Density and speed? My first dégustation stop were HifiMan's Susvara planars. At stout volumes, those had gain to spare. Half the attenuator's range sat unused for big headroom. Now to sound. In many ways it became counterpoint to COS Engineering's H1 which usually hogs the spot. Riviera's take was obviously thicker, denser and chewier. That part followed generic tubular expectations. What didn't was simultaneous speed. Certain transients peeled out from the surrounding thicket like distress flares above low-hanging ocean mist. To these ears, standard 2nd-order THD typical with triodes can act as indiscriminate patina. It applies itself to everything like a textural equalizer. Depending on dose, it can span from attractively shapely to - well, gone to seed. Here the AIC 10 differed. Tone density and earthy grounding coexisted with fiery attacks which cut through brilliant, sharp and steep. Those plus very good S/NR represented more typical transistor virtues. Getting this constellation of qualities which often occupy either/or polarities; and each well developed... that was unexpected. Right off I felt in good hands. Someone clearly knew how to massage his mixed ingredients to best effect.

Another attribute which the Riviera expanded with Fang Bian's flagship planars was dynamic shove. In their review, I'd described how the Susvara were exceptionally spacious; and how on LF bandwidth, they were fully equal to Final's top sealed Sonorous X when Bakoon's current-mode AMP-12R did the honours. However, the Japanese heavyweight dynamics still exerted more pressure intensity. Their sonic physicality was greater. Given how our earspeaker subject can't involve the full body, I won't call it slam. But imagine a localized version and the Sonorous clearly had it over the Susvara. Now the big but surprisingly cool-running Class A integrated shrunk that gap. Now the thin-film membranes showed more dynamic expressiveness and body. In short, with a power supply and output stage designed for 30wpc @ 1%THD of speaker drive, the AIC 10 in headfi mode behaved like a hyper muscle amp; just with a small amp's noise floor and speed. Cynics would call it massive overkill. But even they couldn't argue with these results. The Riviera/HifiMan combination really was superlative!

A quick stop in the media room was next. Here Cube Audio's Bliss C held court. Their extraordinary agility and open-throatedness proved even better served by our petite April Music Stello separates. In this smaller room, the AIC 10's considerable harmonic weight and density slightly congealed the Koreans' superior speech intelligibility on talk-heavy brain fare like Tea Leoni's Madam Secretary. Given that the Polish widebanders too were new review loaners to add more variables, the Riviera quickly moved...

.. into the big room's status quo behind our Audio Physic Codex. Given this nearly 100m² space, lift-off at just 9:00 for solid room levels felt unexpectedly swift. I actually revisited their website to confirm 89dB sensitivity. Sticking it to kilowatt beliefs like an operatic knifing during a heroic aria—on a high C perhaps—Riviera's official 10 watts could easily have clipped our ears long before hitting 18:00 on the dial. Again, let's call the power rating conservative and move beyond associated flea-power concerns. Appreciating how little actual power most of us use is key to not over shop. Having replaced the 30wpc Pass Labs XA30.8, it was easy to tell that the Riviera's low bass was a bit bloomier; and that its virtual needle on the elasticity-to-damping metre sat slightly deeper in the elasticity zone. This was similarly rich generous sound but the Pass held the tighter reins of control. With Wyred4Sound's top 10th Anniversary DAC and the matching STP SE II linestage as two resolution monsters, the Pass emphasized the transmission of musical energy and in-room projection. Off the same DAC without linestage since the AIC 10 integrates its own, the Riviera shifted grippiness and forward adrenaline into a heavier low end, more relaxed gestalt and even more burnished tone. Any associations one might entertain, between transistors and dryness, were obliterated. Mostly one could paraphrase this as very accomplished valve sound; from one lone 12AU7. In fact, to make proper sense and track the actual performance, we would need to change that to a 6SN7 type. That's because elsewhere, 12/7 bottles don't have that effect. Actually, we'd have to nearly invent a 6SN7 power not small-signal triode to explain it. That Luca Chiomenti pulled it off with transistors to avoid output transformers is no everyday trick. [For a real pop-quiz fact, Philips started in Holland's Eindhoven in 1891 and concentrated on the mass production of incandescent lamps. After a joint venture in vacuum tubes, they acquired Mullard in 1928. So the connection between lamps and valves really is most direct.]

Back to valve aromas. Riviera's wasn't that peculiar from-within-glowing spaciousness which the best micro SETs using 45 and 50 can set in high relief. Instead, this was earthy ballsy push/pull flavour. That even tracked with the actual output stage to feel more Stravinsky than Debussy. In speaker terms, it was big boisterous Trenner & Friedl not ceramic driver sound. Unlike many pure valve amps with their hundreds of metres of phase-shifting transformer windings and limited HF bandwidth, treble reach and brilliance were solid-state linear. It's in the slightly underdamped bass with its concomitant minor fat where pure tube notions with low feedback factored. Unlike pentode pushiness—certain UL tube circuits which play it forward in the presence region—this sound was triode sweet. This triangulation gives us the general fix on the sonic map.