Spec candy. It's imperative to revisit that -155dB noise floor value. According to Cees, in direct connection, it presents itself to any amp at 2V out. That's already well in excess of what it takes the vast majority of amps to achieve their full rated output. Some of the quietest current amps like Mola-Mola's Kaluga and Benchmark's AHB2 achieve ~130dB whilst many SET (if spec'd honestly) might hit ~80dB. Metrum's forthcoming Forte stereo and mono amps should be hard-pressed to breach that 130dB threshold. If team Metrum use passive voltage gain from step-up transformers, a seriously tweaked power supply and cool operational temps, they might just manage? Regardless, on pure magnification power as the ability to resolve very minuscule signal, the Adagio would seem to be well ahead of what 2017's amplification can match. Relative to hifi's relentless pricing ascendant, getting there at well below 5 figures counts. If Dutch origins imply potential frugality, we shouldn't expect equivalent eye candy. Actually, that became a consensus dealer demand to face Metrum past launching the awarded Hex. High-street vendors complained that it looked too plain. That's why the Pavane grew the posher fascia and cheeks plus glass lid. All three migrated to the Adagio. Just so, it's where Dutch value thinking stops to lay on the thick sauce. The true bling brigade pursuing Boulder/Constellation type metal work will still go elsewhere. I can't help but think with a grin that this would be perfectly fine with Cees. His focus is real engineering, not marketing voodoo or cosmetic excess. [At right, a DAC TWO module containing two R2R ladders, FPGA and I/V conversion stage, all more petite than a credit card.]

It's how his entry-level Musette DAC could sell for €1'033 and should still miss the €2'000 mark once it's been upgraded to DAC TWO specs. Other companies which push digital boundaries—Chord, dCS, Meitner & Co—tend to work in the rather pricier end. By pushing boundaries I mean proprietary tech which advances the art, not reliance on turnkey solutions, be they an ESS Labs or Asahi Kasei chip or a Soekris OEM R2R board. Because digital can't be listened to but only measured, advancing the art, by definition, must mean superior specs. That implies lower noise, lower distortion, wider bandwidth, higher slew rates... the works.


Metrum's forward-thinking approach led to their forward error correction algorithm and separate data processing of the less/least significant bits. This works them at higher amplitude which post conversion gets reduced again to match. Such temporary magnification is said to improve the linearity of the digital-to-analog conversion process just above the noise floor. The result is improved audibility of the very tiny stuff which otherwise falls through the cracks.


Owning that tech only to see it invariably diluted by downstream losses in the analog domain—here Cees' prior graph demonstrated how the current limits of thermal noise in resistors prevents analog from eclipsing 24-bit resolution—explains why he was so keen to eliminate one analog component (the preamp), then improve its core functionality of volume control. Why work so hard on improving digital specs only to see them subsequently squandered?


That then is the Adagio's entire raison d'être: to minimize signal losses before they hit our amplifier and loudspeakers which, as genres and particularly so the latter, involve exponentially greater distortion and nonlinearities. That Cees Ruijtenberg has some new ideas on how to improve the former we already know. His Forte stereo and mono amps in the existing Pavane/Adagio casing are imminent. About their general concept, Cees had the following details:


"They're an idea from two years ago. At the time I had hybrid amps for pleasure listening and critical evaluations. What I did not like was the fact that triodes suffer aging problems. In simulation software, I started to imitate triodes with fets. The good thing with triodes is their low gain at a lot of bandwidth without needing feedback. That was my goal. It took me a few weeks of simulations but I hit upon the desired result. It falls somewhere between a 12AU7 and 12AT7 on gain but with a bit more bandwidth. The disadvantage is that it runs on the same high 350V rails as real tubes would. From there we drive a class A/B output section (10 watts in class A, 50w/8Ω, 100w/4Ω, 170w/2Ω). Damping factor is 200. Compared to my old hybrids, there's no sonic difference, just less noise. Two years later, we've now turned this into a commercial product with a lot of added protection because during the startup process, we must charge up 176'000µF. And that, in a nutshell, is the Forte platform."


Back to the Adagio, I kicked off with a potential white-out for a first pulse. This ran it into nCore 500-based Nord Acoustics monos with Sonic Imagery discrete opamp input buffers, thence into Albedo Audio Aptica speakers. DAC-direct into ~130dB S/NR class D into ceramic Accuton drivers could be a quick trip to the Cold White North of frost bite and snow glare. Whilst the sound was leaner than the Pass Labs XA-30.8 class A amp, it was far from lean per se. In fact, letting this settle in over a weekend, I was shocked by just how rich and juicy it sounded at already low volumes. Once I'd grasped that lovely fruitiness, I turned the volume down a bit more to see whether it'd collapse. Not. And again. Not. Of course eventually it did, not because the hardware was faulty but because my own psychological threshold was breached and the Fletcher-Munson curves were hard at work. This was my first solid giveaway that Cees' spec candy wasn't empty if sweet junk food. More micro detail equaled higher fullness into lower volumes. Only SPL shrunk. Nothing else did. If you think that's normal, not. Granted, it's how things should be—and would be in a perfect world—but rarely are. Particularly with digital volume which gives its converters less and less raw signal to work with and soon throws away bits, gradual white-out kicks in progressively. It gets demonstrably paler and flatter as attenuation compounds and the less significant bits stop being audible. To not have that happen moved the direct-connected Adagio into rare company.