FULL REVIEW: MSB Gold Link III/PowerBase and Bel Canto DAC2

[jude]

MSB Gold Link III/PowerBase and Bel Canto DAC2 Digital-to-Analog Converters
Redbook CD In High Resolution

Sam Tellig of Stereophile called upsampling the "magic bullet" of redbook CD playback. Still others claim "upsampling" is nothing more than marketing mumbo jumbo, and no different than what oversampling CD players have done since the early days of digital audio. Which is it? I'm not exactly sure, but I do know that the two "upsampling" DACs (digital-to-analog converters) I'm reviewing here shared some obvious sonic similarities that set themselves apart from all of the non-upsampling CD players I've listened to, and mostly to their advantage.

The first two upsampling DACs I've had the good fortune to live with in my rigs were the Bel Canto DAC2 and the MSB Gold Link III with PowerBase power supply upgrade. The DAC2 upsamples to 192 kHz (using an Analog Devices AD1896 sample rate converter), and the Gold Link III to 96 kHz (using a Crystal CS8420 sample rate converter). In terms of form factor, these two DACs could hardly be more different. The DAC2 is a smallish black, brick-shaped metal box (measuring 3.5" x 3.5" x 9.5") that tips the scales at only 3.5 pounds. The Gold Link III is a wide, deep, low component (17” x 14” x 1.75”) that weighs 20 pounds; the PowerBase is the same size, and weighs 21 pounds -- so coupled with the PowerBase, the Gold Link III is quite a physical beast (flimsy shelves need not apply for the job of supporting this combo). Aesthetically, neither the Bel Canto nor the MSB will inspire prose, both with very simple, utilitarian appearances -- if you want Rowland, Wadia, or Levinson visual impact, y'ain't gettin' it with these boxes. One thing both of these units offer is good build quality. Both use rugged casework and connectors, the latter being of particular importance to me, as I frequently switch out cables and components; and lesser jacks have occasionally been problematic for me over time, despite my babying of gear.

The Bel Canto offers both S/PDIF (via RCA jack) and Toslink digital inputs (there's a button that lets you choose between the two), and it outputs to single-ended RCA analog outs. The MSB offers AES/EBU (and MSB Network 192 kHz) digital input via a XLR jack, S/PDIF digital input via RCA jack, and Toslink (it chooses the digital input automatically). The MSB outputs to both balanced analog outs (via XLR jacks), and single-ended analog outs via RCA jacks. The MSB also has a very useful analog pass-through (via an additional pair of RCA input jacks) that allows SACD enthusiasts like me to listen to both SACD and redbook without requiring two sets of inputs at the headphone amp -- almost every available SACD player outputs SACD audio via its analog outputs only, and the analog pass-through will allow that SACD audio to be passed through to its analog outputs. The MSB will output whatever is coming through the analog pass-through if no digital signal is otherwise present -- again, a great feature for SACD'ers, or for anyone who might have a headphone amp with only one analog input, but more than one line level source component to hook up to it.


MSB Gold Link III with PowerBase Power Supply

The first of all the digital review components to arrive, and so the first "upsampling" source I'd ever heard, the MSB was an ear-opener from the get-go. Using my Sony SCD-C333ES as both a standalone source for comparison, and also as a transport connected to the MSB via Acoustic Zen “MC^2=Zen” and Cardas Neutral Reference Video S/PDIF coaxial digital cables, I was able to perform easy direct comparisons between the Sony's built-in DAC/analog output section to the MSB's since both of my primary headphone amps each sport at least two sets of inputs.

After having lived with my Sony SCD-C333ES for quite some time now, I would characterize its standalone CD playback as generally neutral and fast, but on the dry side -- pleasant, but not as convincingly organic as a motivated audiophile might aspire to (especially one who listens to a lot of acoustic and live recordings). Even without the upsampling activated, the Gold Link III possessed a more authoritative, meatier tonal balance than the SCD-C333ES’s redbook playback. With its upsampling switched off, the Gold Link III gave up some bass neutrality to the SCD-C333ES, with recordings of a tonally heavier nature exposing a little bit of bass overhang from the Gold Link III in direct comparison. This very slight sense of boominess did not appear to be caused by the MSB's deeper extension (the MSB was certainly more impressively reachy down there than the SCD-C333ES), as this mild over-ring continued up through the mid-bass region as well. From the lower-mids on up, and with upsampling still off, the MSB exhibited greater image opacity and rounder image edges than the SCD-C333ES. The Gold Link III also maintained more neutrality through this broad expanse (and arguably the most important portion) of the audible frequency spectrum than it did from the mid-bass down. I don't want to say too much more than what I've now said about the MSB with its upsampling off, because I can't imagine why anyone who buys a Gold Link III wouldn't use the upsampling feature; and anything I've described so far that might read negatively disappears, for the most part, the moment you flip that upsampling switch to its "on" position. So onto upsampling, a la MSB....

On the back panel of the Gold Link III is a simple toggle switch labeled simply “Option”. In this instance, “Option” is as much a misnomer as just about any I’ve seen in audio -- it should be labeled “Mandatory”, as a simple flip of this switch fires up the Gold Link III’s 24-bit/96-kHz sample rate converter; and, friends, the Gold Link III goes from quite good to call-your-friends-and-tell-‘em-all-about-it outstanding when the upsampling’s engaged. If you’re like I was before I tried the MSB, wondering if these modern “upsampling” DACs do anything sonically special, I’m here to tell you yes they do (at least the MSB and Bel Canto DACs in this review do anyway).

The natural questions that might follow are "how evident is upsampling?" and "what does it sound like?" To answer the first question: it was evident enough that my colleagues at the office helped me conduct an informal, blind comparison, and I was able to pick it out every time. To answer the second question: its most obvious effects involve individual sonic object edge definition (sharper, more defined, more natural); also, the sense of venue is noticeably improved -- especially with well-executed live recordings, the sense of the acoustic space is significantly more evident with upsampling than without. Long story short, to me, CD playback with the upsampling engaged sounds more convincing, more real.

To elaborate further on my answer to the second question, I think it important to briefly discuss some of my views on hi-fi versus the real thing. <wordy discourse mode on> First of all, it is my opinion that folks who have very little experience (or no experience at all) with the sound of live, unamplified performances tend to like their hi-fi at least a bit too analytical, too hot, too treble spotlit, because their point of reference is primarily -- and sometimes only -- electronic reproduction; so the concept of "if I hear more, it's better" comes into play in the absence of reality as a reference. If one could hypothetically turn up the treble on a live performance -- say a symphony orchestra -- one might hear the triangles and cymbals more clearly and loudly, but it wouldn't be representative of these instruments' proportional energy within the scope of the unaltered performance. More isn’t always better. However, without having experienced the real thing, oftentimes more is equated with better. Thus endeth my sermon on that. Thus beginneth my sermon on something else, starting with another example: if you've ever been to a live piano recital, particularly in a smaller venue (like a largish living room), you know it can actually get quite loud (depending, of course, on the material being played). Assuming it's a nice piano, and the room acoustics are at least somewhat good, that the performance gets loud isn't bothersome; that is, there's something, despite the high amplitude, that keeps the piano from becoming sonically grating -- there's a certain ease of realness that, despite the power, keeps it very listenable. That great presence of the real thing also means full conveyance of information during even very quiet passages -- I'm talking about the virtually silent notes that still have a fleshy, three-dimensional presence you can almost reach out and touch the trailing wisps of, because, well, you're there. When it comes to trying to duplicate that magic of real electronically, a good analog source comes closest, to my ears. And, after this lengthy discourse, that's what I'm getting at: digital has generally, up until somewhat recently, not done this very well. Fortunately, the technologies and techniques for getting music from the actual performance to the CD has improved over time, which has been the most significant step, in terms of fidelity, we've made with CDs over the years, and which provides audio enthusiasts with more to take advantage of with the CD format.

So just what does a good analog rig do that digital doesn't? Before I answer that, I'll admit the following: I'd be willing to bet that most analog rigs fall short of digital in terms of measured frequency extremes. CD also measures better than vinyl in terms of channel separation, signal-to-noise (S/N) ratio, dynamic range, and has no measurable pitch errors from things like wow and flutter. Despite what the measurements say, it's what good analog does do, within an almost certainly more limited frequency range and dynamic range, that makes it better, more live, more real, than CD. What really good vinyl can do, it does better than any digital that I've heard (no matter what the measurements say) -- it sounds more dynamic, more accurate, more extended, more real. In virtually all instrument-measured performance parameters, digital does more, yet good vinyl still sounds better and more genuine -- most digital comes off as a factory-made-sounding reproduction; good analog sounds as if crafted by an artisan, the performance carved out like a sonic sculpture in your listening space. With good analog, sonic image objects not only take on more opacity, they are framed out more three-dimensionally, have more natural texturing and timbre than digital; and image edges with good analog aren't merely anti-aliased, they're natively smooth and, well, pretty darn real. Listen to a very good analog rig, and I think you'll agree that it sets a very high standard for reproduction and musicality. If you’ve never heard a good analog rig, you owe it to yourself to beat feet to an audiophile salon near you that specializes in selling good (or better) analog gear, and staffed with at least one person who knows how to set up such equipment properly.

The new high-resolution digital formats, SACD and DVD-A (up to this point, I've only had significant experience with SACD), have raised the digital bar, getting closer to (but still falling short of) a very good analog standard. Good SACD provides more of the analog-like natively smooth edge definition; gets much closer than most CD audio I've heard to analog's image opacity; provides better dynamic performance than regular CD; but still falls short of analog in terms of the you're-almost-right-there fleshiness -- that is, the sonic image's front-to-back dimension is more truncated than with good analog, but still impressively good. One of the most significant issues with the high-resolution formats is, of course, the still-limited library of high-rez software; so further optimizing playback of regular CDs will still be of tremendous value to those possessing a significant collection of regular CDs, which would include most fellow audio enthusiasts I know. Upsampling -- as exhibited by the Gold Link III and DAC2 -- doesn't get redbook CD to analog (or even SACD) levels, but it sure does move it in the right direction. (takes a deeeeep breath) Which brings me back to the review....</wordy discourse mode off>

With its upsampling on, the Gold Link III/PowerBase is among the most easy-to-listen-to, pleasant-sounding, analog-like CD playback devices I’ve heard. The first thing I noticed when I turn the “Option” switch on is the immediate expansion of soundstage. It's not that the instruments and voices are spread any farther apart, as much as it is that the venue itself becomes more spacious, with a greater sense of breathable, excited air between the image objects -- and I don't mean just laterally. With upsampling on, I perceived that images placed toward the back in many recordings were more alive, more defined, due to the presence of more anchoring spatial and textural detail; this without betraying their background placement. If you want to hear an impressive sense of soundstaging, several of the Mapleshade Records CDs I’ve purchased do the trick nicely. Mapleshade’s recordings are made using only two to four microphones, and no cables longer than 20 feet; and (from Mapleshade’s site) they use “no add-on EQ, reverb or noise reduction electronics.” Start by picking up Mapleshade’s Mapleshade’s Music Festival (Wildchild!/Mapleshade 08132) and cue up “Nick of Time” by Sunny Sumter with the Larry Willis Quintet (track four). Versus all of the non-upsampling players I’ve heard so far, the Gold Link III/PowerBase did a better job of more distinctly layering, front-to-back, Sunny’s vocals in front of the bass that the track opens up with; and without the slightest sense of her forward vocals obscuring that accompanying bass. Both clearly occupy their own individual physical places in the soundstage in all three dimensions, Sunny being way out in the fore. The piano gently trickles in at 0:55 of the track, and grows in prominence as the track progresses -- but even the firmest piano notes stay spatially behind Ms. Sumter more convincingly with upsampling than without. Keeping that CD in the tray, advance to The Blue Rider Trio playing “Gallows Pole,” and, with any CD player you should be able to tell that Ben Andrews’ head is moving left to right while he sings. The Gold Link III/PowerBase with upsampling reveals that his head is not only moving side to side, but also seems to be rocking back and forth a bit; you can even tell when he’s briefly looking down at his guitar. Are we talking SACD-on-a-good-day-and-good-disc soundstaging here? In the comparisons I could do (with hybrid discs, and with single-layer SACDs for which I have the CD counterparts), the answer's no. The soundstage improvements over non-upsampled, however, are notable, and get redbook that much closer to high-rez performance.

What I consider one of the biggest strengths of the Gold Link III/PowerBase is not just the expanded three-dimensional soundstage, but the image solidity it maintains even as it achieves that. There is very little sense of image diffusion, and so each sonic image object maintains its intended opacity. There is no sense of artificial etherealness applied, and it makes for a meatier, more real presentation. In addition to its very extended driving bass, this image solidity means an emphatic “yes” when asked if the Gold Link III/PowerBase combo rocks. Want solid? Want driving? Buy Tomahawk’s eponymous album (Ipecac Recordings IPC-18). Another Mike Patton project, Tomahawk is a raucous, obscenely fun album that all of Patton’s Faith No More fans would almost certainly love. Track 10 (“Malocchio”) is a dense, mid-bass-heavy romp that doesn’t let up from the first note. Driven by the very deliberate, distorted, plodding bass of Kevin Rutmanis (also of The Melvins), opacity is key with this track, and the Gold Link III/PowerBase serves it up solid as a bar of lead. I feel confident assuming Patton and the boys never had the words “light” and “airy” in mind as they laid this beast of a tune down.

I mentioned earlier that the Gold Link III/PowerBase, with upsampling off, had a wee bit of excess bass weight. When the upsampling’s switched on, however, that slight sense of heaviness disappears. It seems as though, with upsampling off, the Gold Link III/PowerBase concentrates the bass energy into too tight an area; but when the upsampling is turned on, the low frequency brawn becomes more naturally and widely lavished upon the soundscape, and so the problem solved. Thankfully, the Gold Link III/PowerBase bass extension isn’t reduced in the least by the upsampling, and still seems to reach down to the Earth’s inner core. Always a fun track to wring a system out is Weinberger's "Polka and Fugue" from Schwanda the Bagpiper on XLO/Reference Recordings' Test/Burn-In CD (XLO/Reference Recordings RX-1000). It’s a macrodynamic bruiser of a track, and, in terms of low bass, it’ll throw things at your rig (like sub-bass organ notes) that are more tactile than aural -- notes and sounds that shake and huff at your ears and head, and the reproduction of which are crucial to extracting the fullest, most lifelike headphonic experience of this piece. The Gold Link III/PowerBase combination kicks this track into full blossom, and meets this track’s macrodynamic bursts without a drop of sweat. In terms of bass performance, let me equate the Gold Link III/PowerBase combination’s prowess with a headphone-geek-specific reference: in terms of bass, the Gold Link III/PowerBase is the HeadRoom Max of digital source gear -- deep, solid, reference-quality.

Upsampling on or off, the Gold Link III/PowerBase combo’s tonal presentation from the midrange through treble is generally neutral, with just a hint of midrange richness that I find very attractive. Treble extension is very good, if not as wow-that’s-exospheric as the Bel Canto DAC2’s treble reach (more on that in just a bit). One of the remarkable things to note about treble performance in upsampling-on mode is the near absence of treble grain -- the bane of redbook-CD-as-audiophile-source, and one of the primary causes of what many refer to as “digititis.” It’s a treble-smoothing effect sans any sense of roll-off, and it’s one of the things that gets the Gold Link III/PowerBase closer to analog than most redbook digital components I’ve heard.

Throughout this review so far, you’ve probably noticed that I keep referring to the Gold Link III and PowerBase together. Why? Because if you’re going to drop the dough for a Gold Link III, the PowerBase (or some other high-quality, beefy, compatible power supply) is mandatory, in my opinion. Replacing the PowerBase with the Gold Link III’s comes-with-it wallwart power supply means less bass, slightly diminished soundstaging abilities, and a general decrease in overall refinement. Simply put, a noticeable amount of the Gold Link III’s magic takes a hike without the PowerBase. As far as I’m concerned, the Gold Link III/PowerBase is not so much a combination as a single product in two separate chassis. Getting the Gold Link III? Get the PowerBase. Simple as that.

Is there anything I don’t like about the Gold Link III/PowerBase? Of course -- I’m an overly critical audio enthusiast, just like most of you reading this. First of all, the combination is physically heavy and quite large. It’s also what most would consider an expensive combo, with a retail combined price of nearly USD$ 2,300.00 ($1,895.00 + $399.00). Treble extension is good but not standard-setting. Soundstaging is excellent, easily edging out all of the non-sampling CD players I’ve heard, but, again, not standard-setting. It also doesn’t decode HDCD. On the whole, these are minor quibbles considering the overall performance, but worth noting.

To be continued, with the full review of the Bel Canto DAC2, a couple of photos, and the summary/conclusion of the dual review of the MSB Gold Link III/PowerBase and Bel Canto DAC2 (and some notes about my experience with digital cables).