This review page is supported in part by the sponsors whose ad banners are displayed below

Song-kyung Han again
: "Excellent review on the Stello U3. Frankly, when it was released in Korea about a month ago there was a lot—and I mean a lot— of hype about how great it is. Naturally I dismissed most of it as noisy marketing. Your review now gives it credibility. I really did not expect that a $500 DDC could be a 'magic bullet' that makes many quality USB/Firewire DACs with S/PDIF input better by a noticeable and significant margin. The first batch of DP1 was distributed in Korea less than a week ago. Unfortunately I wasn't so lucky. There is even much greater hype about it now. I suppose more burn-in time will be required but everyone—and I mean everyone—who received theirs is saying it's way better than expected and easily outperforms DACs or preamps costing up to $10.000. They say it is best used as DAC+PRE."

Left to right, top to bottom: Eximus DP1, Resonessence Invicta, Burson HA-160D, Weiss DAC2, Stello U3,
Metrum Acoustics NOS Mini DAC Octave, Centrance DACmini, Antelope Audio Zodiac Gold/Voltikus

In use the DP1's tiny signal lock LED turns color. This reflects the incoming sample frequency - yellow for 44.1/48kHz, red for 88.2/96kHz and green for 176.4/192kHz. To distinguish between the paired frequencies within this scheme would have added a light and likely confusion. The same display convention applies to the defeatable upsampler - without yellow of course. The pointer shape of the volume control gives visual confirmation of setting if you remember which end is which. Class A bias means the box gets warm in operation, low-level signal voltages mean warm never gets hot.

The filter switch is specific to the 6.3mm headphone port. It does not relate to a defeatable digital filter as might be assumed. Rather it operates in the analog domain. Bypass is the default. Engaged and confirmed with a red LED, bass response is up 6dB @ 30Hz "good for certain headphones like AKG. This is not a x-over type filter but a simple boost at 30Hz. Most headfi systems don't resolve this during playback but you can instantly feel it in the change of ambience. The knee of this filter is at ca. 75Hz."

Top of the heap. DP1 | Invicta. Updated on features and sonics since my pre-production unit review, the Invicta holds one ace card the Eximus can't match. Remote control. Today any upscale converter proposed for amp-direct use will be—perfectly justified—expected to offer infrared or RF volume change. And admittedly the Canadian machine is $1.000 more. With competitive/equivalent sonics, many shoppers could—again perfectly reasonable—prefer not to pay a grand for their new couch-potato membership.

FirstWatt S2 proto, Voxativ Ampeggio with upgraded drivers

Speaker sonics. These converters were so close as to make any decisive honing in on differences a matter of the first few bars. Afterwards my ears adapted and I had to remind myself of this difference rather than perceive it as directly and clearly. The Eximus was slightly sweeter, warmer and 'analogue'. The Invicta was slightly more metallic, fresh and 'digital' (major emphasis on 'slight' in either case). This distinction was clear by how the first tones arose from silence. They did so in more sharply edged outlines and contrast with the Canadian, in rounder more transitioned back/foreground integration with the Korean. Little amusical noises nearly highlighted by the Invicta proved to be just as present with the Eximus albeit not as obvious.

In Nelson Pass speak—and this is purely a descriptive ruse, not any actual cause/effect diagnostic—the DP1 seemed to exhibit minute remnants of 2nd-order THD, the Resonessence 3rd-order. In the FirstWatt amplifier catalogue these would be echoes of J2 vs. F5. In modern digital speak, the final Invicta would rightly be called an extreme resolution machine though it accomplishes that with a newfound smoothness as apparently the result of minor refinements. The Eximus had a very fine softness/sweetness about it which made the contrast ratio a tad less intense. In concert-hall speak this would be the difference of just one or two rows not on apparent stage distance but as the effect it has on the mix of direct vs. reflected sound and thus the center point between transient and bloom.
Headfi sonics. That this small distinction was intrinsic to the converters rather than line-level output stages got confirmed when headphones duplicated it. Both machines had sufficient drive/gain to whip into shape even more hard boiled loads like the HifiMan. Stout but still civilized levels with the HE500 and K702 had the Eximus at 10:15 on the dial, -15 to -10dB with the Invicta. The Korean was thus endowed with more ultimate headroom and horsepower to likely be sufficient even for diehard headbangers where on truly inefficient loads the Canadian might just run out of raw crunch. If you extrapolate from the above and remember the narrowness of the difference window; and if you know the voicing of these cans... you'd correctly predict that the Sennheiser HD800 would be an even more perfect match for the Eximus, the LCD-2 with the Invicta (if we strategically cross-match that is, not emphasize like with like).

CEntrance | Eximus*: The $795 DACmini from the authors of the AdaptiWave™ firmware and JitterGuard™ circuitry licensed to Bel Canto, Benchmark, Empirical and Lavry is a sober reminder of diminishing returns. It's a very high-value proposition which additionally demolishes the async supremacy racism so prevalent in current USB audio discussions. What the thrice-high investment of the DP1 buys is more of three things - ambient recovery, lateral expansion and overall scale. By recovering less venue action, the CEntrance sounded starker. With it Dulce Pontes on "Verde pino verde mastro" from her Momentos twofer for example had a more disassociated quality. The DP1 showed clearly more reverberant interaction around the voice. This produced a richer more interconnected sound embedded in vibrant space that wasn't a cutout voice overlaid against a black vacuum as the DACmini did. The latter was squeaky clean but by comparison stripped away something ephemeral but interesting.

Or as a friend put it, "the CEntrance is very matter of fact." Indeed. And it really should be about the facts and nothing but. With upscale hifi the facts simply often turn out to have multiple layers or 'hidden meanings' which give them greater context and completeness. Here I'm not talking about the silly but mythical artists' intention as though anyone could really know what that was. I'm talking about what occurs acoustically around and between the performers in the recorded space. More of that enriches the timbres. It also makes the presentation bigger and denser. Take these sentences. Mary walks her dog. Mary walks her nervous dog. Mary walks her nervous dog under an ominous sky. The first sentence has all the facts relevant for a police statement. The last paints a more complete moody picture for a psychologist looking for meaning. Music and mood are inextricably intertwined. The pure music facts are just tones in proper amplitude and time. The mood facts need more subliminal tertiary data. That in essence was the difference between Eximus and CEntrance.

* This and subsequent comparisons involving DACs without variable line-outs used my remote-controlled Bent Audio Tap-X autoformer passive preamp also on the Eximus to eliminate extra variables.