John Grandberg
Comparisons. In my humble opinion, they are critical in helping others understand the sound of a device under review. Comparisons give us context which other descriptions – as eloquent as they may be – lack the ability to convey in isolation. I certainly still want those descriptions, but the comparison is often the key metric by which I decode the opinion of a given reviewer.
Let’s walk through an example. When reading a review of a headphone I have yet to experience, it’s great if the author goes into detail about the sound signature. They tell me all about what they hear, using musical examples when appropriate, or whatever it takes to paint a vivid picture of the sonic landscape. Let’s say their ultimate conclusion is that the headphone has exceptional clarity but also sounds a bit thin and tipped up, and requires very powerful amplification to properly drive. That’s a reasonable conclusion, and it gives me a general idea of what to expect from the headphone being evaluated.
But in order to take things to the next level, it would help tremendously to understand where this headphone sits in comparison to other options. The well-known Sennheiser HD800, for example, is notorious for having just that sort of signature. How does the headphone under review compare to that established product? Moving on to amplification requirements, HiFiMAN’s Susvara is infamously the most demanding option on the market. Is the review headphone similarly challenging, or more/less so? These comparisons (and many others) give us vital context to help calibrate the author’s perspective against our own.
The Project
For this article, I wanted to explore the dedicated DAC space around the US$2,000 mark. Gear in that range tends to give listeners compelling performance without being so pricey as to become out of reach. Products in this category have left the “budget” zone and thus have more breathing room for quality parts and construction, yet must still rely on clever engineering solutions if they have any hope of standing out amongst the crowd. Competition here is extremely fierce, and all that pressure tends to produce some diamonds.
In addition, I also wanted to focus on stand-alone DACs which stick to a single purpose. Readers have no doubt heard me opine about my love for integrated units which handle D/A conversion, headphone amplification, preamplification, and even network streaming. Those still make a lot of sense in many situations, and I have not wavered in my enjoyment of those multi-purpose machines. That said, assembling a system using a US$2000 DAC likely also involves some rather high-end headphones (and/or speakers), an equally nice transport, aftermarket cables, etc. The target market in this space likely already owns or has their eye on, a quality stand-alone amplifier, whether it be for headphones or speakers. It therefore might make more sense to throw all our eggs into one dedicated single-purpose basket, instead of duplicating efforts on features that may not be used very often.
I do want to clarify that this is most certainly not a “shootout” type comparison where we end up with a ranked list of absolute winners and losers. That sort of granular thinking is great for click-bait web reviews or audio-forum arguments, but here in the real world, we know that different listeners value different things. One person’s welcome “detail” is another person’s dreaded “harshness”, and we clearly aren’t all chasing the same result. Not to mention the massive array of variations in ancillary gear, musical preferences, room acoustics, preferred listening levels, and a few dozen other uncertainties which make it impossible to assign a definitive ranking. Instead, I attempt to flesh out the character of each device, on its own as well as in relation to one another, and let the reader decide which one might best fit their particular use case.
The Contenders
First up is the DAC3 B (US$1799) from Benchmark Media Systems. Benchmark has been making studio gear since 1983 and became highly popular in audiophile circles in the late 2000’s thanks to their original DAC1 device. That unit was a DAC with integrated headphone output and became something of a benchmark (ahem…) in the sub-US$1000 category. The firm built on that success with upgraded models such as the DAC1 USB and DAC1 Pre, helping popularize the compact “do-it-all” approach we so often encounter today.
Benchmark continued evolving with the DAC2 and now DAC3 models, both of which had multiple sub-variants based on feature set. To my knowledge, the DAC3 B is the first Benchmark model to offer just the DAC itself, without preamplification or headphone outputs – and with a corresponding cost savings of course. I find this significant as Benchmark now offers a dedicated power amplifier and matching pre/headphone amplifier. Assembling such an all-Benchmark system using any other DAC3 variation would involve overlapping features and thus unnecessary expenditure.
The DAC3 B can be purchased directly from Benchmark or through their rather large dealer network, spanning most of the globe and including both online and brick-and-mortar dealers. All Benchmark products are made at their facility in Syracuse, New York.
Benchmark offers a one-year standard warranty on the DAC3 B. Registering the product at their website extends that to 5 years for residents of USA/Canada and 2 years for everyone else around the world. That’s some of the most generous coverage you’ll find out there, and given the Benchmark pedigree, you can be sure the company will still be around if you ever need them. In addition, a 30-day risk-free trial is offered for those unsure if the device will fit their preferences.
Next comes the Pontus II from Chinese firm Denafrips. At 12 years old, the firm is a relative newcomer to the hi-fi scene, but has made a rather large splash with its line of DACs – all of which feature hybrid R-2R technology, custom FPGA processing, beefy enclosures with excellent build quality, and smooth, inviting sonics. Denafrips has expanded over the past few years and now has a full line of power amplifiers, preamplifiers, master clocks and reclocking devices, power conditioners, a dedicated headphone amplifier, and even a set of monitor speakers. It is now possible to assemble a full Denafrips system, starting at around US$5000 for the simplified entry-level setup and climbing to nearly US$20,000 for a more complex system using the top model from each category.
The Pontus II sells for around US$1700 depending on the exchange rate, and is distributed around the world by Vinshine Audio care of the ever-helpful Alvin Chee. This is actually the second most affordable DAC in the Denafrips lineup, with a single model below it and three above. Still, it weighs in at over 8KG and is built like a proverbial tank, giving users the feeling of substantial bang-for-buck from a physical perspective.
All Denafrips products are made in China where the company has its own facility and does everything in-house. They offer a 3-year warranty which thankfully does transfer to the second owner. And the firm recently partnered with a company in Texas to handle repairs, meaning North American customers like me need not foot the bill for overseas shipping if an issue pops up. The downside is that, unlike Schiit and Benchmark, Denafrips does not offer any sort of in-home trial period. Apparently, it just isn’t feasible with the low margins involved.
Last but not least is the interestingly-named Schiit Audio Yggdrasil. Schiit was formed back in 2010 by industry veterans Jason Stoddard and Mike Moffat and has released a startlingly large number of products in that time: multiple generations of DACs, speaker and headphone amplifiers, hardware-based tone control devices, preamplifiers, and even gaming-oriented USB DAC/amplifiers with microphone inputs. The company embraces a direct sales model which they say allows for better pricing and ultimately better products too. It also allows them to be more nimble, often trying out interesting products to see how they sell and then discontinuing whatever doesn’t find solid demand. At the time of writing, there are 26(!) products in the Schiit catalog (not counting various options or sub-models), but that number is constantly in flux.
The Yggdrasil is Schiit’s flagship DAC and sells for US$2699. Yes, I realize that is significantly more than the other two DACs represented here; and quite a bit higher than my self-imposed US$2,000 target. But hear me out as there’s a story behind this discrepancy. You see, Schiit actually sells a few variations of the Ygdrassil, each using different arrays of DAC chips/output stages, and coming in at correspondingly different prices. Their new(ish) “Less is More” version sells for US$2299 when purchased in black (silver costs $100 extra), putting it much closer to our arbitrary goal for cost. To make matters more interesting, that variant is actually considered by many to be the best sounding of the bunch. I of course requested a review unit in just that configuration and waited patiently for it to materialize. But after quite some time with all “LIM” units being spoken for by paying customers, and the other two DACs for this roundup already in-hand, I agreed to accept this “OG” version to keep my project from falling apart. So yes, it is technically the most expensive entry into this collection, but the existence of a more affordable and possibly better-sounding variant hopefully offsets that fact.
Side note – just prior to my submission of this project, Schiit updated their flagship DAC to the Ygdrassil+. The main change is a very similar-looking yet structurally redesigned enclosure that can be opened more easily (remove 4 screws instead of 39) to access the innards. Those guts are virtually identical and the sound should thus remain unchanged. Owners of a non-plus Ygdrassil can pay to swap to the new design or stick with their original chassis, and in either case, Schiit will continue offering upgrades whenever they develop a new board — which they do from time to time, helping this ~8 year old product remain as relevant as ever.
Schiit products are made at their own facilities in California and Texas. They sell direct to the USA-based customers and have a small group of distributors handling other regions around the world. Schiit offers a 15-day risk-free trial period (with restocking fee, US only) and a 5-year warranty for the original owner.
The System
All three DACs were extensively burned-in prior to listening, and all remained powered up for the duration of the evaluation period. I fed each DAC straight from a Euphony Summus music Server (US$2600), paired with a Keces P8 linear power supply (US$899), using Roon to play files from my NAS plus streaming from Tidal and Qobuz. I also sometimes augmented that with a Matrix Audio X-SPDIF II (US$439) D-to-D converter to round out my digital connectivity options. That meant I had USB, coaxial, AES, Toslink, and I2S via HDMI at my disposal, so I was able to test each DAC via the majority of its inputs. Amplification was handled by a Niimbus Audio US4+ (US$6500), Cayin HA-6A (US$2500), and a Pass Labs HPA-1 (US$3500). I listened via multiple headphones including the Meze Elite (US$4000) and Empyrean (US$3000), Sennheiser HD800S (US$1599), Kennerton Audio Thekk (US$3000), Audeze LCD-24 Limited (US$3500) and LCD-2 (US$999), Focal Elex (US$799), and custom in-ear monitors including the Empire Ears Zeus XRA (US$2700) and 64 Audio A18t (US$3000). I also broke out my electrostatic system featuring a custom-made KGSShv amplifier driving the Kaldas Research RR1 earspeakers ($500). Balanced power was provided by an Equi=Core 1800 conditioner and all cables were from Audio Art.
The Benchmark
The Benchmark DAC3 B is a great place to start as it represents the most “mainstream” of the products in this roundup. I say that because it comes from the oldest and most well-known brand of the three devices represented here, and has by far the widest distribution.
But I also mean it from a design perspective. Benchmark has always been very upfront about its desire to reconcile measured performance with subjective listening results, and chief designer John Siau has published many papers on that topic over the years. His design philosophy centers around best practice methods rather than audiophile trends or exotic implementations. That sometimes leads to less traditional choices like the use of a switch-mode power supply, which Siau says clearly outperforms the linear PSU style used in earlier Benchmark DACs. Or the upsampling/oversampling scheme which sees everything converted to an extremely high rate, run through a special reconstruction filter, and then downsampled to 211kHz which is apparently the sweet spot for this particular DAC chip. Audiophile philosophy in some circles is “the less processing, the better”, yet Benchmark goes where they feel the evidence leads in order to produce best results. While it does use the popular ESS Sabre ES9028Pro DAC chip in the usual quad-mono fashion, the company avoids using any of the integrated ESS digital filters, and they further claim to not use the DAC chip itself in a “conventional manner” – though they stop short of telling us what exactly that means. Still, it seems Benchmark is focused on optimizing its design at all costs, even when that may go against the grain of established audiophile wisdom.
The DAC3 B ends up being small and lightweight, capable of easy integration into any audio system. Build quality, connectivity, and general fit-n-finish are all plenty satisfying if not quite ground-breaking – again, the “mainstream” approach where Benchmark gives you most everything you need and not much you don’t. I can find only two exceptions to that general rule. First, the IEC power receptacle is rather crowded by the USB input, to the point where my Audio Art AC cables (with their beefy Furutech Rhodium connectors) would not work with even my lowest profile USB cables. This somewhat limits the use of aftermarket power connectors, but only in combination with the USB input as all others have adequate clearance. The second (and more interesting) exception would be the “passthrough” option, transforming one of the coaxial inputs into an output and effectively making the DAC3 B an extremely capable USB to SPDIF converter. This opens up options like easily integrating a second (hopefully very different sounding) DAC into a system for an alternate presentation.
In order to activate that mode, Benchmark advises users to remove the top panel to move a specific internal jumper. While inside, I’d also encourage most people to switch jumpers on the XLR outputs, thus reducing output voltage to more sane levels. The DAC3 B is still technically intended for studio use, where voltage levels significantly exceed that of consumer audio devices. I prefer the middle setting which gives a roughly 3.9Vrms output, not far off from the 4Vrms used by the Schiit and Denafrips devices. Note that this doesn’t impact the RCA outputs which are also hotter-than-average but not so far as to cause issues with most gear.
With those settings properly dialed in, my initial impression of the DAC3 B was that of extreme clarity, but also surprising balance. I was never a huge fan of the Benchmark DAC1 because in my system it always emphasized crisp treble at the expense of convincing tone. The DAC2 was better, yet I still heard a tendency towards glare and stridency. I didn’t necessarily mind the tonal balance but did take issue with certain characteristics of said presentation. So when I cued up some favorites like Billy Cobham’s Spectrum, Mordechai by Khruangbin, or Damien Rice’s Live from the Union Chapel, I was very impressed to hear the DAC3 B giving a well-rounded performance without any obvious treble trouble. The overall sound was fast, punchy, nicely textured, and had a great sense of realism. So despite my complaints about the previous-gen models, the latest version felt like a significant improvement.
Having established that, I would still categorize the DAC3 B as being slightly on the analytical side. Again, it is nicely balanced overall, but stops short of the middle ground in terms of calling attention to detail versus bathing the listener in rich sonics. I tend to lose interest as things diverge too far towards either extreme, but thankfully the DAC3 B is not drastically tilted. This also means the device pairs well with most other gear. I quite liked it with the relaxed, full-bodied pairing of Cayin HA-6A amp running its EL34 tubes and driving Meze’s Empyrean headphones, as its speed kept things from getting dull and mushy – a problem the Denafrips Pontus II sometimes had with that particular combo. But when I switched it up and paired Benchmark with the insanely clean Niimbus US4+ driving the Audeze LCD-24 Limited Edition, the Benchmark remained enjoyable and did not sound overly sterile. I confess I was initially worried about tonal density and impact, but the thunderous drive of the Niimbus amp made that a non-issue.
Another thing I felt the DAC3 B did extremely well was to create the sense of an open, precise soundstage. I admit headphones differ in this regard compared to speakers, but folks who listen extensively with high-end headphone rigs learn to adjust to these nuances in spatial cues after a while. In any case, the DAC3 B gave a very spread-out presentation with performers nicely localized in their respective placements. There was plenty of width and even some depth which I consider a very lofty challenge for a headphone presentation. This was most evident with the Meze Elite and Audeze LCD-24 but also very noteworthy using my custom in-ear monitors.
Still, as versatile as the Benchmark thus far sounds, there are limits to its matchmaking. Keeping the US4+ on amp duty but swapping the Audeze for a slightly brighter Kennerton Thekk had me calling foul, with the treble taking on an unpleasant tone that grated after a few tracks. I do enjoy the Kennerton Thekk in many contexts but this was simply not a synergistic pairing. Substituting the Niimbus amplifier in favor of the Cayin HA-6A running KT88 tubes in ultra-linear mode went one step further, making even the highly-neutral Audeze LCD-24 sound a bit ragged and thin. This means brighter headphones like the Sennheiser HD800 family or many of the higher-end Audio Technica models will simply be off limits unless counterbalanced by an extremely euphonic amplifier. I personally wouldn’t bother – the DAC3 B does so well, with so many different headphones/amplifier combos, that I wouldn’t waste time trying to bend the less-than-ideal pairings into shape.
In the end, my favorite combination involved mating the Benchmark with my electrostatic rig. The underappreciated Kaldas Research RR1 sells for US$500 but to my ears has a very similar character to the US$2200 Stax SR-007 mk2 – to the point where I sold the Stax and kept the Kaldas. My custom-built KGHHSV no doubt plays a key role in driving the RR1 to perform at such a high level. Here the Benchmark speed and bite combined with the lavish tone of the Kaldas RR1 in spectacular fashion, resulting in superbly calibrated sonics on everything from Pablo Moses to Porcupine Tree to The Postal Service. This was a case where neither the Schiit nor the Denafrips DACs spoke to me in quite the same way as the Benchmark.
Moving away from specific headphone/amplifier pairings, the DAC3 B was generally rather unfussy in its nature. It sounded equally composed regardless of input used and was very forgiving of low-quality transports. The latter aspect comes courtesy of Benchmark’s proprietary “UltraLock3” jitter reduction system which I’d say works as advertised. I experienced excellent results running Roon straight from a Microsoft Surface Laptop via USB, which was only faintly improved by switching to a high-end dedicated music server. I also found the sonic performance virtually identical between XLR and RCA outputs. The cumulative result is that the DAC3 B can comfortably fit into almost any system. Users can focus on choosing matching gear to fit their sound signature preferences rather than worrying about technical aspects which might inhibit usage.
The Denafrips
The Denafrips Pontus II is a very different DAC compared to the Benchmark. In both design and execution, it clearly has different priorities, and nobody would ever confuse the two devices. Where Benchmark aims for compact efficiency, Denafrips brings what I’d call old-school audiophile sensibilities which seem to measure quality by the kilogram. Buyers who value that approach will find plenty to like here.
Removing the thick, seemingly bullet-proof top plate gives us a glance inside which again reflects those old-school themes, but with a modern twist. We get a fully balanced dual mono design using an array of four discrete R-2R ladders flanked by a massive bank of over 100 low ESR capacitors. The overbuilt linear power supply with dual O-core transformers and another huge array of capacitors take up the entire lower chamber of the double-decker enclosure. Those classic design elements are mated with modern touches like full galvanic isolation of all inputs, automatic voltage switching for universal compatibility, and custom FPGA-based digital signal processing. Denafrips even designed their own proprietary USB audio receiver rather than relying on the usual suspects like XMOS or Amanero. Much of this is trickle-down technology from their flagship Terminator DAC, which is very highly regarded and sells for roughly three times the cost of Pontus II.
Before we go on, a quick word about NOS. Denafrips claims their device can be run in “non-oversampling mode”, aka NOS, accessible via a front panel button. But various folks online have taken measurements confirming the Denafrips NOS mode actually does involve some sort of manipulation. It may not be the same oversampling as a modern AKM or ESS chip, but nonetheless there is a process called “linear interpolation” taking place – which simply would not happen in a true non-oversampling design. The only exception comes when feeding the device DSD signals or PCM at 768kHz (or higher), as it does seem to give the expected behavior in those cases. Thus, people who use HQplayer or Roon for advanced upsampling actually do experience Denafrips DACs as truly NOS. Yeah, it’s complicated. While I’m not thrilled about this discrepancy, Denafrips’ use of the term “NOS” does end up being technically correct in at least certain instances. OS mode sounded better in my system anyway so I don’t consider it a deal breaker regardless.
While on the topic of preference, I’ll also add that the Pontus II was pickier than the Benchmark DAC in terms of extracting the best performance. I preferred the XLR outputs by a reasonable margin as they gave a more tonally saturated and liquid presentation. For inputs, my order of preference would be I2S, then USB and AES roughly tied, with coaxial and Toslink taking up the rear. Again it was a question of saturation along with soundstage focus. Lastly, I found that feeding DSD512 (via Roon upsampling) sounded notably more natural and organic, with PCM at 768kHz being not far behind (and presenting a much easier CPU load for those with less potent Roon server hardware). So my recommended approach would be I2S or USB, with upsampling applied, and XLR out if at all possible.
Using the above recommendations along with OS mode and the “slow” filter gave me a rich, palpable presentation, with the midrange in particular feeling meaty enough to cut with a knife. Big orchestral works such as the Reference Recordings 24-bit/176.4kHz release of Rachmaninoff: Symphonic Dances gave an impressive sense of scale – assuming my chosen headphones and amplification were up to the task. The Pontus II’s low-end drive was potent, and the overall presentation seemed focused on giving a visceral, almost primal response. This brought back fond memories of classic devices from Sonic Frontiers, Assemblage, and Enlightened Audio Designs, but without the more obvious treble roll-off that those all possessed to some degree.
Top-end air was present and actually quite articulate on the Pontus II, though not exactly the focal point. So when listening to the masterful drumming of Simon Phillips on the 2011 Hiromi Trio release Voice, my attention was drawn more towards propulsive bass drum “thwack” and floor tom sustain than cymbal shimmer. Meanwhile, the solo track “Haze” (from that same album) has Hiromi doing some incredibly complex runs on her Yamaha CFX grand piano, and the Denafrips DAC showed a beautiful sense of air and fluidity. It did seem to focus more on the scale and the resonant qualities of the grand piano, yet remained satisfyingly insightful on the upper registers. It was a relatively subtle thing, just enough to pull attention towards that full-bodied midrange without feeling too dark or muffled – a great compromise I’d say.
This was clearly a very different-sounding DAC to the Benchmark, and accordingly, it paired best with different equipment. After rotating through my collection I settled on the Cayin HA-6A as my favorite dance partner. The versatile Cayin has several modes of operation but running KT88 tubes in triode mode seemed to best complement the Denafrips DAC, making for an effortlessly tactile combination. The duo painted rich tone colors with a smooth, fleshy midrange and a regal yet buttoned-down top-end. For headphones, Meze’s flagship Elite made a superb pairing, as did the Focal Elex. This Cayin/Denafrips combo would also be a great way to tame overeager headphones from Grado, Sennheiser, Beyerdynamic, etc, which have a lot to offer but can also feel a bit exhausting to my ears.
On soundstage presentation, the Pontus II was again quite different from the DAC3 B. It was also very open and spread out but in a rather distinctive fashion. The Benchmark seemed to put me right in the front row, with performers directly in front of me to easily discern their placement. Meanwhile, the Denafrips device had me sitting much farther back away from the stage, the performers thus taking on a more spacious, large-scale sound but with less specificity to each individual instrument. Each presentation had its own appeal and my preference often came down to music selection and gear matching. All else equal, I generally preferred this aspect of the Benchmark with things like intimate singer/songwriter performances, jazz trios, string quartets, and the (somewhat elusive) well-recorded hip-hop in my collection. Meanwhile, the Denafrips favored big band, most metal, electronic music of all sorts, and large-scale orchestra pieces. But factor in associated equipment and it’s tough to definitively categorize the results by genre.
I did find situations where the somewhat relaxed nature of the Pontus II proved too much of a good thing. In my collection that meant the Meze Empyrean and Audeze LCD-2 were at times a bit too relaxed and soft up top to be a good fit. Using a more incisive amplifier such as the Niimbus or Pass Labs helped with those models, but I typically still preferred using the Denafrips DAC with a fairly well-balanced or even slightly brighter headphone, regardless of my amplification choice. Again, this is not an overly dark, syrupy presentation like some of the vintage NOS models I’ve owned throughout the years. There’s certainly a bit of euphonic coloration to the sound, and a noticeable softness to the higher frequencies, but I would not call it extreme. Those with visions of vintage TDA1543-based DACs with tube output stages and slow, gooey tonality should probably look elsewhere – hitting the “NOS” button on the Pontus II tilts the sound very slightly more in that direction but perhaps not far enough to scratch that particular itch.
After more component swapping, a trend emerged: headphones that didn’t respond well with the Benchmark in the chain tended to perform nicely when using the Pontus II. And vice versa. Which makes sense given their contrasting sonic natures. While it was possible to assemble a middle-of-the-road setup in which both models were enjoyable, the best results were attainable by playing to their respective strengths – and/or the weaknesses of the chosen headphones. The Benchmark would thus favor somewhat darker-sounding models like the Sony Z1R, vintage (pre-Fazor) Audeze, Sendy Audio Peacock, or the ZMF Atticus. Meanwhile the Denafrips played best with the previously mentioned Meze Elite and Focal Elex plus the Audeze LCD-24 Limited and the Kennerton Audio Thekk – all of which are well-balanced with superb treble articulation. Obviously amplifier choice would be a significant factor here but you get the idea.
Is it still worth using the Pontus II in a more straight-forward system, using something like a CD transport via coaxial connection? Sure. It remains a very enjoyable device even in that context, as the core nature of the sound signature will still shine through. But I’d argue the potential of the DAC isn’t fully realized until one feeds it upsampled data via the USB or I2S inputs. That could be seen as a lofty requirement, with Benchmark making no such demands from their users. Or, with a glass-half-full approach, we could take it as a sort of built-in upgrade path, allowing the device to grow with the user as their system expands. That all depends on one’s perspective. Either way, Denafrips offers a compelling value in the Pontus II.
The Schiit
The Schiit Ygdrassil stands out in this crowd by offering a platform rather than a singular DAC product. Benchmark and Denafrips (and the majority of other DAC makers) offer firmware updates as necessary, and that’s about the extent of it. Meanwhile Ygdrassil has seen an almost dizzying number of upgrades over the years, made possible by the modular design which separates digital inputs, USB input, output stage, and DSP engine into separate cards and thus allowing swaps to the latest/greatest versions while keeping the remaining components intact. This means someone who purchased a Ygdrassil at launch way back in 2014 could have upgraded at any point and brought the device up to then-current spec for just a fraction of the cost of buying a whole new device. I can’t overstate how much I appreciate this approach.
As far as circuit design, Schiit goes in a different direction than both the Denafrips and the Benchmark, instead using their own somewhat unique philosophy on audio reproduction. Ygdrassil employs some rather expensive Analog Devices AD5791 multibit D/A converters which to my knowledge have never been used in any other DAC. These are R-2R ladder DACs in chip form, typically used in medical devices rather than audiophile gear, here employed in a fully balanced configuration consisting of four chips total. An Analog Devices SHARC DSP runs their proprietary closed-form digital filter which (unlike almost everything else out there) retains all incoming samples in their original form. The power supply is a beefy shunt-regulated dual mono dual transformer design with extensive local regulation and complete separation between the digital and analog sides. Digital inputs are subject to Schiit’s proprietary Adapticlock system featuring precision clocking for use with quality transports as well as a fallback clocking system for more pedestrian sources, thus ensuring maximum compatibility. USB is handled by yet another proprietary system called Unison which, much like Denafrips on their Pontus II, is implemented with a custom microcontroller rather than using an off-the-shelf solution. The analog stage involves fully discrete, DC-coupled Class A FET buffers optimized for high current output. With nearly every aspect of this device being completely custom or fully proprietary, Schiit clearly likes to think outside the box.
Physically, Ygdrassil is a rather imposing device that weighs around 21 pounds – slightly heavier than the Pontus II though less dense considering the size differences. Like the other two DACs represented here, it sports fully balanced XLR outputs plus a single-ended RCA option. Input choices are also on par for this group. The aesthetic is in keeping with Schiit’s numerous other products, although something about the chunky full-width design makes this (and the matching Ragnarok amplifier) the most aesthetically pleasing of their lineup in my humble opinion. I especially liked the ornate array of ventilation holes on the top panel – a bit of functional flourish on an otherwise understated machine. On the downside, I had a hard time making sense of their tiny and obscure front-panel markings to see which input was active. This does not appear to be improved on the Ygdrassil+ despite a very mild styling refresh.
Schiit recommends the device be left powered up continuously for thermal stability and therefore maximum sonic performance. I believe the impact of this recommendation has been somewhat exaggerated on various audio forums, but I nonetheless kept all three DACs powered up for the entirety of this project just to be on the safe side. Only when listening was finished and it was time for photography did I break the devices down and stage them as needed.
My first impression of the Ygdrassil was one of superb tonality. I was immediately drawn to the dynamic yet natural, organic presentation of the device, and nothing I heard over the next several months of listening did anything to change that opinion. While the Benchmark grabbed attention right away with its stunning detail extraction, and the Denafrips impressed with its lush, lavish tone, Schiit seemed more concerned with the uniform fluidity of the presentation. It seemed so unforced, so effortless, that despite the considerable amount of information being presented with respect to resolution and fine detail, it remained supremely approachable. And did I mention Ygdrassil rocks? It confidently pulls off bass “punch” and deep low-end authority with a sense of ease.
Starting with the Pass Labs HPA-1 amplifier and Audeze LCD-24 Limited, I was treated to convincingly haunting melodies and well-defined instrument layering on Thirteenth Step by A Perfect Circle. Generally more cerebral than aggressive, this album has some excellent harmonies along with interesting chord progression and time signatures, all wrapped up in a package that initially might seem pretty straightforward. Ygdrassil was spot-on in letting me listen on my own terms – dig deep into the complexities or just sit back and let it wash over me.
I had a great time rotating through three different takes of the above-mentioned album (thanks to Roon’s handy “versions” tab) with the original CD release being acceptable yet a bit dynamically muted in comparison to one ripped from the 180g vinyl release. Interestingly enough, my favorite of all was the live version, which is one of those rare instances where both the musicianship and the recording quality actually edge-out the studio version in various ways. Not to say the Benchmark and Denafrips DACs gave me a different preference, but it was the Schiit DAC that called the most attention to the nuances between each release.
Circling back to Hiromi’s Voice, I perceived the Ygdrassil as being somewhere in the goldilocks zone between the caffeinated exuberance of the Benchmark and the full-bodied silkiness of the Denafrips. There was potent drum impact in the lower registers, and graceful piano shimmer up top, plus that liquid midrange which again had a beautiful sense of projection into the listening space. It’s not that the Ygdrassil necessarily sounded more “neutral” than the other two DACs represented here but rather that each had its own interpretation of what neutral, natural music reproduction should actually sound like.
Soundstage again split the difference between the intimate, front-row DAC3 B and the more distant, large-scale Pontus II presentations. It didn’t have the most width I’ve ever heard, nor was it the most grand in size, though neither aspect was particularly lacking in absolute terms. What it did do better than either competitor was give a palpable sense of depth, as well as a more fleshed-out feeling of the actual performance space or venue itself. This obviously won’t apply to a good portion of the recordings in my library which just don’t contain much of this information in the first place. But I was nonetheless surprised I could discern it as often as I did – keeping in mind that I listen to a wide variety of music, much of which would never be played at a hi-fi show.
Swapping out amps and headphones, I struggled to find anything that made a truly poor match with the Ygdrassil. It got along swimmingly with tubes and the warm, rich Audeze LCD-2 or Meze Empyrean. Yet it did not sound overbearing in the treble region even with the lit-up Sennheiser HD800S driven by the uber-detailed Niimbus US4+. Of the three DACs represented here, Schiit’s entry was probably the most universally agreeable in terms of system matching.
If I’m picking nits, I’d say the vigorously dynamic nature of the presentation could sometimes come on too strong in certain contexts. An example involves the punchy Meze Elite, which generally pairs well with Ygdrassil and just about any amplifier. But when playing delicate jazz or classical, it could sometimes feel like too much momentum in the pacing, where the experience would be better served with a more nuanced touch. Again, it’s a small thing, but the Denafrips and especially the Benchmark both did better here.
Oddly enough, my favorite pairing ended up being the brilliant Pass Labs HPA-1 headphone amplifier driving the Empire Ears Zeus XRA in-ear monitors – each custom molded shell sporting 14 tiny drivers along with an 8-way crossover. I would not have anticipated these stout class A designs to synergize so well with diminutive IEMs but, thanks to the extremely low noise floor in both devices, here we are.
I felt Ygdrassil’s XLR outputs sounded slightly better than their RCA counterparts, but not drastically so. If using an amplifier featuring both options I’d choose the former, yet it wasn’t a big enough deal to stop the (single-ended only) Pass amp from being such an excellent match. On the input side, AES seemed to sound best in my chain, but USB was very enjoyable as were all the other inputs for that matter. The Ygdrassil will readily show improvements in transport quality, despite still sounding decent fed by a basic laptop. I tried upsampling via Roon and found the result… slightly different but not really better. I ended up just sticking with native sample rate for most of my listening. So once again Schiit splits the difference between the competitors in terms of fuss factor.
Worth noting is that the Ygdrassil doesn’t do DSD at all – which is almost shocking considering the specification arms race we see from most brands these days. But guess what? This “limitation” ended up not bothering me one bit in real-world listening. Roon makes it a breeze to convert DSD to PCM whenever necessary (which is honestly not very often despite my ~10,000 album collection) and I feel like little or nothing is qualitatively lost in the experience.
The Ygdrassil has been around for long enough to amass an extensive collection of reviews from all over the web. You’ve likely encountered a few of them during your internet travels. It’s important to keep in mind that due to the modular platform and numerous upgrades over the years, many of these impressions will no longer be all that relevant to the current state of the device. Even first-hand impressions also become out of date. I had spent time with an Ygdrassil some years back, using the Gen 3 USB solution and the Analog 1 output stage. It was very enjoyable but did not prepare me for how excellent the current version would be.
Conclusions
As a percussionist, I am often asked about my favorite drummers. I could go on for hours if you let me but my simplified response usually involves three names, representing a diversity of styles/genres: Tomas Haake of Swedish metal band Meshuggah, Matt Cameron of Soundgarden (and later, Pearl Jam), and influential session drummer Idris Muhammad who appeared on countless classic jazz performances and later released some excellent funk albums of his own. Can I objectively decide which of these three artists is “best”? Of course not. I like them all for their unique abilities and the varied performances they give.
The three DACs represented here are in a similar situation. Each one is an excellent performer in its own way, and they all have a somewhat unique flavor that works better than the rest in certain situations. Each has a quality build, subjectively good looks, plenty of connectivity, and a hard-earned reputation for excellence, so the choice really comes down to individual preferences and system matching.
Want a blazing fast, highly incisive sound to unearth gobs of microdetail while giving quick, punchy, tuneful bass? The Benchmark DAC3 B does just that. Prefer a richer, meatier take on things, with potent bass slam and a delicate yet nicely extended top end? Look no further than Denafrips with their Pontus II. Looking for a liquid, flowing presentation that prioritizes tonal purity and dynamic gusto? Schiit’s Ygdrassil may be the DAC for you. In the vast ocean of DACs on the market, these three models each stand out as exceptional choices.
