Before going all in with my Ansuz set, a cooler head was needed. I wanted minimized variables. I wanted to isolate the Pascal power amp. I'd long since wrapped the Børresen 02 review. Those boxes could stay. But I wanted to bypass Aavik's DAC. For that I connected our 1'536kHz discrete R2R Denafrips Terminator. Fed I²S over HDMI from our usual Soundaware D300Ref USB bridge with ultra-cap power supply, that plugged into a U-380 analog input. Next came our usual 200-watt LinnenberG Liszt, a DC-coupled pair of 1MHz class AB monos with lateral Exicon Mosfets. Those hooked into the U-380's pre-outs. Reseating speaker cables and matching volume, I could easily compare both amps running off our domestic source and Vibex power delivery. On the Liszt, the U-380 just handled volume.
Versus LinnenberG Liszt, with iMac ⇒ Audirvana 3 ⇒ Soundaware ⇒ Denafrips front end
For context, I'll preface the following with these data points. Our domestic arsenal includes three different class D amps. There's a pair of Auralic Merak as modified UcD400 modules with linear power supplies and Lundahl input transformers à la Jeff Rowland. There are Ncore 500 monos with Revision D input buffers from the UK's direct-selling Nord Acoustics. There's a red Purifi 1ET400A with Hypex SMPS. Unless drafted for review A/B, none of those see any active duty. We prefer our linear amps, be it Kinki Studio's integrated on a Raal Requisite SR1a headfi rig; Bakoon's AMP-13R upstairs; the LinnenberG in the big system; Goldmund's Job 225 for the TV set; or SAG's AIO in Ivette's studio. The only class D amp I would prefer downstairs to the LinnenberG, by a small margin though not for its bling cosmetics, is Merrill's Element 114. That's a zero-feedback stereo amp with GaN power Mosfets which I reviewed a few months prior. Whilst writing this, a commitment to do AGD's Audion and Vivace monos also with proprietary GaN circuits had just been inked. I was most curious whether those would add themselves to our extremely short list. Because at first, the U-380 certainly did not.
With Merrill Element 114
To circle this particular wagon fully, here's one more data point. Also just arrived was a fully played Manley Labs Absolute dealer sample. That's a transformer-coupled EL90 headfi amp which, unusually, is switchable between push/pull and parallel single-ended on the fly. Coming off our Bakoon AMP-13 driving HifiMan Susvara (or the Kinki on the Serbian floating ribbons), the Absolute/Susvara combo reminded me a bit of the U-380. With a low published SN/R of 83dB for just ~14-bit resolution—that part certainly isn't true for Aavik—the Manley behaved decidedly thicker, slower, warmer and less articulate. It was energetically duller because microdynamic range had seriously narrower. Everything sounded more steadily loud, not variably loud across far finer hues. HF sparkle was padded down, separation distinctly lower. It made for a very different gestalt from what I was used to. That included our direct-coupled Vinnie Rossi L2 Signature with Western Electric 300B; and our Nagra Classic with 12AX7. To appreciate the Manley on its own terms, I first had to get off my higher horse, chase that into the wilderness, forget about it then start anew.
Contrasting our German monos with the U-380, the Dane's microdynamic nuances were plainly coarser, its transients more soft-nosed lead pencil than micro ball point. So it felt slower. I think of the conveyance of musical energy as a crossed function of treble brilliance, leading-edge rise times, power zone impact in the upper bass and overall dynamic contrast. All that was weaker. Jump factor was duller, illumination into the outer soundstage quadrants dimmer. In short, the sound behaved more damped and energetically restrained. However, just as I'd found with the latest crop of Ncore and Purifi amps, it was very easy to listen to. I'd call it a warmish chunky comfort sound reminiscent of thicker class A. Given that our Bakoon with true variable gain is €6'000 and our LinnenberG €8'500/pr, I didn't think that the far pricier Aavik had shown its best yet. It also didn't match what I'd heard in Aalborg though that had involved Aavik separates. The U-380 hadn't been ready then.
With t|AQWO transport and d|AQWO converter, each with its own dedicated Kalista Elektra power supply. Source material CD.
In my review of Métronome's ambitious new AQWO digital separates with native DSD, selectable resampling of CD to DSD256* and switchable tube buffer, I'd already ascertained that Aavik's DAC was on par though different. As you'd expect, it wasn't as harmonically rich or materially dense. But as the doctored overlay means to suggest, of an Eugen Jochum Bruckner 9th cover for DG onto our carpet, symphonic fare over the U-380 went native. It means that any voicing which thickens tone with more bloom and decay, enlarges macrodynamic swing whilst softening transitions, separation, localization focus, microdynamics and ultimate localization specificity approximates the symphonic concert experience. It also shrinks dramatically in scale since nobody's living room is a concert hall. But…if one loves large-scale classical symphonic or operatic, such a voicing approximates the gestalt of an orchestral experience. The photo of the LA Philharmonic's Walt Disney hall visualizes that reality of space, distances and large instrumental forces heard in the far field. The 'feel' of the U-380 conformed to that ambient-rich perspective, not a drier more resolute near-field perspective of predominantly direct sound. It's a very different aesthetic from that of the close-mic'd studio production but is equally valid. It simply didn't match my recall of Aalborg. It was time for the full Danish hit.
* By lacking an I²S input, the U-380 couldn't process pure DSD from the t|AQWO like our Denafrips Terminator could. For DSD, I would have needed virtual media over USB instead.
