"Having designed pure transistor and pure tube machines and now a hybrid, what to your engineering mind and listener's ears are things which your specific triodes and transistors add respectively to this particular circuit?" It's a very obvious question. I was most curious about the engineering half of Mario's self assessment. We expect tubes and transistors to measure dissimilar. Do those behaviours neatly dovetail with listening? Is popular perception about THD correct and in fact paramount? What about other factors which non-engineers never consider? Whilst awaiting Mario's reply, I gussied up my small rig with Kinki Studio's 2.5MHz DC-coupled class AB EX-M7. It's the stereo version of the downstairs monos with the same lateral Exicon Mosfets. The difference over the crisp dry Danish class D rejects was instant and crass. Though small triodes might indicate a more moody than crystalline treble of bronze never platinum hues, the clattering and ticking cymbals on an ECM album of the Tord Gustavsen Quartet had fabulous shimmer, brilliance then gushing airy spray. Clipped-off decays had been banished to the boiler room.

That's key. A quick direct-coupled amp without phase shift can hold up a loupe on how instantaneous leading edges rise; how long falling edges trail. There's no time confusion, lack of slew or any polite rounding off. Here the Ultra Plus behaved very 'transistorized' so capable of cutting through any thicket of pedal tones, stacked legato washes or excess recorded reverb when percussive elements demanded it. Exploiting a small AMT's innate 4:1 velocity lead over classic dome tweeters added another magnifying glass on the dynamic part of said discipline. When cymbals crack hard, their dynamic charge should flicker like bright lightning against a dark sky, not tickle timidly in the undergrowth. Here the Venetian's colour saturation exists in perfect parallel to a fencer's rapidly flicking blade. To this extent that's quite atypical. If you silently questioned whether these tubes can fire sharply when called upon by a recording so crack a whip – if the amp downstream for them allows it and your multi-way speakers aren't badly confused in the time domain, yes they can. That's the adrenaline half of the Ultra Plus proposition. Click album covers to hear two tracks. The black one is thick, dark and nearly oppressive, with plenty of octave-doubled chicane between cello and clarinet or vocals to require keen timbre separation. That's for the Rembrandt half of the Ultra Plus so a rich colour palette with deep half shadows. The green track is upbeat and peppery to rely on snappy timing and julienned transients. Two very different tasks, both aced by the hybrid ZeroUno.

Being able to listen in 'hifi blackout' mode is key especially at night when idiot lights of varying colors and bright displays can become distracting.

Mario: "I started with a very elementary, raw and crude idea. A single-ended triode amp sounds different from a transistor amp. Even a child knows. The first is warm, elegant and fascinating but lacks substance. It's beautiful but without soul. The second is powerful, authoritarian but never magical or elegant. And I wondered. Why not create a bridge between them? I was looking for a 'sex bomb', a circuit that would be beautiful, sexy and very intelligent. A parallel is the car with early turbo boost. It was a revolution but suffered because efficiency was limited to a minimum of RPM. Many years later the twin turbo arrived to apply boosted acceleration to any RPM. I think that paralleling a triode and JFet works similar. I'm very familiar with winding my own transformers so prepped one to bridge 'holy water' and 'devil', i.e. my 211 amp with Tango iron and an FM Acoustics 801 solid-state amp. The transformers for that bridge derived from what I had already designed for my LaScala and Olimpico amplifiers. That's how I summed two sounds and the result was exciting. When I listened with others who didn't know what was playing, comments were 'whoa'. So I decided to develop a DAC/preamp and power amp along those lines. The latter is still in R&D but might be ready for HighEnd Munich next year. The preamp became the Ultra you have."

"From a technical point there's little else to say. It just works. My triode/JFet summing occurs at a 40:60 ratio. As engineers we apply models of transistors and triodes which unfortunately don't explain everything. It's certainly possible to predict whether a circuit will work but no single model can as yet explain why these architectures sound different. PSPICE can't explain it. We might want to point at harmonic distortion but even when we reduce THD to almost zero, the differences remain. We might invoke varying behaviour at saturation close to clipping but the differences are present already when the devices still work in their linear region. THD in general is useful at less than 70dB but -90dB THD won't better -70dB. No single conceptual model can explain the different fingerprints which the various parts leave. Gain ICs do exhibit astonishing features but never sound like a single-ended triode. In short, in the books there's no good technical reason to parallel a triode with a JFet. In fact there are only contra indications and complications. A triode followed by a JFet or vice versa is very typical. It's the common definition of a hybrid circuit. Paralleling them instead is very uncommon so an atypical hybrid. How the industry usually works is to apply classic proven circuits that simply work; to spend as little as possible; R&D as fast as possible; then end up with an easy profitable build. As a boutique builder, I can't compete with consumer products on visibility, scale, efficiency or price. I'm also far more limited in my production capacity. So the only way to stand out and make a mark is by creating something that astonishes and stops people in their tracks because it's demonstrably superior." Again, what he said!

This also piqued my curiosity. Far more than hoping to learn how Mario fixes his 40:60 ratio¹—with a transistor value?—I was curious why he set it there. What particular valve flavour did he mean to throttle back a tad to in turn open up particular transistor qualities by just a bit? That falls under tuning/voicing. It's subject to a designer's inner compass. It's where strict engineering for circuit efficiency breaches the artistic realm. Here the only guiding principle is one's personal idea on what's beautiful, compelling or simply feels 'right'. If all available options measure fine to show no operational issues, how else to decide? On what costs the least? If cost is the same because it's about a simple resistor value, that argument dies. Ditto when within reason, cost isn't the prime mover but the sound should most closely conform to the designer's true north. Having had such discussions with other makers over the years, most were surprisingly reluctant to indulge any subjective sonic comments or get specific about 'their' sound. They insisted that as a reviewer all of that is my job then left it there. Having now done my reviewer's job, I can leave it here. In the sandbox where I still play, the CanEver ZeroUno Ultra+ was that rare next-level component which, should we sample lots of kit for ad nauseam lessons in more-or-less sameness, gets ever harder to meet. If I kept an audio diary of hardware dates, I'd have to draw a big fat red heart next to the day this fine machine from Venice knocked on my door. Now that it departs again, another date in the same diary warrants a big black heart. Sniff.

Postscript. CanEver? Can it Ever! Just to be redundant because it's key: this is a monster analogue preamplifier with a very good DAC built in. It's not a DAC with a convenience volume control plus some analogue inputs tacked on; or even a DAC with a very butch but classic output stage; or a basic tube buffer. This is no verbal fencing to be cute. It's a fundamentally different approach. It's at the heart of what makes this 'DAC' different. ZeroUno happens to be the prefix but that's misleading. It's only the case because of the existing portfolio conventions. After all, Ultra Plus ZeroUno doesn't have quite the same ring. It would simply be more correct and should be considered that way.
___________________

¹ The only other design I've read of that seems to do a similar thing is Vinnie Rossi's Brama preamplifier and integrated. They allow users to combine triode and solid-state paths in adjustable 25% increments from pure triode to pure transistor with various mixed ratios in-between all selectable by BT remote.