that various types of speakers, not his, were under-damped and rang (had a too-long sustain); while yet others, not his, were over-damped and washed out inner detail (had a too-short sustain). These guys are not the only ones to attend to this distinction: seems like a lot of guys in the industry notice it. Mike Fremer, the vinyl guy at Stereophile, in his recent review of the "World's Best?" Continuum Audio Labs Caliburn/Cobra turntable, suggested that what made the LP player in question so good was its attention to the attack, sustain, and decay ratios. When some manufacturer gets that right we say the piece of gear has no personality of its own. And that's the highest compliment a reviewer can offer.

HeadRoom's Desktop Millett Hybrid Amp

There is so much information packed into this title that I think I ought to unpack it before going on. HeadRoom is the manufacturer headquartered in Bozeman Montana. Desktop identifies that its size is smaller than a rack mount, but larger than a micro-stack. Millett means the co-designer, along with help from Tyll Hertsens (HeadRoom's C.E.O.), was one Pete Millett, veteran audio designer of some pieces well-respected in only the best audio circles. Hybrid means that the amplifier has some tube elements and some solid-state elements. For example, the line stage is driven by one tube in each channel, but is buffered by a chip (HA 5002), like the other amps in the HeadRoom line. Amp means a dedicated amplifier designed for use with headphones with neither too much power (which would blow out some headphones), nor too little (which wouldn't be enough for high impedance 'phones). All together it means a dedicated headphone amplifier; small enough to sit on your desk; including Pete Millett's tube/transistor circuits, with some assistance from Tyll; and marketed by his company, HeadRoom.

If you consider a first-class headphone system as a first-class room system—in miniature; and you look at the driver in a typical set of 'phones; you'll find a miniature of a standard magnet/voice-coil/cone speaker—much like a miniature woofer. This might lead you to assume that it will behave like a typical woofer. But if you could measure the mass (or weight) of the moving parts, you might find the diaphragm is so light that it begins to resemble an electrostatic in rise-time, decay-time, etc. So, a good headphone driver may be more like an electrostatic speaker in its behavior than like a cone driver in a full sized speaker. If you can suspend disbelief for a minute, considering the above, you might conclude that a high quality tubed headphone amplifier might be a fortuitous match for most high-quality conventional (not electrostatic) headphones, and deliver really superior (not over-damped) sound.


Compared to Headroom's pretty damn good Micro-Stack headphone amp (all solid state), the Millett amp delivers in the ways I'd not quite expected. Subjectively, the rise time was pretty fast, and the decay time was somewhat relaxed, but not floppy. This made sopranos sound glorious right between my ears. No harshness, grit, or grain. Just the way they sound to me in concerts. Trumpets sounded brassy but with no extra tizz to the notes, and when called on for a bit like Harry James's big velvet tone, or when called on for growlly, or spitty, or "whiskered" with a whisper mute's tone, they delivered. Acoustic guitars sound like they do when I play them, with the right balance of attack, sustain, and decay. Bass viols have all the woodiness that is their texture, and the length of decay is right-on. Don't forget, my neighbor Erik Stahl plays the bass viol in our local orchestra, and he sometimes allows me to hear him practicing. I think I have a grip on what a bass sounds like, up close and personal. And my wife (Grammy Dudious) and I attend a good number of symphonic concerts, so we have a good idea of what the whole orchestra sounds like. (This past fall we even caught a concert in Barcelona's fabled Palau Musica Catalana concert hall, Christoph von Dohnanyi conducting the NDR Symphonieorchester Hamburg.)

Live music sounds damn close to what I hear through the tubed Millett amp. For a totally awesome experience, it must be hard to beat listening to Rachmaninoff's Symphonic Dances, Baltimore Symphony Orchestra, David Zinman conducting, (Telarc CD-80331), through my Optimus 3400 CD player used only as a transport, the HeadRoom Desktop Power Supply, the HeadRoom Micro-DAC, the HeadRoom Millett Amp, and the Grado 325i 'phones. It is, in my experience, the best sound I've heard from a headphone system so far ... glorious and magnificent. If you don't believe me listen to Telarc's new recording of the Vivaldi Gloria, and the Bach Magnificat, Boston Baroque, Martin Pearlman, conducting, (Telarc hybrid SACD-60651). If you have any brains left between your ears, check out the SACD hybrid recording of Pink Floyd's Dark Side Of The Moon, (Capitol, CDP 7243 5 82136 2 1-US). That'll getcha.

HeadRoom's Desktop Power Supply

This piece has the same physical footprint and build as the Millett amp, that is to say, they can be stacked. It is a quality power supply, with six outputs. I'm told the total current the li'l darlin' can deliver is about 350 milliamps, and I'll bet it is enough since pieces like the Micro-Amp and the Micro-DAC draw about 25 ma each, at the max.. The little wall-wart that comes with the amp seems adequate and fairly quiet at first listen; but when compared with the dead-quiet Desktop the wall-wart seems considerably noisier, and lacks the zotz the Desktop has. In other words, if you want a truly extraordinary headphone system the Desktop Power Supply ought to be included. You might make do with the wall-wart for a time, if you must, because you can't play the Millett amp on batteries—it draws 50 ma..

We know that wall-warts are adequate to do a passable job. But the Desktop is a higher quality power supply that regulates the voltage so it doesn't sag below +/- 15VDC no matter what goes on in the music; and has enough milli-amperes (350ma) that if 6 typical pieces draw from it they ought never want. In fact, they think, if they think, that they are seeing a considerable bank account of current they can draw on. This power supply is pretty darn good, and it has some interesting features. There is an optional "ground lift switch" which allows you to float the ground, purifying the current, to taste. There is a "power entry module" that allows the user to adjust for other international voltages when abroad for a time. And there are six "power outputs," all of the same voltage that allow the user to use any of the other HeadRoom Desktop or Micro line of amps and DACs.

I should mention that this system isn't designed for trips. None of the components can run on batteries as the Micro-Stack can. But for late-night listening, this dedicated power supply is the best I've heard.

Tube Rolling

This is an expression derived from log rolling, I think, but here it means swapping and evaluating tube sound for subtle differences. Eventually, theory goes, you can tune your system to get just the right amount of finesse or naked aggression that you like in your music. Three different pairs of tubes come with the Millett amp. According to Tyll, the somewhat laidback 12FM6, is a warm sounding tube; the most liquid 12AE6A, is a very nifty tube indeed; and the most neutral 12FK6, has the most extended top and bottom octaves. These tubes, used in this piece, ought to last a long time. But tubes age like light bulbs, using themselves up with use. When the shimmer and sparkle goes, it's time to replace them. When you find the one you like best, get some backups. With this choice, the tubes have more in common with each other than they do with transistors. The differences are subtle, but real. Like perfume. And once you get into perfume, those differences become huge.

Other Options

Class A: If you'd like, HeadRoom can upgrade your Millett amp with a special version of the Home Module. This mother board is a beefier circuit biased up to class "A" for overall cleaner sound, but it also draws more current and can only be used with the Desktop Power Supply. In my experience with other amps, I find class "A" runs a bit hotter, but that is worth the improvement that comes in very loud passages where there is no congestion and each instrument stands out well defined from the other instruments at full cry. If you make it a point to never listen loudly, class "A" is not for you. But who gets a fine system to play at very low levels?

Stepped-Attenuator Volume-Control: If you'd like, HeadRoom can upgrade your Millett amp with a special discrete-resistor stepped-attenuator volume-control. This means each of the stops on the 24-position switch has a resistor of specified value. This is a cleaner sounding way to do a volume control than even the best continuous tracking pots, and the tracking from right to left channel is usually extraordinarily close, for better channel separation and elimination of crosstalk. Better channel separation makes for better imaging. The quality of sound cleans up a noticeable, asymptotic bit. And it is the cumulative value of these bits that make all the difference. Just as the slightest bit of crud is audible in the signal path of a headphone system, the reverse is also true. Every time you are able to subtract a veil it noticeably improves the sound. Adding to the quality by subtracting a veil is harder to notice, at first. It is easier to notice an improvement by addition, because better bass is usually bigger bass. Addition by subtraction is subtler, but you will notice it, because you'll hear those inner details as you never heard them before.

Some Final Thoughts

This piece, the Desktop Millett Hybrid Amp, is another great piece of product engineering by the gang at HeadRoom. It is flexible! You can get it with or without the Desktop Power Supply, though I'd recommend with. You can play it with three different sets of tubes, equally intriguing. You can get it with a "Class A" driver board, or the regular "Class AB" (I guess) board. And you can get it with a good quality pot, or an even better discrete-resistor stepped-attenuator 24 position volume control pot. I think this makes it a highly desirable piece at its price point.

 To be a genuine contender for "state-of-the-art" sound, there are some headphone amps out there that sell for $15,000!?! I don't know what is in them to warrant such a price. I know what's in the HeadRoom piece for its price, and even with the class A upgrade, the stepped-attenuator volume control, and the Desktop Power Supply, it is still a bargain if you plan an assault on "state of the art." As I've said, so far as I've heard, it is the best piece of its type, the one that matches my preferences the best, anyway. Your age, ears, brain might not match up with mine, completely. I've listened at some lengths to some very expensive systems (saving you the time), and I truly believe Headroom's Desktop Millett Hybrid Amp, used as I used it as described above, is a genuine contender for state-of-the-art. I think you will agree if you can get to hear this system.

If you are seriously interested and you want more information, go to Pete Millett's website ( for his technical stuff and DIY projects, and see the HeadRoom web site ( for more details and prices. If you want to order any of the pieces I've run on about in this article, phone HeadRoom at 800. 828. 8184.

And don't overlook the new Grado 325i 'phones either. These will play well with anything, tube or transistor, dedicated headphone amplifier or a Walkman's amp out. They have the "Grado sound" along with much flatter frequency response. I'd say they have the all of characteristics anyone would want for listening to "pop," "rock," "jazz," and classical music. Their personality is they have no personality. For more information see and use it to find your nearest dealer. This is a truly well-engineered product.

And when you and your friends go down to your local audio boutique in a conga line, don't forget to tell 'em Max Dudious sent ya'.

Ciao Bambini!!