|State of the (Non) Union addresses. They're useful. Hindsight commentary creates context going forward. Being the middle of December, I had just finalized the awards graphics for this year's winners in our pages. This de facto overview of 2007 reminded me of component highlights. Where we'd been. Then something struck. There was a highly unusual peak (well beyond linear) of class D amp winners this year. All of 'em -- with only one exception -- were based on lower-power Tripath chips. Had all this occurred by way of the same writer, one would cry personal biases, system context and write it off. Which is why it would never have happened in the first place. What's more, each of the actual five writers awarding class D amps adores and owns valve amps. Let's illustrate this spike with hard figures:|
|That's 9 class D amps out of 13 amps total, 8 of 'em Ts. Doesn't that give you pause? It sure does me. Adya Tripathi's corporation goes bankrupt, Cirrus Logic acquires its assets and Tripath amps go from strength to strength. Granted, this occurs well outside the mainstream. On one side we have the Kingrex, Sonic Impact, Trends and Winsome Labs offerings for a few hundred bucks each. Their appeal is mainly to the cheapskate brigade, with the well-off bothered by perception, lack of bragging rights and suitable speakers. On the other side we have Firenze Audio and Red Wine Audio which offer more power and battery drive. There's less mess-with-perception hurdles but still few bragging rights.
There's more on this topic. In our 5+ years of operation, we've bestowed our Lunar Eclipse award only four times. We apply it sparingly to truly mean something. To be eligible, numerous factors have to align just so. Simply being very good but ultra expensive won't cut it for example. Out of these four awards, two have been for speakers, two for amps. And both amps are class D, one ICEpower, one Tripath. Last year, John Potis voted for the Bel Canto Design e.One Reference 1000s. He also owns 300B push/pull and 845 SET amps and -- among a lot else -- the 96dB sensitive Hørning Perikles speakers to go with 'em. This year I voted for Red Wine Audio's Signature 30.2. I also own Yamamoto's A-08S, Fi's 2A3 monos, Melody's I2A3 and SP3 and Eastern Electrics's M520. For context, John is my senior reviewer with the (here plus elsewhere) most reviews under his belt of any of my staffers. That's why bestowal of Lunar Eclipse awards thus far has been limited to him and me. It requires experience and context. Back to D-class amps.
David Kan this year awarded the Kingrex, Trends and Winsome Labs offerings. He owns a bevy of tube amps by Dared but also chip amps by AudioZone, conventional solid-state by Symphonic Line and high-power class D by NuForce. Marja & Henk awarded Kharma's small D-class monos. They own an Audio Note 300B integrated, a bevy of 300B tubes for seasoning and a Moscode hybrid. Do we see a trend in the making?
On the personal front, FirstWatt's F1 and F3 gave way to a pair of F4s as my current transistor amps of choice. They combine the highest resolution in relaxed and natural form of any amps in my house plus -- by virtue of unusual transparency to what precedes them -- whatever valve-sourced effects a first-rate preamp like ModWright's LS-36.5 or Supratek's Cabernet Dual adds. Things have developed to a point where, most days of the week, I prefer to listen to my Rethmic Lowtharios, the Saadhanas, over the F4 rather than tubes.
In a related vein -- tube man discussing transistors -- my reviews of AMR's AM-77 180wpc hybrid and Coda's 450-watt CX monos reconfirmed the old adage that high-power amps sound better on speakers which need power. Accept the exceptions though. John's encounter with the 1000-watt Bryston 28B-SSTs proved something unexpected. Bryston's claim for true load invariance, sonics as good at 1 watt as at 500, finally came home. One tends to think of load invariance in beastly terms. An amplifier's behavior won't change into unfriendly loads. If you're paying tons of money for a company's crowning achievement and statement amplifier however (with transistor amps, this invariably means the most powerful), shouldn't it sound terrific no matter what - at low volumes and even over 100dB speakers? It really should. It hardly ever does though.
99 out of a 100 times, the very things which create high power -- endlessly paralleled output devices, multiple gain stages in series, high feedback to combat nonlinearities -- also create compromises for applications where this power is excessive and redundant. A pretty logical conclusion a sane mensch might draw from that is that
Before you insert into the 'low-power' placeholder valve amps by definition -- or SETs to be more specific -- hold a moment. Low output impedance and current delivery are audible across the board. But they're not something SETs deliver to the same extent as most transistors. Resolution, among other things, is a function of noise floor. It's another area where transistors, in general, hold the stronger hand. Transistor amps like Nelson Pass' FirstWatt F3 and F4, Vinnie Rossi's Red Wine Audio Signature 30.2 and Bryston's 28B SSTs show how sand amps can offer something as important to audio pleasure as direct-heated triodes though not exactly the same. That lit-from-within thing triodes do -- 'artificial intelligence' as some refer to it -- is still beyond the grasp of transistors. But so is their sense of seamless bottom-to-top linearity, transient definition and dynamic tracking out of reach of most triodes. Those tend to lack the raw current which translates into these very qualities; plus stygian bass foundation and upper-bass attacks to convey soundstage scale and music power respectively. By the time you increase tube power with push/pull and paralleled circuits, the speed and immediacy of simpler circuits tend to get sacrificed.