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

It's important to use the ears. Eyes are for looking. The mind is for thinking. Op-amps. Discrete op-amps. Discrete discrete. Each approach has its champions and detractors. Ditto DAC chips and USB protocol. Associated propaganda tends to the absolutist. It means to prejudge us with our minds. Bling does the same for the eyes. All of it is designed to have us buy in well prior to any audition so we hear what we think we should hear. The real buy of course is about money. How much does it take to satisfy the ears if not eyes? Nobody likes to waste funds. But spending just a bit too little to miss that hot buy is a waste as well. This uncertainty gets exploited such that a Bifrost—small, plain, cheap as these things go—reads too good to be true. The obvious antidote is to trust your own ears. But it does help to audition multiple contenders. Here consumers are often limited by what's locally accessible.

KingRex UD384/Upower. At $479/$189 for the DAC/PSU respectively, this was my closest call. Because the Upower battery supply is optional, I ran the KingRex off its wall wart to keep the Bifrost's price tight company. Here prejudgment clearly favored the Bifrost. It's 10 times bigger. It's got a ' real' power supply. It's got multiple inputs. It seems altogether more serious (though as its name indicates, via USB the UD384 will process 384kHz data and won't skip over 176.4kHz). With Burson's $950 DA-160 as my next contender, twice the money buys a studlier box with more and butcher socketry and higher internal parts density. Shouldn't that factor?

In AudioMidi the Bifrost shows up as Schiit USB Interface plus (strangely) four more devices.
In PureMusic 1.85 the allowable input rates sadly exclude 176.4kHz.

Earsay judged these three direct competitors. In Nelson Pass' scheme of 2nd/3rd-order types—this refers to a device's predominance of either second or third-order THD remnants—the Schiit and Burson were 2nd-order 'triode' devices. The KingRex was a 3rd-order 'pentode' specimen. This meant a sharper snappier more energetic forward presentation for the UD384 and a softer more relaxed flowing feel for the other two. Between DA-160 and Bifrost the Aussie had robuster bass and thicker overall weightiness, the American greater airiness. The first differentiation captured core flavors. Whittled down to basics it was about transient definition, separation and subjective detail focus for the KingRex; more watercolour transitions and a sense of organic flow for the other two. The second differentiation grafted mini echoes of the KingRex atop the Schiit. This sequenced into Burson | Schiit | KingRex whereby, left to right, one proceeded from warmth, mass, relaxation and density to crispness, propulsion, focus and litheness. The distance between Burson and Schiit was smaller than between Schiit and KingRex. The overall delta of difference between the two ends of this lineup was smaller than what separated the lot from my Eximus DP1 reference. That $3.000 machine steps up the game to overall more resolution, impact, energy and presence.

In AudioMidi the UD384 shows up as KingRex USBDAC UD-384.
In PureMusic 1.85 the allowable input rates cover everything from 44.1kHz to 384kHz

USB or S/PDIF. One intended application for the KingRex is as D/D converter. Go in USB, come out coax. This eliminates >192kHz data where S/PDIF maxes out. Would the Bifrost sound any different entered coax, not USB? And if coax were better, wasn't really the KingRex USB implementation superior to Schiit's rather than calling S/PDIF boss? Yes I could have tapped my €30.000 Esoteric/APL Hifi UX1/NWO-M player as legacy transport. And no it wouldn't have been an even comparison as it doesn't output USB (never mind that its price doesn't fit this picture).

In AudioMidi the DA-160 shows up as TE7022 Audio w/SPDIF.
In PureMusic 1.85 the allowable input rates only cover 44.1kz, 48kHz and 96kHz.

Note how in AudioMidi the front right channel is muted. This happens automatically upon connection and must be reset.

As it turned out, speculations were unnecessary. USB direct was superior to performing external USB to S/PDIF conversion, then sending the results down two connectors and a digital cable of Stereovox caliber. The shorter signal path had the advantage on elegance/suavity, depth and ambient recovery. I didn't expect to hear any differences but these were quite obvious. As a computer audio user I'd thus not worry about the—very minor—trash talk Schiit has aimed at USB in general. Their implementation is clearly solid. Whilst I would have pitted the Bifrost against what's built into my Bel Canto C5i integrated, the latter's analog inputs employ A/D conversion to meet their digital volume control. That's a no-no for this review context.

Judging the Bifrost purely on its own merits—inserting it into the usual system and leaving it there to play—I quickly forgot about the Eximus DP1. That's because like Zu's Essence, Schiit's DAC covered the timing essentials of continuous flow to avoid all edgy choppiness or pixilation. As with the Essence, one can acquire more raw data mining and LF fortitude than the Bifrost delivers. Yet one never feels shortchanged on organically nutritious calories. This particular take on what's essential differs from the detail-über-alles obsession that's currently en vogue.

It approaches things from the other end. From the very beginning it gives us more advanced goods like temporal coherence which has music play straight in the pocket. Ultimate bass power, dynamic swing potential/drive, heightened presence and greater illumination come second. Those are reserved for costlier models which—like the Schiit Lyr headphone amp does over their Asgard—build on the essentials already established properly with the base versions. That the Bifrost can cover those bases for $450 inclusive of asynchronous USB 2.0 is unexpected (192kHz files play without fuss). This catches us up with Schiit's engineers. Despite dude talk, they know how to tap true essentials even when allowable parts costs are slim.

A beefier PSU, butcher transformer, multi-paralleled chips and sundry drive up costs. Though it should massage ultimate tone density, bass power, resolution and air, systems priced in sync won't be capable to process more than the Bifrost passes on. The greatest compliment is that in my usual reference system—grotesquely mismatched on price but review etiquette only changes one item at a time—I wasn't at all unhappy to listen to it particularly at standard levels (only at whisper levels did I find myself really wishing for the Eximus' greater overall illumination).

ModWright LS-100 with Psvane CV-181T, FirstWatt SIT2, Aries Cerat Gladius with new Fostex FW305 woofers

This meant that the Bifrost shifted none of my system's carefully calibrated core qualities. If there was a bit less of this and that, it quickly receded from attention as not vital or relevant. That's very similar to whether one does 140 or 160km/h on the still unlimited stretches of the German autobahn. Consider Schiit's sticker. That's really quite the trick. It means that what they're not charging you for is the very necessary experience—isn't time money?—that was required to pull it off. Years worth in fact. This of course gels with my prior findings about their headfi amps. Given that the Bifrost had to tough it out in my big system rather than via headphones, I was simply more surprised than I perhaps should have been. After all, class A bias, no feedback, discrete FET-based outputs - these are suggestive items with in the right hands predictive outcomes. The Bifrost delivers. It thus plays outside what its spartan appearance or low price might predict. So it's simply another high-value proposition? You bet. From Schiit. It happens...
Schiit website