So when I learned, at the 2013 Consumer Electronics Show, of Musical Fidelity's new M6DAC, intended to offer "reference quality" performance at a relatively affordable price ($2999), I asked for a review sample.

The M6DAC
Housed in a black-painted steel enclosure, with ribbed aluminum side panels and a black or silver aluminum front panel, the M6DAC matches the styling of Musical Fidelity's other M6-series amplifiers and CD player. A rectangular, blue-illuminated LCD display dominates the left-hand side of the front panel, showing the source playing, the sample rate, and whether that source is set for a fixed output level or adjustable level. A rear-panel switch allows the M6DAC's output level for each input to be set to fixed or, within a range of ±10dB, independently adjusted with Up and Down buttons on the front panel and remote control.

The M6DAC offers a choice of two reconstruction filters, labeled Fast Roll Off and Slow Roll Off, selectable with a front-panel Filter button duplicated on the remote. The display momentarily shows which filter has been selected, as well as whether de-emphasis has been set to Auto or Off. As well as the buttons mentioned above, the array of 10 pushbuttons on the right-hand side of the front panel allows six different sources of data to be selected: balanced AES/EBU on an XLR jack, two S/PDIF electrical inputs on RCA jacks, S/PDIF optical on a TosLink jack, asynchronous USB on the usual USB Type B jack, and Bluetooth.

On the rear panel, a badge proclaims that though the M6DAC is made in Taiwan, it was designed in England. On the central section are a pair of RCA jacks for the single-ended analog output, and a pair of XLRs for the balanced analog output. To the right of these are the IEC AC jack and trigger in and out jacks; to the left are the digital inputs, as well as a screw connector for the supplied Bluetooth antenna and three digital outputs: AES/EBU on XLR, S/PDIF optical on TosLink, and S/PDIF electrical on RCA. The Bluetooth connection needs to be set up with a paired device in the usual way, entering a code on the device when instructed. The USB2.0 port functions without a driver program with Linux 2.6.33 or later and Mac OSX 10.6.4 or later. The drivers needed for Windows XP (SP3), Vista, and 7 are included on a CD-ROM. The manual doesn't mention Windows 8, but does include a comprehensive section on getting the optimal performance from Windows machines.

The plastic remote includes buttons for operating both the M6DAC and the M6CD player. Two of its buttons are not duplicated on the front panel: Mute and Display, the latter offering two levels of illumination as well as Off.

Circuitry
When I lifted off the M6DAC's steel top panel, I saw an interior dominated by a large, double-layer, green printed circuit board, with a cutout at the front center to accommodate the AC transformer. This transformer is flanked by blue 4.7µF polypropylene capacitors: 18 on one side, 14 on the other. Almost all the circuitry is concealed by two black metal covers, the one on the left marked "Universal Precision Regulated Digital Power Supply" and offering choke smoothing, the one in the center marked "Fully Balanced Precision Dual Mono 192kHz 24 bit Upsampling Quad DAC." Only the digital-input receiver components are out in the open, with a Bluetooth LSI on the main board and the XMOS USB receiver mounted on a small daughterboard behind the digital input jacks. All digital inputs are galvanically isolated, to minimize noise being injected into the circuit ground.

Removing the shield from the signal-handling stages revealed that the i²S data from the digital inputs are fed to a Burr-Brown SRC4392 chip, an asynchronous sample-rate converter found in many of Musical Fidelity's digital products. Whatever the incoming sample rate, this chip upsamples it to 192kHz and feeds left- and right-channel 24-bit data to a pair of Burr-Brown DSD1796 DAC chips, one per channel. This DAC is a two-channel part; Musical Fidelity uses the two channels in differential mode to provide a balanced mono signal path, with each channel's circuit physically separate from the other's. I/V conversion appears to be performed by a Texas Instruments LME49720 dual–op-amp chip, this a high-precision, high–slew-rate, low-noise part—followed by another LME49720.

All passive components are surface-mount types. The output stage for each channel appears to be based on a third LME49720, the two op-amps within the chip driving the hot and cold phases of the balanced output signal.

Tangled Up in Bluetooth
Ever since his company began distributing FoxL Bluetooth wireless speakers in the UK, Musical Fidelity's Antony Michaelson has been a big fan of streaming audio wirelessly via Bluetooth. I am not a fan, for two reasons: First, the limited bandwidth of a Bluetooth connection requires that the audio data be encoded using a lossy algorithm, which, if you're playing uncompressed or lossless-encoded files, imposes an unnecessary reduction in fidelity, and if you're playing files that are already lossy compressed, runs the risk of multiplying coding artifacts. Second, the first lossy Bluetooth codec I played with, A2DP, sounded poor and measured worse. An early version of the reputably superior aptX codec, which runs at 352kbps, and which I tried using a Creative Labs BT-D5 dongle on an iPod Touch, sounded and measured only slightly better than A2DP when I tried it with the Chordette Gem DAC, which Art Dudley reviewed in January 2011.

I was pleasantly surprised, therefore, when I tried using the M6DAC's Bluetooth connection using my 2012 MacBook Pro as the data source. The M6's display confirmed that the aptX codec was in use, and while music files ripped from CD were flattened in perspective and uninvolving overall, they also sounded smooth and inoffensive, as if the edges had been rounded off a little. By comparison, the plain-Jane apt codec, which I used when streaming music via Bluetooth from my iPad 2 to the M6DAC, sounded annoyingly harsh.

According to Musical Fidelity, while aptX is limited to 16-bit/48kHz data, the M6DAC provides the aptX chip with its own power supply and directly couples it via i²S to the sample-rate converter chip. Certainly I found that the M6DAC's Bluetooth connection worked well for casual listening, and especially for streaming spoken-word Internet stations such as Manhattan's NPR station, WNYC, regardless of cascaded codecs.

Sound Quality
For critical listening, however, I used the M6DAC's USB or AES/EBU connections, both of which worked reliably up to 192kHz. And with either connection and with all kinds of music, I could hear no difference at all between the Slow and Fast Roll Off filter settings. One reviewer said the difference was "subtle"; I think it nonexistent. I discuss the behavior of the two filters in this review's "Measurements" sidebar, where it appears there was no measurable difference between the two filters below 96kHz. It is not surprising, therefore, that they sounded identical. A more cynical reviewer might suggest that the Filter button was included for marketing reasons; I continued my auditioning using only the Fast Roll Off filter.

Several of the DACs I've recently reviewed include a high-performance volume control; three that spring to mind are the NAD M51 Direct Digital ($1999, July 2012), MSB Diamond DAC Plus ($21,995, October 2012), and Resolution Audio Cantata Music Center ($6495, November 2011). Each is an excellent-sounding DAC topped off with digital-domain attenuation. Other than the NAD, each also comes with a somewhat steep price tag.

Enter Wadia
At $1299, the Wadia 121decoding computer is more in line with M2Tech's Young DAC ($1499, May 2013), which I've also been listening to recently—though the Young lacks a volume control, and adding its Palmer battery power supply ($1249) bumps it out of the Wadia's price range.

At exactly 8" square (and 2¾" high), the Wadia has a modest footprint very similar to that of my Benchmark USB DAC, and the same appearance and dimensions, as Wadia's 151PowerDAC mini DAC–integrated amplifier and 171iTransport. The 121decoding computer has an external power supply that keeps its chassis size and weight low, as well as the amount of heat it produces. In fact, the Wadia ran extremely cool—it never felt more than barely warm—and is surprisingly light for such a solid-looking block of metal.

The case has a dark, sandpapery, matte finish, and a conical rubber foot at each rounded corner. The look is attractively modern and understated, and the side panels are nonresonant. On top, there's a coin-like disc of gray metal at each corner; a dimple at the center of each disc accepts one of the footers of a Wadia companion product, for easy stacking.

On the no-nonsense front panel are, starting at the left: a blue LED power indicator, a ¼" headphone jack, an IR sensor for the remote control, and three stacks of six small blue LEDs each: the first to indicate the volume level, the second to show the sample rate, and the third to indicate the digital input chosen and the phase setting.

On the rear panel, from left, are: left and right balanced and unbalanced audio outputs (both sets are active at all times), BNC and RCA coax digital inputs, TosLink and asynchronous USB input jacks, a three-pin AES/EBU socket, and a seven-pin connector for the external switching power supply. All digital inputs, including the USB, can accept data rates up to 24-bit/192kHz.

Included is a gorgeous and simple aluminum remote to control the volume, input selection, mute, display brightness, phase, and other settings. It also has buttons for controlling Wadia's 171iTransport.

I'll get on my soapbox for a second here and wish again that there were some way to control the volume on the product itself—I often leave the remote near my listening position, which is several steps away from my music sources and the DAC. When I used the Wadia as a preamp and volume control, I often started playing some music, then leapt for the remote to turn the volume down. Or up. And when I held down one of the remote's volume buttons, the response was a little pokey. I much prefer a knob I can grab and set quickly. Since the volume display on the 121's front panel is only a crude representation of the actual volume setting—120 steps of 0.5dB each divided by six LEDs equals 20 steps per LED—there was plenty of back and forth to get things right before I settled down to listen. Even a simple numeric display would have greatly helped. Sorry to go on about this, but MSB really got this right—their Diamond DAC Plus has a physical volume knob and a numeric display.

Setup
When you first plug in the 121decoding computer, its volume is set to 0 (–60dB), to prevent any unpleasant surprises. No printed instructions are included; instead, you get a 27-page owner's manual as a pdf file on a thumb drive. I decided to try setting everything up without referring to the instructions and mostly succeeded, but did need to read the manual to fully understand the remote's Mode and Enter buttons.

A certain sequence of button presses lets you set the headphone jack's sensitivity to match your particular headphones; another sequence lets you fine-tune the DAC's overall output level, for system matching. The latter is especially useful if you use the 121 with another preamp. I found that the default setting matched my other DACs perfectly when run through a preamp, so I didn't change anything.

According to Wadia, the 121 has no volume-control bypass and no need for one. Digital inputs are upsampled to 32-bit/1.4MHz, then processed with Wadia's proprietary DigiMaster interpolation filtering algorithm, which, they claim, maintains resolution at all volume settings. Sure enough, I could hear no difference between using the 121 as a preamp or through another preamp, at any volume level. In the latter case, I'd just max out the volume and go.

Squonk
The 121decoding computer was installed in my system for about a week before I got down to any serious listening. During that time, its sound seemed pleasant and polite overall, reminiscent of the sound of the dCS Debussy ($11,499), which had been here a few months before (January, December 2011; February, September, October 2012).

M2Tech's Young, the DAC that immediately preceded the Wadia 121 in my system, proved a hard act to follow. The Young's seductive, luxurious sound had successfully drawn me away from the ultimate accuracy I usually prefer. One of the last albums I'd been listening to with the Young was a remastering of Genesis's A Trick of the Tail—I'd ripped both the CD and DSD tracks from the SACD/CD (EMI 0 65541 2)—which features glorious bass, though a tad extra hardness all around. The Young tamed the top end wonderfully; when I switched to the Wadia, some of the hardness crept back in, along with a slight opacity. I didn't fault the Wadia—this remastering is clearly not perfect—but nonetheless noted the change.

John Atkinson dropped by for a few days in March to present a talk, "Just How Absolute is Recorded Sound," to our local audio club. (Packed house. They loved it!) The Wadia was in the system, and the evening before the meeting, we listened to some music.

One album I'd just downloaded from HDtracks was the new remastering of War's The World Is a Ghetto (24/96, Avenue/Select/HDtracks). Not only is it a stunning example of early-1970s recording prowess, the HDtracks version includes an awesome bonus: a rehearsal take of the title track. And it's staggering, alone worth the price of admission—it showcases what a stellar group of musicians can do with a live take. Sure, it isn't technically or musically perfect, but the sheer sense of the musicians' presence in the studio is remarkable.

From $43,325 to $449. Yowseh!!—the MSB costs almost 100 times as much as the Schiit! Was this even fair?

The Bifrost wasn't warmed up, and it certainly hadn't settled in—but who could resist a little listening? I switched on the Bifrost, selected the S/PDIF input, and tapped the screen of my Sooloos music server to bring up a bunch of Turtles tunes we'd been listening to only moments before. "I really want you, Eleanor, near me. / Your looks intoxicate me, / even though your folks hate me . . ." Hmmm.

But I'm getting ahead of myself.

Schiit Audio is an interesting company. They've set some limits in how they operate that others might charitably call suicidal. For example, they design and make everything in the US, and claim that the bulk of the materials used in their products are also sourced from US manufacturers. They also claim to not even want to know if overseas suppliers (read: China) could better these costs. They sell direct, keeping their prices low. The build quality of the products I've seen so far lives up to Schiit's stated desire to make "something you can pass down to your children." They also provide a five-year warranty.

Apparently, company founders Mike Moffat (formerly of Theta) and Jason Stoddard (formerly of Sumo) have a sense of humor. On the front cover of the Bifrost's owner's manual is the following: "In Norse legend, Bifrost is the flaming rainbow bridge connecting the land of the gods (Asgard) to the earth (Midgard). Yes, rainbows, ha ha. Tell that to Odin and see what he thinks. I don't think you'll be laughing at him."

'Round the Chassis
Like most Schiit products, the Bifrost is a gray metal box containing the circuit boards, plugs, and switches; smoothly folded in a U shape around this is a slightly lighter sheet of bare, brushed aluminum. On top, to the left, is the silk-screened company logo; to the right, a grid of perforations. If you like, you can affix to the bottom four small rubber feet.

The front panel is simplicity itself: other than the Schiit logo and the Bifrost's name, there are only a single input selector switch (a metal disc about half as big as a dime), and three white LEDs to let you know which input is selected.

On the rear panel, from left to right, are: two RCA L/R output jacks; digital coax, USB, and TosLink inputs; a power switch; and a jack for the detachable AC power cord. Unlike with a lot of other DACs at this price, the power supply is built in, which gives the Bifrost a bit of heft. If feels very solid, and gets warm (but never hot) during operation.

Guts (and Glory)
The Bifrost can handle signals of any resolution up to 24-bit/192kHz at all of its inputs. It can be purchased for $349 without its asynchronous USB board; if you change your mind, Schiit's USB Gen 2 board can be added later. Also worth knowing is that the first-generation USB boards couldn't handle 24/176 data (24/192 was no problem). The USB boards are plug'n'play; if you've got the earlier board, I'm guessing you can replace it yourself; $100 if you do it, $150 if Schiit does it.

Another upgrade option is the new Uber Analog output stage, available for $70 ($100 if Schiit installs it). The Uber board also snaps right in, and sports the "more advanced" discrete analog output stage from Schiit's Gungnir DAC ($749/$849). At present, no Schiit DAC can handle DSD, though the company says that's coming.

I pulled the Bifrost apart, to verify Schiit's claim that they use all discrete components for the analog output stage, standard or Uber, and was impressed with the build quality. (You'll hear that a few times more before I'm done.) Seeing a product designed and constructed so well, and then checking the price again, is a little disorienting. Comparing the Bifrost's innards to those of the similarly priced and nicely built (in China) Peachtree DAC•iT ($449), you'd think the Schiit would have to cost more: The Peachtree's switching power supply is external, it doesn't include asynchronous USB, and it's not upgradable—but it sounded oh, so sweet in my system exactly one year ago. More on the Bifrost's sound in a bit.

The heart of the Bifrost is an AKM4399 chip—a 32-bit, delta-sigma D/A converter—coupled with a fully discrete (no chips) JFET analog section. There's no sample-rate converter, which means that data are not upsampled but are processed at their native rates. Indeed, each time a track was followed by one of different resolution, I heard quiet clicks from the muting relays inside the Bifrost as it made its adjustments.

No upsampling seems to be a point of pride for Schiit—another line they've drawn in the sand and promised never to cross. From their website: "Not just no but hell no. None of our DACs will ever do sample rate conversion. Our goal is to perfectly reproduce the original music samples, not to throw them away and turn everything into a mystery-meat soufflé. . . . We worked hard on a microprocessor-controlled, bit-perfect clock management system to ensure that all the original music samples going into Bifrost are delivered to the D/A converter. . . ."

Yet more good news comes in the form of Meridian Audio's Explorer ($299), a 4"-long USB digital-to-analog converter from a company that many hobbyists would name as one of the industry's premier digital specialists. That the Explorer is available not only at traditional Meridian dealers but also at a growing number of single-brand Meridian boutiques—locations now including Fort Lauderdale, Moscow, and Kuwait—may be seen as icing on the cake.

Ken Forsythe, Meridian America's director of product development, says his company hasn't turned its back on the high-end audio and video markets. "But if we want to be around 100 years from now, we have to go beyond our core. We think of computer-centric users as the new enthusiasts, so the question becomes: How, over time, can we grow them into core customers?" The answer, Meridian believes, is in the form of this, their first portable processor.

Description
Shaped like a Bic disposable lighter and sized like a Pez dispenser, the Explorer is built into a lightweight aluminum alloy tubewith a hard-anodized finish. A plastic cap at one end incorporates a USB mini-B jack—chosen because a full-size B jack would subject the internals to excessive stress—while a similar cap at the other end holds two 3.5mm jacks: one for headphones, the other for line-level audio output. The latter is combined with an optical digital-audio output—rather like the headphone jack on the back of an Apple iMac—to address the TosLink input of any outboard D/A converter. One might see that as an effort on the part of Meridian to emphasize both the Explorer's portability and its usefulness in a domestic system that already contains a high-end processor from Meridian (or anyone else, for that matter).

The Explorer requires only 5VDC, which it gets from the USB bus of the associated computer. It operates in asynchronous mode, using Meridian's proprietary software to reclock the incoming datastream. The converter chip of choice is the 24-bit/192kHz PCM5102 from Texas Instruments, followed by an analog section that Meridian describes as containing especially good-quality parts for one so humble. (As I could find no way to crack open the Explorer without destroying its aluminum shell, I didn't go poking around.) The Explorer's 130mW headphone output incorporates a 64-step analog volume control, while the line-output jack itself is fixed in level.

While I wasn't able to see for myself the Explorer's build quality, I can nonetheless comment on the whereabouts of its construction: Meridian's least expensive product is, like the rest of their line, made in England. Ken Forsythe relates this, too, to the company's efforts at "building their brand," so that new customers might someday step up to Meridian's more expensive gear: "We couldn't build products overseas and still be able to look our new customers in the eye and say, 'This is built in the same place as our finest products.'"

Installation and setup
In addition to a black-velvet travel pouch—another inducement to portability!—the Meridian processor is supplied with a very flexible 6" cable, used to connect the USB-A socket of the associated computer with the mini USB-B socket of the Explorer. As far as I can tell, there exist no aftermarket, perfectionist-quality versions of this digital cable; I'm keeping my fingers crossed that, if and when that day comes, the industry will keep stiffness, expense, and speculative fiction to an absolute minimum.

There do, however, exist aftermarket cables for use with the audio-out jack at the Explorer's other end: a genre in which AudioQuest has recently become a major player, owing to the use of a 3.5mm output jack on their own DragonFly DAC. For the Meridian converter, I used the same 5m length of AudioQuest Yosemite—a three-conductor interconnect with RCA plugs on one end and a 3.5mm mini-plug on the other—that I used when I reviewed the DragonFly in October 2012.

As with most contemporary USB DACs, getting the Meridian Explorer and an Apple iMac computer to play nicely with one another was as easy as losing one's health insurance. After connecting the USB cable and opening the Sound window of my G5's System Preferences menu, I found the review sample listed as "Meridian Explorer USB DAC Out"; once I'd selected it, neither the Explorer nor my iMac ever seemed to forget the other. Explorer owners who wish to use their new converter with a Windows operating system must first visit the Meridian website and download and install the appropriate driver file. (The OS specs for PCs listed on Meridian's website are "Windows XP SP3, Windows 7 SP1 or Windows 8.")

Once the Explorer is powered up, three small, white LEDs on its upper surface light up; after data streaming begins, the pattern of lights changes to inform the user of the resolution of the incoming music file: one light for 44.1 or 48kHz files, two for 88.2 or 96kHz, and three lights for 176.4 or 192kHz. This is in marked contrast to those processors on which a single light is likely to correlate with mathematically related combinations of frequencies; eg, 48 and 96kHz.

Because the Meridian Explorer weighs just slightly more than a cookie, placing it atop any sort of "isolation" accessory seemed even more ridiculous than usual. So I didn't.

Listening
A few days after my review sample of the Meridian Explorer arrived, I set about running it in. In retrospect, given how little this changed its sound, the new converter didn't particularly need it, but I nevertheless enjoyed the time I spent using it to hear my favorite Internet radio stations, during which casual listening the Explorer's tonal balance and spatial presentation were almost indistinguishable from those of the similarly priced and sized AudioQuest DragonFly. (I was helped to that early conclusion by the fact that the two devices are also very similar in apparent output voltage.)

The first serious listening I did with the Explorer was to the classic bluegrass album Appalachian Swing!, by the Kentucky Colonels, featuring Clarence and Roland White on guitar and mandolin, respectively (AIFF ripped from CD, Rounder SS31). The Meridian was instantly impressive, with a sense of scale that was pleasantly big but still appropriate to the ensemble and their setting. The original recording is a bit light, but the Meridian Explorer retrieved from it almost as much timbral color as one might hope for. The same was true with the weight and color of Roger Bush's double bass—the Explorer was clear and unambiguous in portraying the pitches of individual bass notes, down to being coldly candid about Bush's dodgy intonation.

Subtle differences were apparent between the Explorer and the DragonFly, the former having considerably better channel separation. Although not as severely "two-channel mono" as, say, those early Beatles albums, Appalachian Swing! doesn't have a lot of center fill, a characteristic made all the more plain by the Meridian. Comparisons between the Explorer and the Halide DAC HD showed the latter to be a little meatier in the timbral sense—though one could, I suppose, turn that around and describe the Explorer as "airier." That said, I did prefer the richness of the Halide—which costs almost twice as much. All three products got across the essence of the White brothers' highly charged musical interplay, yet I dare say the Meridian was the most explicit, being clearly upfront about such subtle musical—not merely sonic—details as the bass lines that guitarist Clarence sneaks in behind brother Roland's mandolin solos.

Far be it from me to tell the players on Buena Vista Social Club (AIFF ripped from CD, World Circuit/Elektra Nonesuch 79478-2) that there's an overabundance of trebly percussion instruments in the opening measures of "Amor de Loca Juventud." That said, there was something in the sound of the Explorer that brought that quality to the fore. The Meridian wasn't bright, wasn't etched, and didn't lack bass, but there was a lightness—or a responsiveness to a lightness in the music, if you will—that highlighted those high-frequency overtones. The difference between the Meridian and AudioQuest processors was exceedingly slight in this regard: Even through the DragonFly, I found those opening bars a bit too mosquitoey, but the effect was ever-so-slightly more pronounced through the Explorer.

And yet—listening through the Meridian Explorer to Lee Feldman's brilliant "Do You Want to Dance?," from his Album No.4: Trying to Put the Things Together that Never Been Together Before (AIFF ripped from CD, Bonafide UM-130-2), I again heard a sound with a more silvery, more detailed treble range than through the DragonFly. Here, however, the slight distinction definitely favored the Meridian: The British converter made clearer the descending figure in the tremoloed electric bass, revealing a greater frisson of feeling.

Speaking of bass, the Explorer proved capable of communicating low-frequency tones with appropriately generous weight and power. The Meridian allowed just the right amount of force and purr to the double bass and bass drum in "Polly Come Home," from Robert Plant and Alison Krauss's Raising Sand (AIFF from CD, Rounder 11661-9075-2). The timpani that open the third movement of Mahler's Symphony 2, performed by Gilbert Kaplan and the London Symphony Orchestra (AIFF from CD, MCA Classics MCAD 2-11011), had fine attack and a degree of timbral richness in their decay that, while not the best I've heard from a USB DAC, was satisfying. Later in the same recording, the Meridian was more than satisfying in the way it communicated the sheer, monstrous weight of the assembled instrumental forces, organ and carillon included. Heard in concert, the ending of that symphony should leave one, if not in tears, then at least slightly misty; heard in my home, the Meridian did the job—which, at the end of the day, is the most important thing I could ask of it.

A brief mention is due of the Explorer's performance with headphones: a style of listening with which I'm less than experienced. My feelings belong in the file folder labeled "Musically Satisfying and Free from Gross Distortions, Although the Sound Was A Little More Opaque Than I Expected."

Conclusions
When it comes to inheriting the Earth, or at least that portion of it that wants perfectionist-quality sound from music files stored on computers, I think The Small are doing a damn good job of things this time around. Fonder though I am of analog playback, I take heart at digital audio's recent efforts in making products that normal people desire and can afford—a trend of which there is no finer example than the Meridian Explorer. It is robustly competitive in its price range, and although bettered by the considerably more expensive Halide, the Meridian is not embarrassed by it. And the choice between it and the similarly fine AudioQuest DragonFly may, for some, come down to nothing more mysterious than aesthetics, ergonomics, and the question of whether one wants a headphone jack or not.

Also as with the US-made DragonFly, I can't deny being impressed that the people who assemble the Explorer live and work in the same country as those who stand to profit from its sale.

A remarkably good addition to a burgeoning field—and an excellent value. Very highly recommended.

The Aida is a prime member of Bel Canto's "operatic" series, which includes the Tosca and the Fidelio. John told me that the Aida was developed to provide a digital audio source that could compare with and be enjoyed alongside a high-quality analog system. In developing the Aida, the goal was to identify and correct the primary error mechanisms which afflict DACs.

A case of the jitters
As described in Stronczer's White Paper on the Aida, his first priority was to reduce the noise on the critical word clock used by the DAC. For theoretically correct conversion, this clock must be virtually free of noise or jitter—a task made exceedingly difficult by the digital interface standards (S/PDIF and AES/EBU), which specify that the critical clock be referenced to the incoming data stream. Because the data stream is susceptible to many types of corruption from the digital source through the interface electronics and interconnect cables, jitter creeps into the conversion process.

Like many commercial DACs, the Aida uses the Crystal 8412 input receiver chip, which can generate a clock with under 500 picoseconds of jitter. Unfortunately, this level of performance is inadequate for high-end performance. To improve jitter reduction, a clock regeneration method must not only deal with external jitter caused by modulation at the digital source or in the interface, but also must cope with Logic Induced Modulation (LIM)—a major, recently identified source of jitter produced by variation of the digital data representing the audio signal. These data variations modulate the reference clock through the phase-locked loop (PLL) typically used by the clock regeneration circuitry. The simple expedient of providing a low-pass filter in the PLL is inadequate to address this type of jitter. By its nature, LIM is correlated to the audio signal and can produce discrete tones at or below the system noise-floor. This is plain and simple sonic garbage that obscures low-level detail.

So how low an induced or LIM jitter level is required for high-end performance? Steve Harris at Crystal Semiconductor has shown that jitter levels as low as 200ps generate extraneous tones at or above the noise floor of a 16-bit DAC. Therefore, to push jitter artifacts a full 20dB below the noise floor would require jitter levels on the order of 20ps. Stronczer feels that LIM jitter is a potential source for the lack of "soul" in digital audio, and a major impediment to long-term listening enjoyment.

The Aida is outfitted with a proprietary dejitter circuit to provide the cleanest high-frequency clock to the Crystal 4328 DAC chip. The circuit is claimed to maintain digital signal–induced jitter to below 10ps. Separate power supplies are used to prevent modulation of the clock, and to isolate the critical clock from the rest of the digital circuitry of the processor.

The 4328 single-bit Delta-Sigma DAC was selected because of its inherent low-level linearity and circuit simplicity. Stronczer points out that this latest offering from Crystal Semiconductor has several advantages over earlier single-bit converters: the use of a fifth-order modulator to reduce quantization noise artifacts to lower than –120dB at frequencies where the ear is most sensitive; use of switched capacitor techniques instead of continuous-time op-amps for the initial filtering of the one-bit DAC output; and the inclusion of two MOSFET ICs in a single multi-chip module to optimize both the analog and digital circuitry, and to reduce radiated high-frequency noise.

The Aida's analog outputs are buffered by a class-A open-loop follower circuit using separate power supplies for each channel. This unity-gain output circuit was chosen over a stage offering gain to keep the signal path as pure as possible, and because it was felt to be unnecessary when using the processor in a typical system with an active line-level preamp. As a consequence, the nominal analog output is only 1.4V—3dB lower than the 2V CD standard.

The analog ground plane and power supplies are separated from the digital circuitry to prevent digital noise from coupling to the analog output. A broadband RF filter is also used to prevent RF noise from leaking into other components in the AC circuit through the power-supply cable.

Preliminaries
After several months of flawless performance, my sample of the Aida developed a hiccup: it sounded as if the input receiver IC was malfunctioning. I suspected that electrostatic discharge (ESD)—perniciously common in New Mexico's dry climate—had done the chip in. This was confirmed by Bel Canto Design. They found that their on-chip ESD protection circuitry is inadequate under extreme ESD conditions. Supplemental protection is now being added in the form of high-speed diodes in parallel with the input IC. The unit performed fine thereafter, but as cheap insurance against static shock, I now run a small humidifier continuously in the listening room.

At first it was difficult for me to accept the inevitable conclusion, but after several days, there was no escaping the truth: a sonic coup d'état was being consummated in my listening room. My long-term reference, the Theta Generation III DS Pre, was being unceremoniously booted out of office—the Aida had established itself as a more capable resolver of low-level detail. It was unearthing more information on recordings I was intimately familiar with, and the soundstage was more transparent, easier to step into, more convincingly fleshed-out in 3-D relief.

Hooked on detail
This business of detail resolution needs some clarification. Back in 1982, JGH was the first kid on the Santa Fe block with a CD player—a Sony CDP-101, the embodiment of Sony's prophetic promise of "perfect sound forever." We were struck by some positive attributes of God's new digital gift to humankind: bass lines sounded awesome; detail simply sparkled to the surface of the soundstage.

In hindsight, however, I can tell you that this dastardly machine transformed JGH's listening room into a sonic horror chamber: we cycled through one CD after another, and almost nothing sounded right. Bright and edgy was how musical textures were typically being translated by the Sony. Because we were intimidated by the technology, we tended to blame the software rather than the hardware; it seemed in those days that no recording engineer could get it right.

That was my first lengthy exposure to digital sound, and even now, many years after the fact, my feelings of disgust remain clear in my memory. In two words, the Sony's sound was musical vomit. Morsels of partially digested detail, especially treble transients, were expelled from its bowels with an edgy disposition, and with such a tinge of sizzle and brightness that they screamed and splattered at me. There were also textural artifacts—synthetic detail added to the natural fabric of the music.

This is a far cry from a live performance, where detail is perceived as an integral, organic ingredient of the musical tapestry. It's all in there, but it doesn't assault the senses. The sort of detail I'm after has to do with such musical nuances as the discerning of the various micro-resonances that musical instruments possess; or the ability to enjoy the delicacy of a musical chord as its harmonic envelope blooms to full glory, then decays into the noise floor of the hall; or the ability to discern the ambient signature of a recording venue; or even the various reverb settings used in a multi-track recording.

Another barometer of detail resolution is the ability of a system to delineate the individual phrasing of various instruments in an ensemble. Fuzzing over a complex passage so that individual voices are lost in the blend is the mark of a low-resolution device—the layers of an ensemble ought to be preserved in natural fashion by the reproduction chain. This is not to say that I'm always looking for that sort of detail. I may choose to focus on the whole of the music the first time around, only later shifting my attention to the substrate of the music. With the Sony, however, detail was relentlessly forced in my face—a constant barrage of exaggerated, synthetic musical minutiae.

You've come a long way, baby.
The Aida did its detail thing naturally—I was able to sift through music at my leisure, picking up detail naturally as I might at a concert hall. The treble registers were reproduced with an airy refinement, although the flavor through the top octaves was a bit on the solid-state side of reality—namely, a shade hard and grainy. Processors with a tube-based analog buffer/gain stage tend to do a better job of smoothing out and slightly softening harmonic textures. I noticed, but rarely objected to, the Aida's somewhat transistory treble presentation—probably because I consistently used tubed line stages with it.

The upper mids were gloriously pristine. My litmus test in this regard is soprano voice and violin. The sweetness and gorgeous bloom of Arturo Delmoni's violin on Music for Violin & Guitar (Sonora SACC 102) shone through unabated. And to cite one of many examples, Kathleen Battle's timbre and upper-register purity on Baroque Duet (Sony SK 46672) were undiluted.

The lower mids and upper bass projected a convincing rhythmic expressiveness, and were nicely balanced with the rest of the midrange. Whether coping with Bach, Mozart, Brubeck, or Delta blues, the Aida propelled the music forward with consummate persuasion and exquisite dynamic bloom. Bass lines were tightly defined—as long as I was careful in my choice of power amp and matching speaker cable. For some reason, the Aida did far better with Acrotec's 6N-S1040 than it did with TARA Labs' RSC Master. Check out Rob Wasserman's backing of Jennifer Warnes on Leonard Cohen's "Ballad of the Runaway Horse" (Duets, MCA MCAD-42131): if there's such a thing as a passionate double-bass, this is it.

The soundstage was projected with convincing depth and width. Transparency? The soundstage was transparent to the point of readily illuminating every inner recess. Image outlines were tightly focused in space, and stayed that way through the full dynamic roller coaster of the music. Massed voices were resolvable into their individual constituents. Commendable indeed; so many inexpensive players don't get the job done here, and smear the living daylights out of a chorus.

Final report card
In addition to engaging in hand-to-hand combat with my Theta Gen.III, the Aida did battle with CAL's Alpha processor (review forthcoming), the original PS Audio UltraLink, the Sonic Frontiers SFD-2, and, just for the hell of it, the Accuphase DP-90/DC-91 combination that I recently purchased.

Against the PS Audio UltraLink, the Aida highlighted just how far digital processors have advanced in the last few years. The UltraLink's upper registers sounded significantly grainier, and I was constantly aware of a reduction in soundstage depth. There was a forced, almost mechanical seasoning to its delivery—very transistory in its treatment of musical textures. Bass lines were exceptionally strong and forcefully delineated.

The Aida also fared well against the Sonic Frontiers SFD-2—the unit to which RH recently gave high marks, especially when used balanced. I've tried it both single-ended and balanced, and have thought it to be the finest processor I've heard (even when operated single-ended)—at least until the Accuphase came along. The Sonic Frontiers' reproduction of the treble octaves was smoother than that of the Aida, its midrange textures were more liquid, and image outlines were portrayed with greater palpability within the soundstage. However, if the SFD-2 were a perfect 10, then I'd rate the Aida a 7—not bad for a unit retailing at less than 40% the price of the SFD-2.

Then came the Accuphase DP-90 transport and DC-91 processor—Holy cow! A total of 16 hand-selected, 20-bit Burr-Brown DAC chips operated in parallel!—and my life changed forever. Well, at least my pocketbook got a lot lighter; this combination was, quite simply, head and shoulders above anything else I'd heard. The degree of detail retrieval was stunning; the multiple multi-bit–based processor was unearthing more information than I'd heard before, and doing it with elegance and grace. What clinched it for me was the following sonic episode:

I was comparing a digital master of Anyone in Love—my wife, Lesley's, latest album (Lesley & the Santa Fe Sound Machine, Vital Music VF003)—with a standard-production CD. This involved listening to the DAT master through a TASCAM DA-30—a pro DAT machine—then switching to CD playback through the Accuphase DP-90/DC-91. Imagine my shock when I realized that the production CD actually sounded better than the master itself. By better, I mean that the CD came oh, so close to capturing the essence of the live mike feed. Now, that's magic!

If I were then to rate the Accuphase DC-91 a perfect 10, the Sonic Frontiers SFD-2 would rate a 7, while the Aida would earn a 4. Still not shabby for a unit costing not much more than a tenth of the DC-91's asking price.

I suspected that the Aida's toughest competition in this price range would come from CAL's Alpha processor. Not surprisingly, the Alpha's sonic character was more tube-like, with slightly softer, more liquid harmonic textures.

Final thoughts
The Aida represents a new breed of digital processors, with access not only to the latest in DAC technology, but also to the latest technical thinking in the field. Using the latest in digital circuitry and paying close attention to clock-jitter performance, the Aida redefines what can be achieved for moderate cost. The unit served me well for many months, and I can confidently recommend it—it's a must-audition at the $1700 price point.

In a CD player?

Century-old technology embedded in a modern digital design?

I realize that Aesthetix's Saturn Romulus is not the first disc player or D/A processor with tubes, nor will it be the last—but does combining these technologies even make sense? Are audiophiles working at cross purposes to themselves, looking for modern perfection but preferring a little old-school sweetening here and there?

To date, I've stuck with purely solid-state DACs and disc players, but I was curious enough to jump at a chance to audition the Aesthetix Romulus DAC/CD player and get my ears on this oddest of technological hybrids.

The Saturn Romulus is named after one of the mythical founders of Rome (Star Trek fans will recall Romulus as the namesake planet of the Romulans), and is essentially the same product as Aesthetix's tubed Pandora DAC, with the addition of a CD disc drive and a $1000 hike in price, to $7000. For another $1000, either product can be upgraded with an optional volume control based on Aesthetix's Calypso preamplifier, rendering a preamp unnecessary in an all-digital system.

I've had in for review another CD player/DAC in the Romulus/Pandora price range, the Resolution Audio Cantata ($6000, reviewed in the November 2012 issue), and here's the weird part: after setting up everything, the soft-spoken, well-mannered designers of both products asked to hear metal/rock band Tool at prodigious levels. Any other price range, and we're mostly listening to normal audiophile fare. But something about the $6000–$7000 range and they start going all Maynard on me. It turns out that Romulus designer Jim White's wife is a retired race-car driver who was pretty good at keeping the other cars in her rear-view. Audiophiles living on the edge!

Under the hood
The DAC sections of both the Pandora and the Romulus sport a full complement of digital inputs that decode every resolution up to 24-bit/192kHz, and include: Gordon Rankin's asynchronous design for USB inputs, a Motorola DSP56362 in the filter section, and a Burr-Brown PCM 1792A DAC chip. The analog section has four tubes: my unit included a Russian Electro Harmonix 6922EH and a Slovakian Teslovac E83CCS in each channel. The analog circuit is a zero-feedback design; there are both balanced and unbalanced outputs.

The Romulus arrived carefully packed in a large, heavy, 2'-square carton. The unit itself weighs a hefty 40 lbs and is 18" square—it pretty much takes up every spot of depth in my cabinet. Its cleverly removable pop-off top—no tools needed—grants easy access to the tubes.

With the top off, and peering in from the front, you'll see the plug-in digital input cards at back left, and the beautifully laid out analog output stage on the right. Both the Pandora and Romulus feature Rel-Cap polypropylene coupling capacitors and Roederstein metal-film resistors. Jim White clearly takes time to lay out everything with precision, and the build quality is as good as I've ever seen.

At the front of the chassis, taking up almost half the interior and encased in shiny metal for extra shielding, is the hefty power supply. Also at the front, housed in its own Faraday cage to isolate it from the rest of the DAC, is a "Red Book" disc drive made by TEAC.

I normally don't go on about the innards of products I review, but under its hood, the Romulus is a thing of beauty. White has been at this awhile—he designed for Theta Digital back in the 1990s—and his experience shows.

You can order the Romulus with a silver or a black faceplate, but the surrounding case is always black, with two large open areas on top protected by metal screens. The exterior design is solid and clean, without gratuitous slabs of metal or flash, and resembles the other preamp products in Aesthetix's Saturn line.

One detail that stands out is the triangular shape of the company's logo, and of every button on the front panel. Starting from the left are the Standby button and indicator light, then the Input selector button and display on/off button. The display, in the middle of the front panel, runs through self-check messages whenever you power up the Romulus. That done, it provides track and timing information for CDs, or input and sample-rate information when running as a DAC. There are also indicators for setting CD functions like repeat and indicating Phase setting and digital lock. A nice touch is that the display senses the room's light level and adjusts itself accordingly.

Perhaps the coolest feature of the display is that it also functions as a touch volume control, assuming you've paid the extra $1000 for the volume option. The plastic display panel itself rocks slightly left and right; tap it on the left and the volume goes down; press it on the right and the volume goes up. Ingenious, though a bit tricky to figure out if you don't read the manual. Alas, my Romulus was delivered sans volume upgrade (though the display still rocks back and forth).

The CD drawer is a ½"-high slot in the middle of the front panel, below the display. One difference between the Romulus and the Pandora is that the DAC-only Pandora has a button for each input instead of a CD drawer, while the Romulus has a single button that cycles through the inputs.

To the right of the display is a Mute button with indicator LED, and to the right of that are the standard CD-transport buttons. The front panel ran only slightly warm, with the hotter components toward the rear. Still, the Romulus never got more than moderately hot around the tubes.

I was a little surprised that such a beautifully engineered product comes with a generic plastic remote control silk-screened with the Aesthetix logo and button functions. That aside, it has a button for everything, including direct selection of inputs and some extra preamp controls. I found that everything I needed was actually on the front panel of the Romulus, and so didn't use the remote much. So perhaps this really was the perfect place for Aesthetix to save a few bucks—and if you drop this remote, it won't dent your furniture or break your toe.

Around back, starting at the left, are the analog outputs, both balanced and un-, and in the middle are the AC power connector and power switch. Below those are RS-232 and trigger jacks for home-automation applications. On the right are three removable plates for the various input configurations. Mine had the first plate blank, with the USB plate next and, on the third plate, the TosLink, coax, and AES/EBU jacks. The blank plate can be replaced with another USB input.

On the USB plate is a pushbutton that switches between Class 1 and Class 2. Class 1 allows operation up to 24/96 for all operating systems without requiring any special drivers. Class 2 is USB High Speed mode for higher data rates up to 24/192, if you're using either the Windows drivers supplied by Aesthetix via their website, or a Mac with OS 10.6.4 or higher. I was in the latter camp, so I left the switch in Class 2 position.

Getting Our Ears On
After running the Romulus in the system for about a month, it was time to get down to serious listening. First things first: Whenever comparing DACs, carefully match their output levels. I often wonder, when reading a review or comment in which the sound of one DAC is described as "far and away" more striking than that of another, if the better DAC was simply a tad louder. The Romulus, though close to my system standard level set by the Benchmark DAC1 USB, was still 1dB louder—noticeable when I made close comparisons in which even such a small difference in loudness could obscure relevant details.

And so we have the Audio Research Corporation's Reference CD9 digital-to-analog converter and CD player, which, at $12,999, costs about the same as my Krell Cipher SACD/CD player ($12,000), but is so different in design and function that it offers a good test of my theory on converging sound qualities. In the May 2012 issue, I called the Cipher "a great CD player: the best I've heard in its price range, and the best I've heard, period, in my home system." How does the Reference CD9 stack up? What can each player do that the other can't, and does it matter? What does the very existence of such machines, near the peak price for a single-chassis player, say about the future of high-end audio?

Description
The Reference CD9 is a top-loading player that uses Philips's CD Pro2 transport. The drive rests on a ½"-thick metal I-beam that extends the player's full depth. Four 6H30 dual-triode tubes drive the analog section; a fifth 6H30 and a 6550C regulate the power supply. ARC claims that its use of four digital-to-analog converters (two per channel, each in dual-mono mode) reduces the digital noise floor by 3dB (ie, doubling the number of DACs doubles the signal's amplitude while keeping the noise the same). Power is handled by a custom transformer that, according to design engineer Dennis Petrich, has no gaps or breaks in the material of its R-core and thus more efficiently contains the flux field, with much less leakage, resulting in less distortion and a higher signal/noise ratio. Jitter reduction (to <10 picoseconds, according to the spec sheet) is handled by a crystal-controlled reclocking mechanism.

In its most up-to-date feature, the CD9's rear panel has not only digital outputs but also five digital inputs—S/PDIF on RCA, BNC, and TosLink, AES/EBU on XLR, and USB—for outboard digital devices or digital streaming. (The player comes with a CD-R carrying the necessary driver programs for Macs and Windows PCs.) Once you've pushed a button (on the CD9's front panel or its remote control) that activates one of these digital inputs, you can set the bit rate to 24 and the sampling rate to 44.1, 88.2, 96, or 192kHz, depending on the source signal. (You can also set it to play at the source's native resolution, whatever that may be.) The DACs are also equipped with dual oscillators, so that music at 44.1kHz (or doubled to 88.2kHz or, from there, to 176.4kHz) uses one of the oscillators, while music played at 96 or 192kHz uses the other.

Finally, when you're spinning discs, the CD9 allows you, with a push of a button, to upconvert the sampling rate from 44.1 to 88.2kHz. Another button lets you select the digital filter: Fast (a brick-wall filter at the highest frequency) or Slow (a filter with a more gradual rolloff).

For those familiar with Audio Research CD players, the Reference CD9 is the same as the CD8 (released in 2010 at a price of $9995), except for its digital inputs, the selectable filters and upsampling options, and the circuitry that facilitates them: the four DACs (vs the CD8's two), the two master oscillators (vs the CD8's one), the greater bandwidth of the analog circuitry, and the sample-rate display on the front panel.

Setup
I hooked up the Reference CD9 to a system that included the Simaudio Moon Evolution 700i integrated amp, Revel Ultima Studio2 speakers, and Nirvana cables. I usually place Black Diamond Mk.4 Racing Cones under all electronic components and plug all hi-fi gear (except amps) into a Bybee Technologies Signature Model Power Purifier. But David Gordon of Audio Research urged me not to do this, at least at the outset, so I let the CD9 stand on its own Sorbothane feet and plugged its power cord straight into the wall—in my case, into hospital-grade sockets wired to a dedicated 20-amp circuit. He was right: the player sounded a little better without the usual aids.

I played lots of CDs, occasionally pushing the buttons on the remote control to switch between straight 44.1kHz and the same datastream upsampled to 88.2kHz, and from the Fast (brickwall) to the Slow (gradual rolloff) filters. (About my findings on these matters, more later.) ARC recommends 600 hours of break-in; I gave the review sample three weeks of continuous play before settling down to serious listening. I conducted several A/B comparisons with the Krell Cipher. To check out the CD9's digital inputs, David Chesky, of Chesky Records and HDtracks, brought over a MacBook Pro and an external hard drive loaded with Audirvana software and lots of high-resolution tunes from the Chesky catalog (mainly in 24/192, but a few in 24/96), which I compared with CDs and, when possible, LPs of the same titles.

Sound
As a CD player, the Reference CD9 was simply excellent. This wasn't a surprise, given Audio Research's track record. What was a surprise—and what takes me back to the proposition at the top of the review—is that it sounded almost exactly like the Krell Cipher, even though the Cipher has a front-loading transport (it's not a top-loader), transistors (not tubes), two DACs (not four), and proprietary circuitry that manipulates audio signals in the current domain (not the voltage domain). Often, two similarly high-priced components may both "sound good" but have different strengths and weaknesses, which in many cases stem from their designers' different preferences or trade-offs. A review that compares two such models generally identifies those trade-offs, weighs the differences, and concludes which player might be better or worse for various kinds of music or taste. Usually, the differences are pretty clear.

That Electrocompaniet amplifier was impossible to find in the UK, where I then lived, and in the absence of actual experience, the reputation of its sound quality grew to mythic proportion. But when I did finally hear the Electrocompaniet amplifier at an audio show, I was not disappointed.

Which makes it all the more peculiar that, in more than 37 years of working at audio magazines, I have never reviewed an Electrocompaniet product. With this review of the company's ECD 2 digital/analog processor, which costs a dollar short of $3100, that streak of inattention has come to an end.

The ECD 2 . . .
. . . is a wider-than-usual (18.3") component, with a black-painted steel chassis and a thick acrylic front panel. Into the latter are set a blue fluorescent display on the left, a gold on/off button in the center, and, on the right, four gold buttons grouped in a diamond pattern and labeled Navigation. Of these four, the top and bottom buttons are for volume up and down (from "0" to "100" in 1dB steps), the middle two for selecting the input. The buttons are duplicated on the remote control, which also has a Standby button. When you select a source with valid data, the display shows the sample rate on its left and the volume on its right for about five seconds, after which it reverts to showing the input name.

On the left of the rear panel are the balanced and single-ended analog output jacks; and at its center, the five digital inputs: USB 2.0, and two each coaxial and optical S/PDIF. On the right are the AC inlet, trigger input and output jacks, and an RS232 port. The 18-lb ECD 2 stands on three compliant feet. Its USB input doesn't need a driver program when the ECD 2 is used with Macintosh computers; a Windows driver, and manuals for all current Electrocompaniet products, are provided on a CD-R.

Internals
The ECD 2's digital and analog circuitry are carried on a large, blue printed circuit board mounted behind the rear panel and running almost the entire width of the chassis. This board is stuffed with surface-mount parts. A large XMOS chip handles the USB input, and the digital data are fed to a Burr-Brown SRC43921, the same asynchronous upsampler chip used in the Bryston BDA-2 and Musical Fidelity V-DACII. This feeds two Cirrus 4398 multi-bit, delta-sigma DAC chips. Although the 4398 is a two-channel part, the ECD 2 has one per channel, presumably used in dual-differential mode to give a 3dB increase in dynamic range. This chip has an integral volume control, and a separate port for DSD data; perhaps an upgrade will be offered to enable the ECD 2 to decode DSD datastreams.

The analog circuitry appears to be based on discrete transistors, with local regulation for the power-supply rails and much decoupling in evidence. The power supply is based on a toroidal transformer mounted behind the front panel.

Sonics
Although its volume control makes it possible for the ECD 2 to be used straight into the power amplifier, I stuck with the traditional arrangement of feeding it to a preamplifier, in this case the Pass Labs XP-30 that I reviewed in March. I left the Electrocompaniet's volume control at its maximum setting of "100." Although I occasionally used the Marantz NA-11S1 media server's or Ayre Acoustics C-5xe^MP disc player's coaxial outputs to send audio data to the Electrocompaniet, I mostly used the ECD 2's asynchronous USB input with Pure Music running on my Mac mini.

The ECD 2 is specified as handling PCM files with sample rates of up to 192kHz. As the Electrocompaniet has a DAC chip capable of decoding DSD data, I tried playing some DSD files using Pure Music. However, while these files played, it looked as if Pure Music was converting them on the fly to 176.4kHz PCM to make them compatible with the ECD 2.

I've always been a bit suspicious of upsampling converters, concerned that the benefit of driving the DAC chip with higher-sample-rate data would be offset by the possibility of mathematical imprecision in the upsampling process. But the ECD 2 allayed any such fears I might have had. In Trevor Pinnock and the Royal Academy of Music Soloists Ensemble's excellent new recording of Erwin Stein's chamber-orchestra version of Mahler's Symphony 4 (24/192 ALAC files from Linn CKD438), the acoustic around the piano in the opening movements was deliciously apparent, while the images of the solo oboe, clarinet, violin, and cello were presented more forward in the soundstage, and were well defined and stable. The single double bass that carries the responsibility of providing the work's tonal foundation was weighty yet well defined. In the first part of the symphony's third movement, Ruhevoll, poco adagio—where, after several solo instruments serially intone a forlorn melody over a pizzicato bass riff, the harmony resolves on a low bass note—the ECD 2 made sure it was goose-bump time.

Illusonic Immersive Audio Processor
Beginning with its mouthful of a name, the Illusonic Immersive Audio Processor, from Switzerland, is a high-end preamplifier-processor whose developers have a unique idea of what such a device should do: First, at a price of 19,200 Swiss francs, or about $21,300, it should be well constructed, with sound quality and transparency that are beyond significant criticism. Second, it should eschew features made redundant by their inclusion in other system components. Third, it must include features that are either missing from or incompletely implemented in competitive products. (Illusonic has no US distributor; the IAP is available directly from the manufacturer.)

When I first learned of the IAP, I dismissed it because it lacks features, such as dts-HD and Dolby TrueHD decoding, that these days seem de rigueur. In fact, the IAP accepts only PCM or analog signals—it can't even decode plain-vanilla Dolby Digital. The manual states "Up to 8-channel/192kHz/24 bit audio over HDMI is supported. Most Blu-ray players will decode Dolby Digital, Dolby True HD, DTS, and DTS HD and deliver the decoded multi-channel signal to the Immersive Audio Processor." And, indeed, most Blu-ray and universal players will.

On the other hand, the IAP has features that most pre-pros don't. The inclusion of a five-band parametric equalizer in each of its 16 output channels is rare and welcome. Truly unusual is the IAP's ability to process, via those 16 output channels, two- and multichannel sources into one of 168 loudspeaker setups with the ability to manipulate the presentation's spatial distribution using three parameters: Center, Depth/3D, and Immersion. Yes, my Meridian Reference 861 v8 pre-pro will do something similar, but with fewer channels, and not as easily as is possible with the IAP's real-time utilities for Windows and Mac computers. Finally, Illusonic touts the Swiss-made IAP as delivering "high-quality sound, otherwise only found in 2-channel high-end audio equipment," without including unnecessary parts and features. All of this made the IAP increasingly enticing to me. I asked for a review sample.

The IAP's front panel is disarmingly simple. To the right is the volume-control knob, which can be pushed to select among other options reached by using the navigation buttons (Back, Menu, Forward), which sit just below the eminently clear, three-line, alphanumeric display. A small Apple remote duplicates these controls, and a standby LED and an IR receiver complete the panel. While these controls work reliably and exactly as described in the owner's manual, the nesting of menus and the range of controls demand tightly focused attention lest one lose one's place. Illusonic comes to the rescue with a superb control app that, via a USB connection to a computer, gives the user a nice graphic interface to all of the IAP's functions. This is in lieu of a Web-based control app, and avoids the unpredictable response latencies of such apps. The IAP's app works in real time—you can immediately hear the effects of your manipulations.

A look at the rear panel suggests the range of the IAP's processing. There are four HDMI inputs and one HDMI output, three coax and two optical digital inputs, three analog stereo RCA inputs (configurable into a 5.1- or7.1-channel input), and an XLR analog stereo input. Across the bottom of the panel are 16 (count 'em) 16 XLR output connectors labeled, from left to right: SUB, RRCH/FRWH, RLCH/FLWH, RRH, RLH, FRH, FLH, FRW, FLW, RRC/RCH, RLC/RC, RR, RL, FC, FR, and FL.

The XLRs whose labels include slashes can provide different outputs depending on the IAP's setup, but overall, the 16 outputs permit the common arrangements or 2, 3, 4, 5, 7, or 9 channels, with options for subwoofer, height and top (ceiling) speakers, up to a full 9.1+4H+T or 9.1+6H array. With all of these options plus bass management, channel trims, and EQ, it is appropriate that the IAP can store three (soon to be five) preset configurations, as well as Bypass.

With such a range of options, it's likely that every user will seek a different path through them, and unlikely that any will try every possible option. I was limited to 2.1-, 3.1-, and 5.1-channel arrays, supplemented by another pair of channels used as FLH/FRH (front height) or FLWH/FRWH (front width/height) , or a 6.1 setup with a single RC (rear center). Also, had this been a permanent installation, I would have tried harder to optimize very useful bass management. In the interests of time and sanity, I reverted to full-range operation for all main channels. I also used the IAP's parametric EQ for room correction, supported by RoomEQ Wizard (REW) and OmniMic as measurement and filter synthesis tools.

First, I found the straight-through sound quality of the IAP remarkably clean, transparent, and balanced, which qualified it as competitive in its exalted price class. With stereo or multichannel signals, the clarity and soundstaging were open and reminiscent of a good analog preamp, though I still detected a bit of blurring of detail with SACDs and 24-bit/192kHz PCM, as compared to running the analog output of the SACD player through my Audio Research MP1 multichannel preamp. Although the "hardware is principally designed to be able to operate up to 192kHz," per Illusonic, and the IAP can accept signals up to 192kHz via HDMI and 96kHz via S/PDIF, the IAP's internal signal path is 24/48. Nonetheless, I did not find this a limitation except on paper, and would not have suspected the signal path was limited to a 48kHz sample rate had I not been told.

My experiments with adding height and/or width channels were thrilling—the IAP's use of DSP to derive the additional channels and the dependent modification of the main channels was judicious in expanding the soundstage without bloat or loss of spatial resolution. The implementation of the IAP's Center, Depth/3D, and Immersion controls was essential to avoiding those common foibles. For example, the IAP removes front center (phantom center) signals from the FR/FL when a center speaker is present, but the Center control let me optimize the degree of transformation. I could sit at the primary listening position, laptop on lap, and listen while tweaking the processor to my preference, all in real time.

I particularly enjoyed the effect of adding the width and height channels with choral recordings: they expanded the array of voices without losing image specificity. The front high and front wide channels were less effective in my room with FR/C/FL speakers on the short wall, but the addition of a rear center channel really nailed discrete and ambient sounds behind me, and eliminated any ambiguity in those rare source signals that include discrete center-rear information. That said, I have yet to be sufficiently won over by these enhancements (and they really are enhancements) to want to permanently reconfigure my music-only room/system.

Finally, when used with a calibrating microphone (not provided), the IAP can automatically measure and correct speaker levels and distances from the listening seat. I didn't try this because: 1) I already knew those values, and could just plug them into the IAP's set-up; and 2) I was already committed to using RoomEQ Wizard or OmniMic for EQ measurements, which can also provide/confirm these values. Both of those systems are fundamentally mono, but the IAP's ability, in measurement mode, to route the FL input to any output channel made multichannel measurement operations a piece of cake. Run the sweeps, average as needed, calculate and tweak filters, and then enter into the IAP the frequency, gain, and Q values. Further tweaking always seems necessary, and the IAP made it easy to reach a satisfying result by using real-time monitoring and A/B comparisons of multiple presets.

Illusonic's Immersive Audio Processor is not like other processors. Its unique feature set will appeal to sophisticated users who know exactly what they want from their systems: the ability to implement flexible and complex speaker arrangements, a graphic user interface with real-time configuration operations, an effective parametric equalizer, and outstanding overall sound quality. The quality of support has been excellent, with two flawless firmware upgrades during the brief time I've had the IAP; promised enhancements include support for two independent subwoofers, expanded preset functions, virtual channel remapping, and Mac/Windows software for EQ measurements and settings. All of this should appeal to the hands-on audiophile, but only if he or she is willing to forgo high-definition codecs, multiple remote zones, and other bells and whistles commonly found in AVRs. If that's you, the IAP is the way to go.

miniDSP 10x10Hd Digital Signal Processor
For some years, miniDSP Ltd., based in Hong Kong, has provided techie audiophiles with interesting digital-signal-processing (DSP) modules and input/output devices to support audio applications. Their initial products were configured as crossovers, subwoofer interfaces, and equalizers, and were accompanied by analog, S/PDIF, I²S, and USB interfaces. However, based as they are on DSP, as the company's name implies, these devices are fundamentally multipurpose, their applications determined by the chosen interfaces and firmware, which miniDSP calls plug-ins.

These products have fascinated me, stirring up long-buried DIY urges from my past. Think of the DSP capabilities incorporated into all modern audio products—if made discrete, as in miniDSP's components, they could be under your direct control. Megalomania aside, I felt compelled to see whether such things could be used by a nominally normal audiophile who has long forgotten machine code, high-level programming languages, and all things related to programming.

The Scarlattis improved on the Elgar Pluses in cosmetics and, more important, technical and sonic performance that, in the opinion of many, put it at the top of the digital audio heap. It will be obvious to those familiar with the Scarlatti stack that the Vivaldis take visual appearance and fit'n'finish to new heights. These sculpted boxes in a silver matte finish, each with a unique, flowing wave pattern machined into its front panel, are pleasing to the eye and silky-smooth to the touch. Who said engineering geeks aren't sensual?

The four components are: the Vivaldi DAC ($34,999, slightly more than the original Elgar-Purcell-Verdi combo), which, using can decode every digital resolution, from MP3 to DSD to DXD (a 24-bit/358.8kHz PCM format used primarily for editing DSD files) and everything in between; the Vivaldi Upsampler ($19,999), which can upconvert the lowest-resolution MP3 data to 24/384, DSD, and DXD or any format in between; the Vivaldi Master Clock ($13,499), containing two groups of four clock outputs, which can be independently set; and the Vivaldi Transport ($39,999), a smooth, quiet, quick-booting SACD/CD drive based on TEAC's Esoteric VRDS Neo disc mechanism, considered by many to be the world's finest (footnote 1). This is controlled by dCS-designed signal-processing electronics, and can also upsample CDs to DSD or DXD.

The total price is an astronomical $108,496, though of course you could start with just the DAC and save $73,497. Considering the cost of some preamplifiers, audiophiles with only a single source component can drive their power amplifier directly with the Vivaldi DAC, which has a digital-domain volume control, and save the cost of a standalone preamp.

Most audiophiles prepared to drop $35k on a DAC will head straight for the full stack, I feel, especially those who want to add a music server, since the Upsampler includes an Ethernet connection for network streaming via UPnP. (dCS makes available a free iPad app to work with UPnP-based servers.) While both the DAC and the Upsampler have an asynchronous USB-B port, to allow them to accept audio data from a computer, the Upsampler also has a USB-A port, for use with USB flash drives or i-Devices. The USB-B ports can operate in Type 1 mode, with data sampled up to 96kHz, or in Type 2 mode, which allows DSD and PCM up to 192kHz to be streamed to the Vivaldis. Type 1 operation doesn't require a driver program for Windows PCs or Macs; Type 2 operation doesn't need a driver with Mac OS10.6.3 or later; dCS supplies a driver program for Type2 operation with Windows XP and Windows 7. (They say that it will work with Windows 8 in Windows 7 compatibility mode.) The dCS driver is not compatible with ASIO-type drivers, which will need to be uninstalled.

In other words, while continuing to offer a transport option for those with large collections of CDs and SACDs, dCS has configured the Vivaldi for the discless future—already the present for some, particularly those who never got into SACDs—though it will still be some time before downloadable DSD content catches up with what's already available on disc.

Major Changes Inside
Compared with the older products, the structural rigidity of the Vivaldi components' cases has been upgraded, including top plates of thick aluminum into which are machined asymmetrical cavities that contain damping pads. The combination is said to reduce vibrational modes. But the Vivaldi DAC's insides have been given far more than a cosmetic makeover.

While dCS's products have all been based on the company's proprietary Ring DAC technology, the Vivaldi is based on a complete revision of the Ring DAC concept. The earlier Ring DAC used quad latches (a circuit element that can be instantaneously "flipped" between two stable states) to select current sources based on metal-film resistors. The new Ring DAC design still includes high-speed latches and metal-film resistors, but instead uses individual latch chips said to eliminate between-latch, on-chip crosstalk resulting in lower jitter. The total number of latches has been increased to make better use of the Ring DAC's available dynamic range. A pair of high-speed, software-updatable (FPGAs) replaces the Scarlatti's mapping ROM chips, which allows individual errors in the DAC's current sources to be randomized, reducing the level of distortion and spuriae by 3dB.

The Vivaldi DAC's and Upsampler's digital-processing platform, which runs dCS-developed and -maintained code that forms the core of the entire operating system, has been completely upgraded, including use of a single field-programmable gate array (FPGA) chip with more than twice the capacity of the previous two chips combined. These FPGAs are programmed from flash memory each time the DAC powers up, so that performance improvements or new Ring DAC operating modes can be added with a firmware update.

The Vivaldi DAC's analog circuitry has been redesigned for reduced DC offset, lower noise, and less crosstalk. The completely dual-mono architecture is claimed to improve left/right crosstalk by 15dB at 20kHz. New, higher-output mains transformers in all of the Vivaldi components run cooler than in earlier dCS products, and are mounted on specially damped subchassis to eliminate vibrations. Even the transport has been given a serious mechanical upgrade, which is claimed to reduce the level of acoustic noise by a significant 10dB.

The Vivaldi stack's exterior elegance sacrifices some user-friendliness: the small buttons and their even smaller labels aren't backlit, so if you like doing your digital business in the dark, or even in moderate light, keep a flashlight handy.

The non-backlit remote control suffers a similar problem. While it's a heavy, nicely made piece of aluminum finished pleasingly for the hand, its brushed-aluminum surface makes it difficult to read the labels just below the unmarked buttons unless it's angled just so to reflect any available light. But when you do that, the reflected glare makes the buttons that are marked hard to read.

Setup and Use
Setting up the Vivaldi system can be complicated, but when you spend almost $110,000 on a digital playback system, you can expect plenty of dealer help. The multi-cable system, featuring five AES/EBU cables—two each to connect the Transport and Upsampler to the DAC and a fifth to connect the Transport to the Upsampler—five word-clock connections, and four AC cords, is a cable manufacturer's dream.

But once everything has been set up for you, and the routing for each input and the many screen icons have been explained, you're on your own. Believe me, until you figure it all out and memorize the icons' meanings, you'll feel lonely.

Your options are then seemingly endless. For example, you can route the transport directly to the DAC via the dual AES/EBU connections, which will send DSD signals in their native format and, should you choose, upsample CDs to DSD or DXD—or you can send CD signals to the DAC via S/PDIF or AES/EBU and play them in their native resolution—or send CD data to the Upsampler via the single AES/EBU connection and, by selecting the Upsampler's AES input, play them back at 24/44.1 (the Upsampler automatically outputs at 24 bits)— or you can upsample the CD data there to high-resolution PCM or DSD.

The Vivaldi DAC has six filter options for PCM, with four more for DSD playback. dCS says that with PCM data, the first four filters give different tradeoffs between ultrasonic image rejection and impulse response. Filter 1 gives the most rejection, Filters 2–4 offer progressively relaxed image rejection and better time-domain performance. For data with sample rates from 176.4kHz to 384kHz, two extra filters are available, one with a Gaussian character, the other asymmetrical with almost no pre-ringing. With CD data, Filter 5 is also an asymmetrical type, Filter 6 a very sharp linear-phase type with pre-ringing. The DSD filters progressively roll off the format's ultrasonic noise, in case there are system synergy issues.

dCS recommends Filter 5for CD playback, Filter 2 for 48, 88.2, and 96kHz-sampled data, Filter 6 for 176.4kHz data and above, and Filter 1 for DSD data. For this review, I stuck with Filter 1, which had been recommended by dCS distributor John Quick.

In 2007, I spent time with Bel Canto Design's e.One DAC3 D/A processor. In his review of the DAC3 in the November 2007 issue, John Atkinson quoted my comparison of it with the Benchmark DAC1, which I called "the Swiss army knife of audio" and "one of the only future-proof source components you can buy these days."

Sure enough, in 2013, the standalone DAC-preamp has become an integral part of the audiophile world, and no company has led the way to the high-quality DAC/preamp/headphone amp as confidently as has Benchmark Media. I bought my original Benchmark DAC1 ($995) because I was an up-and-coming musician who didn't have a ton of cash but who knew what high-quality audio was supposed to sound like. I produced recording sessions, and needed something that could play high-resolution files through my headphones. I also wanted a component that could play data from a silver disc and work with my computer—which I saw as the future of the digital front end.

Keeping up with the times and the needs of the digital marketplace, Benchmark has steadily improved and added to its original model, the DAC1. They released the DAC1 PRE, which added an analog preamp, then the DAC1 USB and the DAC1 HDR, which improved the volume control and used higher-quality op-amps. I found the DAC1 HDR to offer slightly better, smoother sound than my DAC1. So when I had the chance to listen to Benchmark's new DAC2 HGC, I was very curious to hear where it fit in Benchmark's impressive lineage.

Whatcha got?
The Benchmark DAC2 HGC ($1995) has a number of features not included in its older siblings. First, the front panel boasts sample-rate and word-length displays. These are invaluable tools, especially when your digital front end is a computer and you want to know if you've configured your audio settings correctly to play hi-rez files. While I can usually hear when a hi-rez file is being incorrectly truncated or decimated, checking the LEDs on the DAC2's faceplate reassured me that I was indeed hearing what I was supposed to be hearing.

The front panel also has buttons for Power, Dim/Mute, Input, and Polarity, a motorized Alps volume pot, and two ½" headphone jacks. Except for the last, all of these functions can also be handled via the included remote control, which I found elegantly simple in design, layout, and function.

The DAC2 HGC has five digital inputs: USB, two optical, and two RCA coaxial. The USB input works in asynchronous mode and will operate in either USB 1.1 or USB 2.0 modes. If you play only 24-bit/96kHz files via USB, no extra drivers are needed. To play DSD or 192kHz files, a downloadable driver is provided for Windows computers, and the DAC2 must be set to USB 2.0 mode. While the DAC2 HGC has a lot of inputs, I miss the DAC1's balanced AES/EBU input and its BNC connectors for the S/PDIF inputs, all of which I used. The DAC2 has two pairs of analog inputs, whose signals remain wholly in the analog domain from input to output (more on that later). These lead to three analog outputs, one balanced and two single-ended. And one of the coaxial digital inputs can be configured to act as a digital pass-through.

The DAC2 HGC offers two new features that should improve sound quality over Benchmark's older DACs. First, it converts digital signals with four 32-bit ESS Sabre DACs run in balanced configuration. Benchmark claims that the DAC2 HGC is a full 10dB quieter than the DAC1. The DAC2's digital processing is also claimed to have 3.5dB of digital headroom when fed a signal of 0dBFS. With today's standard practice of overloud mastering, this much headroom should ensure that no digital clipping occurs, and that the filter runs more linearly when fed high-level inputs.

Also new in the DAC2 is Benchmark's Hybrid Gain Control (HGC), for volume attenuation. The HGC grew out of Benchmark's experience with their HDR volume control included in the DAC1 HDR. As I understand it, the DAC2's volume control combines active analog gain control and passive low-impedance attenuators in the analog realm with a 32-bit digital DSP gain control for digital signals. Unlike many DAC-preamps, the DAC2 HGC keeps all analog inputs in the analog domain, instead of relying on digital conversion at both input and output. Analog purists, rejoice!

Like the Benchmark DACs before it, the DAC2 HGC is half a rack unit wide and comes in a black or silver case. Optional rack mounts are available. The layout is very clear and organized, the styling professionally oriented. The fit and finish were great, and the feel of the volume control was solid and silky—but the DAC2 lacks any pretention of being audio jewelry.

Set-Up
I began listening to the DAC2 HGC using my Sennheiser HD600 headphones and Hewlett-Packard EliteBook 8570p laptop computer. The HP has an Intel i7 chip, runs at 2.90GHz with 8GB of RAM, and runs Windows 7 Professional (64-bit). For all of my critical listening I used JRiver's Media Center 18, which I've found offers the best sound from my computer. The DAC2 was connected to the HP via DH Labs' USB cable, which sounds (and looks) lovely.

The DAC2 HGC doesn't use a separate audio driver to interface with the computer—it really is plug-and-play in USB1.1 mode. [Benchmark does supply a driver package, including ASIO support, for operation in USB2.0 mode.—Ed.]. However, when using a program like Media Center, having a dedicated ASIO driver comes in handy when I try to bypass as much of my laptop's audio circuits as possible. After going through Benchmark's checklist in the comprehensive manual of how to create the best settings for using the DAC2 with Windows 7, I was still unable to entirely bypass my laptop's volume control. Benchmark suggests that users who set their computer's volume level to "100" should get bit-transparent data, but I'm always leery when I can't bypass all of my computer's audio functions. By contrast, when I use the $795 Centrance DACmini D/A converter, which I reviewed in December 2011, I have the option to either plug-and-play or use their optional ASIO driver in Media Player. When I select ASIO, I can easily get my computer's audio functions out of the signal path. With the Benchmark, I was never quite sure I had. However, I was sure of the bit depth and sample rate I was sending to the DAC2: I could read it on the Benchmark's LED display.

Headphone Sound Quality
A hi-rez download of Daft Punk's Random Access Memories (24/88.2 FLAC files, Columbia/HDtracks) provided fantastic listening via my Sennheiser HD600s. The tasty bass playing on the robot's "Give Life Back to Music" was full, controlled, solid, and driven, and Nile Rodgers's classic Stratocaster rhythm-guitar work really shone through the DAC2. What would normally be merely a pop song's background rhythm guitar was brought to the fore by the DAC2, allowing me to appreciate the subtle voicings and strumming patterns Rodgers chose for this groove. Hearing this track through the DAC2 made me appreciate why Nile Rodgers is the man.

Part of what made it easy to pick out Rodgers's rhythm-guitar work was the DAC2's ability to offer wonderful image size and separation via headphones, while still giving each instrument the proper solidity and scale in the mix. I think that partly comes from a low level of self-noise, which the DAC2 certainly seemed to have. But from treble to bass, the DAC2's sound through headphones was also very dynamic and even, highlighting nothing but missing nothing.

That sounded like a DAC I needed to review, but it wasn't until July that AURALiC co-founder Xuanqian Wang hand-delivered a review sample of the VEGA. And it wasn't until October 2013 that I was able to embark on this review.

AURALiC . . .
. . . was founded in Beijing, in 2008, by Xuanqian Wang and Yuan Wang. Xuanqian Wang trained as a professional engineer in both electronics and recording engineering, and also started playing piano at age four, while Yuan Wang studied sociology and management science in the US, before returning to China to start a company that manufactured precision instruments. The two met at a music festival in Berlin; their shared love of music and the "relentless pursuit of superior sound quality" inspired them to design and manufacture audiophile products. (From here on, I refer to the company and its product as Auralic and Vega—Stereophile's style is to reserve the all-caps treatment for actual acronyms and initialisms, and the capitalization and inverted camel-capping usage that Auralic has adopted looks too much like shouting in print.)

The Vega The elegant-looking Vega ($3499) is housed in a slim, brushed-aluminum enclosure. The front panel is dominated by a wide, rectangular, yellow-on-black OLED display, to its right a domed knob and a red LED. The rear panel offers single-ended and balanced outputs, respectively on RCAs and XLRs, and five digital inputs: transformer-coupled AES/EBU on an XLR (the default), two transformer-coupled coaxial S/PDIF on RCAs, one optical S/PDIF on TosLink, and a high-speed USB2.0 port. The AES/EBU and S/PDIF inputs handle 16- and 24-bit data with sample rates up to 192kHz; the USB port also operates with sample rates of 352.8 and 384kHz, and will accept DSD64 (2.8224MHz) and DSD128 (5.6448MHz) data using the DoP v1.1 protocol. No driver is required for correct operation with Mac OS10.6.4 onward; Windows users need to install the supplied driver (detailed instructions are included in the excellent manual).

While there is an on/off switch next to the IEC AC receptacle on the rear panel, the Vega is disturbed from its Sleep or Standby modes by pushing the front-panel knob. A second push brings up a menu, permitting selection of Input, Balance, Phase (polarity) Filter Mode (see later), and System parameters; otherwise, the knob controls the output level, allowing the Vega to be used as a digital-input control preamplifier. Pressing the knob to bring up the System submenu allows the user to set the Display brightness, Sleep mode enable/disable, internal clock mode (see later), and Volume mode: each input can be set to default to a different volume level, or all can be set to default to the same value.

The Vega can also be operated by remote control, a plastic controller being supplied as standard.

Technology
Inside, the Vega's digital and audio circuits are carried on a large printed circuit board that occupies the full depth of the chassis and most of its width. The toroidal power transformer sits behind the front panel and in front of a yellow shielded section that carries the AC input and filtering. A small daughterboard behind the USB jack is marked "AURALiC DSD over USB" and carries an XMOS USB receiver chip. Three surface-mount LSI chips live behind this board, the largest of which is marked "AURALiC Sanctuary Audio Processor powered by Archwave." Archwave AG is a Swiss company; their multi-core, ARM 9–based Sanctuary processor runs at 500Mips and is used in the Vega to upsample PCM input data to approximately 1.5MHz and 32-bit depth, to provide what Auralic calls the ActiveUSB buffer stage, and to implement four reconstruction filters for PCM data and two choices of low-pass filter for DSD data. The filter options are referred to by Auralic as Flexible Filter Mode—the PCM filters include linear-phase and minimum-phase options, as well as two slow-rolloff types; the two DSD options offer different degrees of ultrasonic rolloff, to prevent the format's noiseshaping from contaminating the downstream amplification.

The Vega uses a high-precision master-clock circuit that the company calls the Femto Master Clock. Covered by a hefty heatsink, this uses what is claimed to be an "aerospace grade," temperature-compensated crystal oscillator with an "ultra low noise" linear power supply. The jitter is specified as an extraordinarily low 82 femtoseconds, with phase noise at –168dBc/Hz. The user can choose between four clock settings: Auto (the default), in which the Vega uses the optimal clock window for any source; Coarse, which offers the widest bandwidth of input lock, to allow the Vega to work with very jittery sources; Fine, which narrows the lock acceptance window to give the lowest jitter with high-quality streams; and Exact, which will only give lock with only very low-jitter streams but gives the highest sound quality.

Once processed by the Sanctuary chip, the oversampled 32-bit data are fed to an ESS Sabre³² 9018 D/A converter chip. This is a premium-quality, 32-bit, delta-sigma part with eight individual DAC sections, these operated, I believe, in two sets of four in push-pull parallel to get the lowest possible noise and the highest linearity, even at very low signal levels. Though this chip offers upsampling, this can be bypassed, in which case it will accept data of sample rates up to 1.536MHz, as in the Vega. It offers volume control for both DSD and PCM data. This chip is covered by a plate labeled "AURALiC DSD Direct Stream Digital DXD Digital eXtreme Definition." (DXD is the 24-bit/352.8kHz PCM format introduced by the Swiss company Merging Technologies for its Pyramix DSD workstation some years ago, to allow editing of DSD data without losing resolution.)

Texas Instruments' high-performance SoundPlus OPA1612 dual–op-amp chips are used for the Vega's I/V and analog low-pass filter stages. The balanced analog output stage, based on what Auralic calls the Orfeo module, is said to be "inspired by the Neve 8078 analog console's circuit design." (Orfeo is not used for the single-ended outputs.) The components in this module are claimed to operate in thermal equilibrium with very low open-loop distortion, the output transistors in class-A. Auralic says that the Vega's output stage will have no problem driving loads as low as 600 ohms from both its balanced and unbalanced outputs, which my measurements confirm (see sidebar).

Sound Quality
Xuanqian Wang told me that he doesn't believe in burn-in. However, I found that the Vega took several hours from cold before its sound quality reached a plateau. This is apparently because both the Femto Master Clock and Orfeo output-stage components need at least an hour to establish thermal equilibrium before reaching their specified performance conditions. When I put the review sample in Sleep mode, the clock and output stage remain powered, eliminating the need for any further warm-up when the Vega is switched back into operational mode.

Prior to the Vega's arrival, my listening room saw some superb-sounding D/A processors in 2013. In order of rising price: the NAD M51 ($2000, reviewed in July 2012), the Musical Fidelity M6DAC ($2999, June 2013), the Electrocompaniet ECD2 ($3100, December 2013), the Arcam FMJ D33 ($3200, February 2013), the Marantz NA-11S1 ($3499, October 2013), the MSB Diamond DAC IV with Diamond Power Base ($43,325, October 2012), and the dCS Vivaldi system ($68,497 without its SACD transport, January 2014). But the $3499 Vega was in no way embarrassed by having to follow this company. In fact, though I felt a twinge of loss when the Vivaldi system went back to the distributor, the Auralic Vega proved a very satisfying replacement.

The English composer John Tavener passed away as I was installing the Vega in my system. One of the first recordings I played, therefore, was cellist Raphael Wallfisch's performance of Tavener's The Protecting Veil, accompanied by the Royal Philharmonic Orchestra conducted by Justin Brown (CD, Intersound 2847), feeding data to the Vega's AES/EBU port from my Ayre Acoustics C-5xe^MP universal player. Like Michael Lavorgna, I found that Mode 4 was overall my favorite filter. It allowed the spaces between the notes on this hauntingly beautiful recording to be fully developed.

I was going to follow the Tavener with a needle drop of Joni Mitchell's 1978 album, Don Juan's Reckless Daughter (LP, Reprise K63003)—but I couldn't wait for the transfer to digital, which of course can be done only in real time. So I fed the AES/EBU output of my Ayre QA-9 A/D converter, set to 24/96, straight to the Auralic Vega. OMG! The low F and C that Jaco Pastorius strikes from his Jazz Bass at the start of "Cotton Avenue" were projected with almighty weight by the Vega, but without losing any of the definition to the notes' leading edges. Similarly, when Pastorius swoops down to the open D string at the start of "Jericho," which ends side 1 of this album, the combination of low-frequency weight and higher-frequency definition made it difficult to remember that I was listening to back-to-back A/D and D/A converters in the playback chain. I knew, from my review in November 2012, that the Ayre has superb sound quality, but it was obvious that the Auralic Vega was equaling that quality.

The front panel offers, from left to right: a Mode toggle to switch between internal and external clocks; a second toggle to select the sample rate (44.1, 48, 88.2, 96, 176.4, 192kHz) for the Main and Auxiliary outputs and to select ADAT or AES/EBU data formats; a third toggle to select the meter range (1dB or 6dB/division) and Peak Hold on/off; and a 2x9-LED level meter. Continuing to the right is a group of controls for the first channel, toggle switches to select gain (0, 10, or 20dB), Variable or Calibrated gain, and a rotary level knob; this group is then duplicated for the other channel.

The rear panel offers, from left to right: a pair of balanced analog inputs on XLRs; two unbalanced AES/EBU Aux outputs on BNC jacks; a USB Type-B port; a Main TosLink output working in either S/PDIF or ADAT format; a Main balanced AES/EBU output on an XLR; word clock input and output on BNCs; and the IEC AC inlet. The word clock ports allow the ADC1 to be used in multichannel applications, slaved to other converters. The Main USB, TosLink, and AES/EBU outputs always offer 24-bit data. The Aux outputs can have a different sample rate from the main outputs, to feed a CD-R or DAT recorder, and can be set to offer TPDF-dithered 16-bit data as well as the original 24-bit data. (TPDF stands for Triangular Probability Density Function, and refers to the spectrum of the dither used.)

Inside, the circuit layout is as clean and as logical as I have come to expect from Benchmark. The analog circuitry appears to be based on OP27, AD797, and LME49860 high-performance op-amps, while the A/D converter chip is an AKM 5394. This is a 24-bit, two-channel, 128x-oversampling, delta-sigma chip capable of operating up to a sample rate of 216kHz and with a high specified dynamic range of 123dB. A Texas Instruments TAS1020B is used for the USB interface, which limites operation via USB to sample rates of 96kHz and below. Two AD1896 asynchronous sample-rate converter chips are present, but I'm not sure whether they're used only to prepare the data for the Aux outputs or for a more fundamental purpose. Benchmark uses what they call UltraLock to reduce the level of jitter.

Sound Quality
I prepared some 24/96 needle drops using the ADC1 converter's USB output and the inexpensive Vinyl Studio app running on my MacBook Pro. In level-matched comparisons of "The Lark," from Moving Hearts'The Storm (LP, Tara 1304), with a 24/192 needle drop made with the Ayre QA-9 A/D converter ($3950) I reviewed in November 2012, the Benchmark had a little more weight and authority to its low frequencies, the Ayre a more delicately drawn soundstage.

The differences were in the same direction with a 24/96 transfer of English singer-pianist Peter Skellern performing "The Continental," from Astaire (LP, English Mercury 9109 702). The Ayre's imaging was a touch more palpable, the Benchmark's sound slightly more upfront, with firmer lows. But overall, the transfers from vinyl were more alike than different. I couldn't hear any meaningful differences between the Ayre and Benchmark converters with "Die Tänzerin," from Ulla Meinecke's Wenn Schon Nicht für Immer, dann Wenigstens für Ewig (LP, German RCA 426124).

I used the Benchmark last May as the master converter to record the final rehearsal of Bob Reina's band Attention Screen performing jazz compositions for trumpet, double bass, drums, and church organ. The outputs of the main pair of organ mikes, a pair of DPA 4003 omnis amplified by a Millennia Media preamp, were fed to the Benchmark running at 88.2kHz, with a Metric Halo MIO2882 used for the spot mikes on the other instruments slaved to the ADC1's word-clock output. The Benchmark's 24-bit AES/EBU output was fed to a Metric Halo ULN-2, and both Metric Halos were connected to my MacBook Pro via FireWire. That way, I could record all 10 channels using Metric Halo's multitrack Record Panel app.

Listening to the sound of the organ mikes, it was obvious that the Benchmark ADC1 had done a great job. With its clean high frequencies and weighty, extended lows, it faithfully captured the magnificence of the newly restored Ralph and Alice Greenlaw Memorial pipe organ at the Community Church of Douglaston, Queens. In the work intended to be the concert's finale, Bob gives the organ's bass pipes a workout—the last note, a sustained, lusty 32Hz C, shook the windows of my listening room.

Conclusion
Benchmark's versatile and full-featured ADC1 USB both measures superbly well and produces digital files that sound equally superb. It offers performance for which, a decade ago, you would have had to pay five times its price of $1795. It also offers better resolution than you get from inexpensive converters like E-MU's popular e404, whose nominally 24-bit A/D converter section actually has about 17-bit resolution. Unless the fact that, unlike the Ayre QA-9 it lacks DSD encoding is a problem for you, I highly recommend the ADC1 for transferring your LPs to digital.

We can add another product to this illustrious list: the Kinergetics KCD-55 Ultra. Although the KCD-55 Ultra is the latest processor I've auditioned using UltraAnalog DACs, Kinergetics would have been the first to market with such a product had it not been for a quirk of fate. At the 1989 Audio Engineering Society Convention in New York, UltraAnalog had hundreds of thousands of dollars' worth of custom equipment stolen—including Kinergetics' only prototype of an UltraAnalog-based version of their highly regarded KCD-40 CD player (footnote 1). Rather than start over with the CD player, Kinergetics decided to build an ambitious outboard D/A converter instead.

The result is the KCD-55 Ultra reviewed here, Kinergetics' top-end digital product. As often happens after loss or destruction, the opportunity to rebuild creates something better than what was originally lost. Kinergetics used their months of design experience on the KCD-40 Ultra as a launching platform for this more high-reaching effort.

Let's see what they've come up with.

Technical description
The KCD-55 Ultra (hereafter called simply the Ultra) is an unusually and beautifully built component. The thick front, rear, and side panels are made from CNC machined extruded aluminum. These panels are then attached to a separate sheet-metal chassis that holds the printed circuit boards. This construction technique results in both a very solid build and an expensive, elegant appearance.

Two knobs, two switches, and an LED comprise the front-panel controls and indicators. These functions are common to most digital processors: power on/off indicator LED, polarity reversal switch, and selection between two digital inputs. The knobs, however, are unusual. The first attenuates the level from the "variable" analog outputs, obviating the need for a preamplifier or passive level control in CD-only systems. Rare, but not that unusual. The knob on the panel's right-hand side, however, is another story. Marked "Processor," this control reportedly allows the user to tailor the sound to match system and tastes. I'll have more to say about this function later.

The rear panel has two pairs of analog outputs on RCA jacks, one fixed level and one variable, with the latter controlled by the front-panel volume knob. Although the rear panel is machined for two coaxial inputs, only one RCA jack is fitted. The second input's hole is covered by a plug (the circuit topology can accommodate up to four digital inputs). An AT&T ST-type optical input is provided on newer production, reflecting the general agreement that ST-type optical is superior to either coaxial or Toslink interfaces. Kinergetics commendably added the expensive ST-type input without raising the Ultra's retail price.

Popping the top panel, I was surprised by the Ultra's minimalist design; there were fewer parts than found in most converters. The analog output stage, power supply, and "glue logic" (the chips that make everything work together) were all executed with a minimum of parts. A closer examination, however, revealed some unusual and tweaky design techniques.

Starting with the power supply, three transformers are used, one toroidal for the analog stages and two standard laminated types for the digital supply. These are mounted on a separate pcb in the front left-hand corner, away from the rest of the circuit. The digital supply transformers were selected on the basis of their ability to keep RF noise generated by the digital circuits from getting back into the AC line.

There are a total of eight regulation stages for six supply voltages (two digital and four analog). The analog supply is regulated by standard three-pin IC regulators, then, for the critical output stage, is regulated again by a discrete circuit using a pair of OP42 op-amp chips located right next to the output stage. Wherever a rail supplies a circuit stage, a Roederstein polypropylene decoupling capacitor is employed.

The input receiver is the ubiquitous Yamaha YMJ3623B (footnote 2), but implemented with Kinergetics' proprietary jitter-reduction circuit. A Sony CXD1144B chip provides 8x-oversampling digital filtering. This chip is the most powerful (highest number of taps) and expensive of the digital filters, selected after auditioning a variety of filter chips. The input receiver, filter, AT&T optical input jack, and associated components are mounted on a separate pcb toward the back of the chassis.

The rest of the circuit is contained on a large pcb that consumes about a third of the chassis's real estate. I was particularly impressed by the output stage: it's all discrete, class-A, direct-coupled, and uses high-quality parts. The circuit is essentially a "discrete op-amp," with bipolar transistors and JFETS providing gain and a Precision Monolithics BUF-03 acting as the op-amp's output. This is the same part Corey Greenberg was so enthusiastic about in his DIY buffered passive preamp article last November. It can drive large amounts of current (40mA), and has no problem with low impedances (it was designed as a 75 ohm line driver).

Because the Ultra's analog stage is direct-coupled, an NE5532 op-amp and pair of trim pots form a DC servo circuit to prevent DC from appearing at the analog output. The front-panel level control is a high-quality metal-film type ganged pot (with separate left and right channel adjustment) that attenuates the output level from the variable output jacks. Note that this pot is an attenuator after the gain stage, not part of the gain-determining feedback loop. De-emphasis is passive, switched in by a solid-state device instead of by a relay.

Now, about that front-panel knob marked "Processor." According to designer Tony DiChiro, it's a "hysteresis control." This circuit is found in all Kinergetics electronics, but the Ultra is the first product to provide user adjustment of the circuit (footnote 3). The control's range is quite narrow to prevent overuse and sonic degradation, but reportedly produces enough change in sound for final system matching. I'll comment on the control's sonic effects later.

Where the Ultra really gets elaborate, however, is in the critical digital/analog conversion stage. Not only does the Ultra use the best DACs currently available—the two-channel, 20-bit UltraAnalog DAC D20400—but it employs two of them for differential operation. In this scheme, each dual DAC receives the digital code representing the analog signal and the same code inverted. One channel of the dual DAC converts one polarity to analog, the other channel converts the other polarity. This produces two analog output signals of opposite polarity for each channel. When the two opposite-polarity signals are amplified differentially in the output stage, only the wanted difference between the two signals is amplified. Any noise, distortion, or artifacts common to both channels are thus rejected—a phenomenon called "common-mode rejection."

Kinergetics uses this technique in all their digital products. It becomes very expensive, however, when using 20-bit UltraAnalog DACs. For comparison, the KCD-55p, which uses the same chassis, power supply, pcb, digital filter, and discrete output stage as the Ultra, sells for $1695. The only difference? The KCD-55p uses the Analog Devices AD1860 DACs found in the KCD-40 (and the ST optical input is optional). Incidentally, owners of the KCD-55p can have their units upgraded to the Ultra version for $2500, about $200 more than if the Ultra were purchased initially.

Overall, I was impressed by KCD-55 Ultra's build quality, thoughtful design (particularly the additional discrete regulation stage for the analog output and the output section itself), and use of two very expensive UltraAnalog 20-bit DACs.

Listening
A logical comparison for the Ultra was the Audio Research DAC1-20. The ARC is similarly priced ($3500), uses an UltraAnalog DAC, and has established a reference level of musicality at its price. Any digital processor selling for about the same money must regard the DAC1-20 as formidable competition.

After my first listen, I was pleasantly surprised by the Ultra; it more than held its own against the DAC1-20 in some respects. In addition, the Ultra offered a different perspective on the DAC1 that may appeal to many music lovers.

First, the Ultra had excellent bass—tight, controlled, and with powerful dynamic impact. In fact, the Ultra had the best low-frequency reproduction of any of the UltraAnalog-based processors I've auditioned, save the Mark Levinson No.30. I have previously criticized the bass performance of UltraAnalog-based converters as lightweight, soft, and lacking dynamics. I had mistakenly attributed this characteristic to the DAC, not the implementation. The Ultra (and the No.30) set the record straight. The bottom end seemed to extend deeper than heard through the DAC1-20, with a more taut and less fat rendering. Dynamics were also superior, with a greater sense of slam and power. Listen to the kick drum in the tune "Are You Scary?" from the new Sheffield CD The Usual Suspects. Through the Ultra, it had a punch and depth rivaled only by the No.30 and Wadia 2000.

In addition to being tight and punchy, the low end was round, solid, and had a satisfying fullness. Acoustic bass had a warmth and liquidity I particularly enjoyed. Pitch definition was excellent, with clear articulation of each note, even with fast and complex bass lines. Listen to the acoustic bass on "Round Midnight" from Kenny Rankin's Because of You (Chesky JD63, reviewed in this issue). This is a stunning recording; the track mentioned (vocal and bass only) is very revealing of how well a converter conveys the instrument's roundness, warmth, and fine detail. Through better processors, the instrument will sound more "bass-like," and less flat or wooden. Through the Ultra, the bass was superbly portrayed. Overall, the Ultra's bass reproduction was exemplary, and clearly a step above the DAC1-20's.

When it came to presentation of midrange and treble textures, however, I preferred the DAC1-20. The Ultra had a trace of hardness through the upper mids and lower treble that was contrasted with the DAC1-20's ease and liquidity. On Three-Way Mirror (Reference Recordings RR-24CD), for example, the acoustic guitar, flute, and cymbals sounded slightly edgy. In addition, the presentation was more forward, with less sense of ease. It wasn't a case of not enjoying the Ultra—it was very liquid and had more natural rendering of midrange textures than most processors—but it fell short of the Audio Research unit, which excels in these areas. Long sessions with the Ultra tended to be more fatiguing than with the DAC1-20, and there was less inclination to listen at high levels.

Though the treble was smooth, I wouldn't characterize the Ultra as laid-back. Rather, it struck a good balance between revealing HF detail and being overly soft, with a tilt toward revealing detail. Treble textures had a trace of hardness compared to the DAC1-20, and sounded more "digital." In addition, the treble could at times sound a little on the etched and analytical side of reality, rather than soft and gentle. I wouldn't use the words "refined" and "delicate" to describe the Ultra's treble, characterizations I've used to convey the DAC1-20's presentation.

These drawbacks were more than offset, however, by the Ultra's superb rendering of detail and ability to separate individual instruments from the whole. The presentation was infused with a wealth of fine detail; subtle sounds that were blurred through the other processors became vibrant and alive through the Ultra. The brushed snare drum on Jazz at the Pawnshop (Proprius PRCD 7778), for example, was made up of many finely woven components rather than fused into a single sound. It also had a vibrancy and palpability rarely heard through any processor. Inner detail was rendered with a precision and life that made me feel as though I were hearing more music. Without a doubt, the Ultra presents another level of information to the listener. In this regard, the Ultra approached the No.30's stunning resolution of detail, but with less ease and warmth.

There is one important presentation aspect in which the Ultra is superior to just about every other converter I've auditioned, except the No.30: creating the illusion that the presentation was made up of individual images, not merely variations in a synthetic tapestry. On Robert Lucas's Luke and the Locomotives (AudioQuest AQ-CD1004) there was a convincing impression of the individual band members in the listening room (especially on the track "Feel Like Going Home"). The soundstage was beautifully fleshed out, with superb delineation of each instrument. Images were tight, well-defined, and thrown with pinpoint precision. This superb image specificity was accompanied by a feeling of air and space surrounding the instrumental outlines. The Ultra's portrayal of depth and space was excellent, with a distinct three-dimensional quality. Naturally miked music with subtle spatial cues was well served by the Ultra: the listening room assumed a wide range of apparent sizes throughout the auditioning, accurately reflecting the recording's characteristics. Soundstage width was similarly good; the musical presentation was thrown in a wide arc across the listening room, making it easy to enjoy the music and forget about the loudspeakers.

The Ultra's crystal-clear soundstage transparency further heightened these impressions. The result was a convincing illusion of individual instruments hanging in space around the loudspeakers. I really enjoyed this aspect of the Ultra's presentation; music was more lifelike and less homogenized. If I had to name the Ultra's best attribute, it would be this. Moreover, the ability to separate individual images from the presentation is rare in digital processors, and an aspect I find musically important. The Ultra was the antithesis of synthetic homogeneity.

When it came to dynamics, the Ultra was topnotch. Transient detail was razor-sharp, with fast, clean leading edges. Drums had power and energy not heard from the DAC1-20. The sound of the stick hitting the head was well conveyed, giving a greater feeling of the drummer's rhythmic contribution to the music. I recently recorded a great-sounding drum kit with just two microphones in a live room with a very pure, all-tube signal path. Playing back the DAT master tapes revealed the Ultra's ability to recreate the steep attack that gives drums immediacy and life. The Ultra's dynamic quality and transient quickness conveyed an increased sense of rhythm and energy, especially on jazz and blues.

Most of these impressions were gained with the Ultra going through an Audio Research LS2 line stage. Driving the power amplifiers directly from the variable outputs slightly improved the overall transparency and palpability. The LS2 is, however, extraordinarily transparent; listeners with most other preamps that are not as neutral will realize much greater benefits from using the Ultra's variable outputs and front-panel volume control.

Finally, I heard very little difference when using the front-panel "Processor" control. The presentation seemed a little softer toward the counterclockwise side, and more immediate with it turned up. Its effect, however, was far less then the differences described between the Ultra and DAC1-20. I must also add one complaint about the Ultra: there was a loud click when switching inputs. Users are therefore advised to mute their preamps or turn down the Ultra's level control when switching between digital sources.

Conclusion
If I had to count my favorite digital processors on one hand, the Kinergetics KCD-55 Ultra would be included. I really enjoyed my time with the Ultra: It consistently conveyed the music with life and vitality. The Ultra had many strong points that made it musically involving, especially its ability to differentiate individual instrumental images. I also found that its presentation of detail and transient quickness gave music a heightened sense of energy. Further, the Ultra excelled at throwing a spacious, well-delineated soundstage with pinpoint images. To top it off, the bass rendering combined power and punch with liquidity and pitch definition. Overall, I thought it was a terrific processor.

Having said that, I can easily see how many listeners will prefer the Audio Research DAC1-20's greater ease and gentleness over the Ultra's more incisive presentation. The DAC1-20 was more analog-like in its portrayal of instrumental textures, but ultimately presented less information to the listener. Playback systems that lean toward the etched and analytical may benefit from the DAC1-20's greater ease. I enjoyed listening to both processors; I'm sure their different interpretations will each find an audience.

The Ultra's variable output is a powerful attraction for listeners with CD-based systems. No preamp, passive level control, or extra interconnects are needed: just plug your CD transport and DAT recorder into the Ultra, and plug that straight into the power amplifier. Getting an active preamp and second pair of interconnects out of the signal path—especially if they're colored—will greatly improve a playback system's overall transparency and musicality.

After just having spent many enjoyable hours with the $14,000 Mark Levinson No.30, the KCD-55 Ultra struck me as emulating some of that reference product's best characteristics. The detail, image specificity, and dynamic contrast that made the No.30 so stunning were heard in the Ultra, albeit to a much smaller degree. The No.30 did this, however, with an ease not heard from the Ultra. Nevertheless, this says a lot about the Ultra's special qualities.

If you're considering buying one of the megabuck processors, give the Kinergetics KCD-55 Ultra a listen. It's up there with best of them.

Footnote 2: This Yamaha receiver chip outputs 16-bit digital data, meaning that it truncates S/PDIF or AES/EBU datastreams featuring bit depths greater than 16.—John Atkinson

Footnote 3: The Modern Dictionary of Electronics (Howard Sams, publisher) defines hysteresis as (definition #5) "A form of nonlinearity in which the response of a circuit to a particular set of input conditions depends not only on the instantaneous values of those conditions, but also on the immediate past (recent history) of the input and output signal. Hysteretical behavior is characterized by inability to 'retrace' exactly on the reverse swing a particular locus of input/output conditions...This term literally means to lag behind."—Robert Harley

There. I've said it. I am not an audiophile; ie, someone who's in love with recorded sound for its own sake. The search for ideal sound can leave a person burned out and broke.

That might be why I love Internet radio via Bluetooth. So much content. So convenient, via smartphone or laptop. As for computer-audio downloads, they're too complicated, chaotic, and costly.

I might pay for streaming high-resolution audio, if the content and convenience are there. More than one computer guru has said that digital subscriptions are the future. Who wants to "own" and store physical media?

I love Musical Fidelity's M1SDAC with aptX Bluetooth (see "Sam's Space" in October), even though my iPhone 4, updated to iOS 7, doesn't support aptX.

The aptX codec is said to automatically optimize the Bluetooth receiving device for the best sound quality possible with each incoming aptX signal. I have it now, with my new (June 2013) Macbook Air.

Is aptX a big deal? Some say it's not. As my colleague Bob Deutsch says, "It depends on the implementation."

AptX is back-compatible with earlier Bluetooth codecs. It streams at up to 380kbps, but it can work with devices that stream at lower bitrates, including 128 and 256kbps.

When my iPhone 4 ran iOS4, Bluetooth streamed at 128kbps, if I'm not mistaken. Now, with iOS7 installed, it streams at 256kbps, with better sound: more resolution, more air, fewer dropouts, more there there. If you have an iPhone 4 or later, it's definitely worth installing iOS7.

I'll pass along a couple of iPhone tips.

If you no longer get a Bluetooth connection, or you keep losing it, you may have a dirty dock. Some folks online have recommended that you brush around the dock's connections with a clean, soft toothbrush. Or flatten the tip of a cotton swab with a pair of pliers, dip it in grain alcohol, and wiggle it around. (I didn't tell you to do this.) This worked for a while.

Then, a genius at a bar told me another secret: Keep fewer programs running in the background. That was like a visit to the dental hygienist. Bluetooth became Cleanteeth: brighter, cleaner, more refreshing, less stale. Now the difference between my iPhone 4 with iOS7 and my Macbook Air with Mountain Lion running aptX Bluetooth was less pronounced. Of course, if you really want aptX on your smartphone or tablet, you can look to Samsung.

The Musical Fidelity MS1DAC sounds very good indeed; it's a DAC, a headphone amp, a line stage, and a headphone amp, all in one. But if all you want is aptX Bluetooth, you might not want to pay $1499 for an MS1DAC. You may already have a DAC you like. There must be a cheaper way to do this.

Ho, ho, ho, there is. At $249.95, the Arcam rBlink seems expensive for a tiny black box measuring only 2.9" (75mm) wide by 1" (26mm) high by 3.9" (100mm) deep and weighing just 12oz (350gm). But it will put Internet radio wirelessly at your fingertips, from laptop, tablet, or smartphone.

Arcam describes the rBlink as a "high-performance Bluetooth audio receiver and digital-to-analog converter." It works well, whether or not you already own a DAC. The rBlink has its own digital-to-analog converter and left and right analog outputs. Just connect it to your integrated amplifier or preamp. Arcam even includes interconnects to get you up and running fast. (But consider getting better interconnects.)

Don't dis the rBlink's built-in DAC, which uses Burr-Brown's PCM5102 DAC chip. It's no slouch, as the Brits like to say. Don't feel you need buy a separate DAC.

On the other hand, if you already own a DAC, you can use the rBlink's S/PDIF coaxial output. I used my Musical Fidelity V-DACII, now replaced by the V90-DAC.

I ran the rBlink into the Croft Acoustics Phono Integrated amplifier, first using the rBlink's analog RCA outputs and a decent set of interconnects. I cackled when I heard the sound. Art Dudley and Stephen Mejias are right: The Croft is crazy good in musical rather than in audiophile terms. You can tell from its rich, full-bodied sound that it wasn't designed with test tones in mind.

In the blink of an eye, I had Internet radio in excellent sound, depending on the quality of the stream.

Why are some streams so crummy—and from the same station? Classical New England—now re-rebranded as WCRB in order to show its real share in the ratings—broadcasts live concerts from its Fraser Performance Studio, always in excellent sound. Yet when they broadcast from Boston's Symphony Hall or Tanglewood, the sound deteriorates.

The day after I received the rBlink, John DeVore, of DeVore Fidelity, drove up to my digs with his new beast: the Orangutan O/93 loudspeakers. John groused but didn't growl about the Bluetooth sound, but that was before I got the Macbook Air and upgraded my iPhone 4 to iOS7.

I didn't let on to John, but CDs do sound better than Bluetooth at the moment. But we can expect Bluetooth to evolve and improve. It's one good reason not to plop down more than $249.

I rattled the zookeeper's cage.

"The sound is fine by me," I told John. "I just DeVore it. It's all free. Swiss Radio Jazz. Nostalgie Jazz. BBC 3. Classic FM. France Musique. Symphonycast.com. Radio Dismuke, with music from the 1920s and '30s. If I want Perfect Sound Forever, I'll get off my butt and fetch a CD."

Which brings up a point: I never found the rBlink irritating. Well, I did with a few piss-poor streams, but these are easily avoided. The rBlink seemed to hold on to the signal—fewer dropouts—better with my Macbook Air than with my long-in-the-tooth iPhone, until I installed iOS7.

The rBlink has one oddity: To pair it with a Bluetooth device, you need to push in the pairing button with the tip of a pen. I have no blinking idea why Arcam doesn't provide a simple pushbutton. While pairing, the rBlink changes from steady red to blinking purple. Pairing completed, the purple light glows steadily. When the rBlink is connected to a Bluetooth device, the light glows a steady blue. I love this thing!

Since the Croft's measurements rattled John Atkinson in October, I tried my LFD LE IV integrated amplifier. I got the same excellent results. The aptX codec no longer mattered so much, although you might as well have it, if you can.

If you're looking for a gift idea, and Mom and Dad have Bluetooth devices, put an Arcam rBlink under the tree. While you're at it, get another for yourself. You can bring Bluetooth to just about any audio device, including a kitchen radio.

Highly recommended.

On the mainland across from these isles, in Port Hueneme, is Channel Islands Audio (aka CIAudio), a company of modest size that's been around for 17 years and makes compact power amps, preamplifiers, and digital/analog converters. Like the islands, always sitting quietly off the coast but barely seen, CIAudio had always been in the periphery of my audio world, though I knew little about them. So I was pleased to get the Transient Mk.II USB DAC-preamp and optional VDC•5 Mk.II power supply in my system, to learn firsthand what CIAudio is up to DACwise.

At 4.45" wide by 2.9" high by 5.25" deep, the Transient Mk.II ($699) fits into an open palm and is thus very portable, hence its name. The understated case, made of 1?8"-thick aluminum, is nicely finished, with a 3/16"-thick front panel. All hardware is nonmagnetic stainless steel, and the Transient feels solid and nonresonant.

On the front, starting at the left, is something I love to see on a DAC of any price: a row of six LEDs that indicate the incoming signal's sampling rate, in this case 44.1, 48, 88.2, 96, 176.4, and 192kHz. On the right are stacked two small buttons, with arrows to indicate volume up and down, and next to each of these is another LED. Each volume button's LED flashes as you tap it, then stays lit when you hit its limit. You can also use these buttons to set the DAC to Line Level, if you don't need the volume function.

On the rear panel, starting at the left, are the unbalanced left and right audio outputs (RCA). At the right top is the USB input, and below that the three digital outputs: a 75 ohm BNC connector galvanically isolated for S/PDIF (an RCA-to-BNC adapter is included), a mini-DIN jack for I²S, and an HDMI jack for differential I²S (also used by PS Audio, Wyred4Sound, and others).

With its variety of outputs and 24-bit volume control, the Transient can be used as a USB DAC, a DAC-preamp, or a USB-to-S/PDIF or USB-to-I²S converter. The only thing missing for a desktop system is a headphone jack.

Power that doesn't corrupt
There's one more thing on the Transient's rear panel: to the right of the digital outputs is a DC input, for the optional VDC•5 Mk.II power supply. Although the Transient was designed to be powered by the 5V available from a computer's USB output and still output a full 2V analog signal, and has onboard filtering to clean up that incoming power, CIAudio suggests that if you want the best sound, hook it up to the VDC•5 Mk.II. Housed in a case identical to the Transient's, the VDC•5 costs $329, bringing the total package price to $1028. The VDC•5 is marketed to also replace the power supply that comes with the Squeezebox Touch music server.

CIAudio points out that the quality of DC provided by USB varies quite a bit from computer to computer, and can often contain noise, ripple, or simply not enough current, all of which can affect sound quality. The VDC•5 Mk.II's regulated linear power supply is designed to feed the Transient pure high-current (2.5 amperes) DC with no noise or ripple.

The VDC•5's front panel has only a single power-indicator LED; on the rear are a power switch, an IEC power socket, and a 5V output jack. A short cable is included to connect it to the Transient, the two models together forming a matched set. Both are "handcrafted" in Ventura, California, and come with a five-year warranty on parts and labor.

The Transient's USB input connects to a four-layer custom USB board featuring an XMOS multicore processor and ultraprecise clocks, which CIAudio claims work with the computer to generate a low-jitter I²S signal. The precisely clocked I²S output is then fed to four circuits: a buffer for the I²S DIN output jack, a differential buffer for the HDMI I²S connector, an S/PDIF transmitter for BNC S/PDIF output (based on a Wolfson WM8805 chip), and into the onboard Wolfson DAC chip. Takman resistors and WIMA polypropylene capacitors are used for signal circuits.

CIAudio's Dusty Vawter emphasized to me in an e-mail that they wanted the Transient to be completely portable, which meant that it had to be able to be run from the 5V supply available from a laptop's USB port. "There are very few DAC [integrated circuits] with this capability," Vawter said. "We listened to offerings from ESS, TI, and Wolfson. In the end, we chose the Wolfson due to its superior musicality and benefit of having a built-in 24-bit volume control."

Set-Up
Connected to a Lightning–to–USB Camera Adapter, the Transient didn't power up when connected to my iOS7 iPad Air, iPad mini, or iPhone—not surprising, considering how the DAC's design sucks up every bit of current it can get, and these battery-powered iDevices don't put out much via USB.

But I was surprised to find that, after powering the Transient with the VDC•5 power supply, I still got a warning on the iDevice display that the power draw was too high. Interestingly, this time the warning notice mentioned the CIAudio DAC by name, so more info was getting through, but not enough to include music. Occasionally, the Transient's LEDs would blink. Only a handful of USB DACs have been able to work straight off my iPad, by the way, so this is not a really a criticism, just a note to those who might want to use the Transient this way. To fire it up with an iDevice, you could also try a powered USB hub.

However, when I attached the Transient to my MacBook Pro, everything powered up fine without the VDC•5, and worked right away—no need to reboot or for extra drivers. The DAC also works with Linux and, like all high-resolution converters, requires download and installation of a driver for computers running Windows XP through Windows 8. All of my listening was done with the Transient connected to the battery-powered MacBook.

Music that might corrupt
In honor of the Channel Islands, I felt obligated to throw some California music at the Transient, beginning with the fabulous new 24-bit/192kHz remasterings of the Grateful Dead's studio albums, from HDtracks. These are remastering done to perfection—even less-than-fanatical Deadheads should grab them.

I skipped around a couple albums, then settled down with American Beauty for some extended listening. Everything sounded in the right place, with a beautiful flow from top to bottom. I used the loose, open arrangement of "Sugar Magnolia" to check for any changes in the sound as I switched the external power supply on and off. (The Transient will automatically default to USB power if the VDC•5 is switched off, and automatically default to the VDC•5 when the latter is switched on.)

I was impressed by what I heard from these cards, and concluded in a Follow-Up that "the Xonar Essence STX and its PCI-bus equivalent, the Xonar Essence ST, can be recommended to those on restricted budgets who wish to incorporate a PC into their high-end rigs." So when ASUS announced that it was introducing a version of its standalone Xonar Essence One D/A headphone amplifier fitted with JRC's high-performance Muses op-amps, which I had first experienced when I reviewed the Esoteric D-07 D/A processor in January 2011, I asked for a review sample.

The Muses Edition
The Xonar Essence One is a hefty processor housed in an elegant, black-painted enclosure. The basic Essence One costs $599; the Muses Edition, which can be distinguished by the black color of the stylized lion graphic on the top of its extruded-aluminum sleeve, costs $899. On its front panel are, from left to right: a power button; buttons to select Upsampling, Input, and Mute (the selected input LED turns from blue to red when Mute is selected); a large volume control for the line outputs, to its right an arc of blue LEDs; a smaller volume control for the headphone output; and a single ¼" (6.3mm) stereo headphone jack. Because there are independent volume controls for the line and headphone outputs, the line output doesn't mute when headphones are plugged in. On the rear panel are pairs of RCA and XLR jacks for the single-ended and balanced line outputs, respectively, and input jacks for USB data and S/PDIF data on TosLink and coaxial links.

When the upsampling function is off, the sample rate of the incoming data—44.1, 48, 88.2, 96, 176.4, or 192kHz—is displayed by the arc of LEDs. When the upsampling button is pressed, data at 44.1kHz are upsampled to 352.8kHz; data at 48kHz and its multiples are upsampled to 384kHz. ASUS calls this "Symmetrical 8x upsampling," because the upsampled frequency is an integer multiple of the incoming rate. None of the LEDs illuminates when upsampling is engaged. However, there is an LED at the top of the arc labeled Bit Perfect; though this never lit up when I used the Xonar processor with my Mac mini or MacBook Pro, it is supposed to do so when the Essence One is connected to a Windows PC and the necessary ASIO driver (supplied on a CD-ROM) is installed. (I couldn't verify this, as all my auditioning and measuring was with Macs.)

Inside the box, the circuitry is neatly laid out on a large printed circuit board, with a cutout in the board for the toroidal power transformer. S/PDIF data are routed to an AKM AK4113 receiver chip; USB data are handled by a C-Media CM6631 USB receiver (the same chip used in the Schiit Bifrost DAC reviewed by Jon Iverson in August 2013). The audio data are passed first to an Analog Devices SHARC ADSP-21261 40-bit floating-point DSP chip, then to a pair of TI's PCM1795 DAC chips.

The PCM1795 is a two-channel, 32-bit–resolution device that is pin-for-pin compatible with the earlier and widely used PCM1792 chip; according to its datasheet, the PCM1795 is DSD-capable, but the Essence One will not decode DSD data. Also according to its datasheet, the PCM1795 is intended to operate up to a sample rate of 200kHz, so I'm not sure how ASUS is able to use it at 352.8 or 384kHz when upsampling is engaged. I wondered, as the chip is a two-channel part, each channel could be fed alternate samples, to give an effective doubling of the sample rate, as was once done by Stan Curtis in a mid-1980s Cambridge Audio CD player. However, my measurements (see Sidebar) suggest this isn't the case.

Unusually, all of the eight-pin op-amp chips, a mix of JRC and Muses devices, are socketed. The six Muses 01 dual op-amps, which follow the DAC chips, are made with advanced fabrication techniques said to reduce crosstalk and produce better-balanced left/right channel symmetry, and use oxygen-free copper leads. The headphone amplifier appears to be based on a pair of TI's LME49720 high-performance op-amp chips and an LME49600 high-current output driver; the line outputs appear to be based on TI's LM4562 ultra-low-distortion, low-noise, high-slew-rate op-amps. Other than the headphone output, all the analog audio circuitry is heavily bypassed with local electrolytic and plastic-film capacitors. Overall, the parts count and the quality of those parts are very high for a relatively inexpensive product.

Sound Quality
I used the Xonar Essence One Muses Edition for all my regular headphone listening during the fall of 2013, as well as during the preparation of my review of the Audeze LCD-X headphones elsewhere in this issue. I also used it in my big rig (see the "Associated Equipment" sidebar). Although the Xonar DAC had already been reviewed by Michael Lavorgna and Dinny FitzPatrick on, respectively, our sister websites AudioStream.com and InnerFidelity.com, I didn't read my colleagues' reviews until I had finished my own auditioning. But my impressions of the Xonar's sound quality to some extent echo theirs.

Used as a DAC without upsampling engaged, the Xonar Essence One didn't resolve recorded details as readily as the NAD M51 and Auralic Vega, though it's fair to note that those DACs cost very much more. There was a smooth, rounded-off quality to the Xonar's line outputs that was a benefit with typically overcooked rock recordings, such as the Pretenders'"Talk of the Town" and "Back on the Chain Gang," from The Singles (ALAC files from CD, Sire/WEA; and yes, I am rediscovering and relishing that delicious sob in Chrissie Hynde's voice).

But this character obscured the fact that my reissue of Richie Havens's 1969 album Richard P. Havens, 1983 (CD, Polydor 835 212-2) had apparently been mastered from an LP rather than from the original master tapes. And the dry acoustic of Yamaha's YASI recital hall in Attention Screen's "13 Trojans of Vundo," from their live recording Takes Flight at Yamaha (16/44.1 master file for CD, Stereophile STPH021-2), also seemed a little suppressed compared with the Auralic Vega's presentation. Chris Jones's fretless Fender Jazz bass guitar also sounded a touch softer, but the Essence One still revealed the hit in sound quality resulting from the lossy audio encoding in the video from the concert that I posted on YouTube, to which I had added the audio mix from the CD.

Tiring of dropping receivers off our building's roof or ramming TVs (tied to the back of a pickup truck) into the shop's brick wall, Tom soon discovered that one of our customers owned a machine shop with an industrial press. Pay dirt. Somewhere there are camcorder cassettes of what transpired, but let's just say that even the best casework was no match for this giant squishing machine. Tom's videos would first show the component being crushed. Then he would gleefully pan to the pressure gauge, as it rose higher and higher. Then back to the metal pancake.

Which brings us to MSB Technology's Analog DAC.

This product's design and shape suggest a typical MSB component that has been squeezed tight in an industrial press, then sanded and buffed to a smooth finish. Call it an audiophile pancake. In fact, it resembles in size and thickness the bigger-than-plate-size blueberry pancakes at Hoover's Beef Palace, just up the road from me in Templeton, California (yes, this is true!). I'll bet Tom would be challenged in trying destroy the Analog DAC, and appreciate how well it's made.

I reviewed MSB's Diamond DAC IV (since renamed the Diamond DAC IV plus) in the October 2012 issue, with Diamond Power Base and other upgrades ($43,325), and it remains the best digital I've heard in my home system. When I spied the new Analog DAC at the 2013 Consumer Electronics Show, in Las Vegas, and was told that it's their new, lower-cost product, I was interested before I'd even heard any specifics. And when I did hear those specifics, they were interesting.

Best-case scenario
Let's start with that enclosure. The stealthy-looking Analog DAC ($6995) is CNC-machined from a solid hunk of aircraft-grade aluminum and comes in matte silver or black, with custom colors available for $699. They leave much of the metal in, removing it only where they need to stuff electronics—what's left feels like a solid plank of 7/8"-thick metal. The case is 17.5" wide and 12.5" deep and sports curved sides, with a semicircular bulge at each corner for a little spike foot. Underneath is a hatch to gain access to the main electronics, and there are three slots on the back for the inputs. It looks like something that would fly if tossed like a Frisbee.

On the back, starting at the left, are the balanced and unbalanced analog outputs and analog input, grouped by channel. MSB recommends using the unbalanced outputs if possible—they claim that the DAC is "fundamentally single-ended." Unless the optional volume control is installed, the single-ended analog inputs are passed directly to the outputs. With the volume control, this input can be either volume controlled or not, depending on the menu settings; MSB suggests that it's ideal for adding a vinyl input, if you're using the Analog DAC as a preamp. This input should be shorted when not in use, as it was during my testing.

To the right of the output/analog input section are three slots for the various digital input options. The five possible choices for the three spots are: Optical and coax S/PDIF inputs (on one input block), XLR balanced AES/EBU input, MSB network input (it looks like an Ethernet jack, so is colored bright green), Pro I²S input, and a 32-bit/384kHz PCM/DSD-compatible USB input. I'll go over the prices of these options later; it can be a bit perplexing. My review sample came with the Optical/Coax, MSB network, and USB options.

To the right of the inputs is a jack for the DC power supply. There are two power-supply options: the linear Basic Desktop supply, with two transformers, is included in the basic price and gets the job done; a more advanced supply, the Analog Power Base, is housed in a case that looks just like the Analog DAC and makes a nice stacking companion (yes, like pancakes). It contains five transformers—for complete isolation of digital processing, clocks, and analog DAC modules—as well as a 12V power trigger for remote operation. On the back of the Power Base are an IEC AC power receptacle, a DC out jack, trigger jacks, and a teeny-tiny power switch that glows red when off, green when on. I'm wise to MSB products, so I quickly found this unmarked switchette and figured how to turn it on without help. I had only the Analog Power Base upgrade on hand for listening, so can't remark on what improvements, if any, it makes over the Basic Desktop supply. The Analog Power Base adds $2995 to the price: total so far, $9990.

Back to the Analog DAC. The front of its case is bare, smooth metal, but on top, at right front, the volume control and input selector sit flush with the surface. The volume selector is puck-sized with the input button a small circular indent in the volume puck and held in by gravity. How do I know about the gravity thing? When I first turned the review sample over to check out the bottom, the heavy volume knob and small input button fell out and bounced on the floor. Oops. Luckily, no dents.

To the left of the volume control is a small grid of pinholes in the aluminum; under these is the white LED display. The large letters and numbers are quite bright and let you know the software version on startup, the input selected, the sample rate, and, as you spin the knob, the volume setting. The interaction between the volume and input selector and the display have a great feel, and there's a very satisfying little clicking sound as you bounce the volume up and down. At the rear of the top panel are the MSB logo, and labels for the outputs in light colored type.

This arrangement, with the volume control on top, worked great when I perched the Meridian Sooloos Control 15 (with its small stand) atop the DAC. However, this might prove problematic with a normal component on top, as I found when I added to the stack the MSB Universal Media Transport plus. With the UMT+ underneath, the feet lined up perfectly, and the volume control was visible again. The one ergonomic issue I had with the Analog DAC's controls was when I switched inputs in low light: I would invariably also tick the volume knob a bit. It took some skill to push the barely visible input switch and not hit the volume by mistake.

Filter King
The Analog DAC includes MSB's Femto Clock technology, as well as 80-bit digital processing and 384kHz ladder DACs. When I asked MSB's Vince Galbo for some details about the digital filter used in the Analog DAC, he said that even though the DAC IV has several filters to choose from, "while everyone wants to play with these [filters in the DAC IV], they all come to the same conclusion, that one of the default filters is the best. So the default filter is the same in the Analog DAC as that DAC IV series default filter." Which means they're using a custom-designed, linear-phase apodizing filter designed for minimal pre-ringing. Galbo explained that this is "MSB's definition of the term apodizing in that it has a stop band that starts before the Nyquist limit of the source's sample rate (for example, 22.05kHz for 'Red Book'), therefore avoiding aliasing caused by the Nyquist limit."

The Grand Total
Let's talk price. The Analog DAC is MSB's "lower-priced" DAC, but of course that's only relative to their pricy products as noted above. The Analog DAC's base price is $6995, which includes one input module, basic remote control, and the Basic Desktop power supply. This is all some folks will need to get up and running.

You can add the volume control for $995, turning the DAC into a preamp (if you do this, don't forget that it has just that one analog input!). Next, you can add a remote-control upgrade ($85), RS-232 input ($995), or WiFi control ($995). Additional digital inputs cost $995 each (you can add two more). Finally, you can upgrade to the Analog Power Base supply for $2995. The review sample had three inputs, volume control, and Power Base, bringing its total to $11,980. Note: Unlike the other inputs and power supply, which can be upgraded down the line, the volume-control option cannot be added later—it must be ordered with the Analog DAC itself.

First Attempt
I set up the short stack of Analog DAC and Analog Power Base on my cabinet and ran it overnight to settle it in. It didn't get very warm—a balmy 94.5°F was the hottest spot near the display (MSB's Diamond DAC IV ran so hot I couldn't put it in a cabinet)—so I proceeded to set the Sooloos Control 15 on top and fed the MSB via its S/PDIF input. The two products look great together, and the Control 15's smallish base left the Analog DAC's volume control and input switch right where I wanted them.

I cued up a few albums—standard rips from CDs—and settled in for some first-impression listening. Then I cued up some high-definition music. Silence. I restarted the MSB. It powered up, selected the right sample rate (96kHz), and played. No problem. I switched back to a lower sampling rate. No problem. I went to a higher rate and it locked up again.

I e-mailed Vince Galbo, who noted that a dealer had reported the same problem with the Sooloos, as had users of Logitech Transporters. According to Galbo, "some sources do not switch perfectly clean, and the sample-rate transition may contain a bit of noise. Our inputs have a fairly stringent 'window of acceptance,' so to speak." I put the MSB to one side and reviewed some other DACs.

A couple months later, an update to the Analog DAC's firmware became available and I downloaded it from MSB's website. Updating was simple with the Sooloos: I downloaded the WAV file, added it to the Sooloos, and played it through the MSB once. The DAC rebooted, showed the new firmware number on its display, then played a short snippet of music to show that all was well.

You can also update the Analog via MSB's transport, your computer, or by burning the file to a CD. The only requirement, according to MSB, is that playback of the update must be bit-perfect, with no upsampling, volume, or any other filtering added. This update fixed the problem, but there was still one small glitch: Every time the Analog DAC switched to a higher sampling rate, the volume dropped one dB increment. A second update was soon posted and fixed that.

Fast forward to 2013. Taking a gander at his company's website, I think it fair to say that Michael Grace has become a fan of DSD downloads. He writes: "DSD sounds stunningly natural. . . . If you're the type to seek out this kind of fidelity, we can't recommend it enough." This was on the page devoted to the m904's replacement, the m905 ($3495), which supports DSD 128x (5.6448MHz) via the DoP v.1.1 protocol (DSD packed into PCM frames)—via not only its USB input, but also its S/PDIF and AES/EBU (balanced digital) inputs. Perhaps because I've worn out and retired "Gloriosky!," all I could say was "Woober Joobers!"

Still, out of an abundance of caution, I must emphasize that the m905 handles PCM only up to 192kHz; at present, the m905 can't handle DXD at 352.8kHz, or 384kHz program material (such as that available for download from Norwegian record label 2L). Grace Design claims that, as with the m903, and basic visual design aside, the m905 is, compared to its predecessor, close to a clean-sheet-of-paper redesign in which particular attention has been paid to the analog circuitry and to Grace's proprietary phase-locked loop (PLL) reclocking scheme. Therefore, when I asked Grace if I could listen to his smaller product's larger sibling, unlike last time, he fairly bubbled over with enthusiasm. He assured me that even his design team had been unprepared for how much better than the m904 the m905 sounded.

The Grace M905
The m905 is unlike any other consumer-audio component now offered. It took a little getting used to, but it quickly became very addictive. The differences begin with the m905's form factor. The m905's is a rack-mountable component two rack units (3.5") high and 17" wide. Its front panel is empty, except for a headphone jack, an illuminated power button, their silk-screened icons, and the names of the product and its manufacturer.

On the m905's rear panel are, by my count, 33 connectors and switches. These include balanced and single-ended analog inputs (the m905 handles analog signals entirely in that domain), six digital inputs, two digital outputs, Word Clock throughput, Talkback circuit In/Out, three sets of Monitor analog outputs, Cue analog outputs, and Multipurpose analog outputs. Michael Grace states that the proprietary, high-precision S•Lock PLL digital-clock regeneration used in the m905 is a significant improvement over that used in the m903. The m905's USB connection is the preferred asynchronous type.

Except for the power switch, all of the m905's controls are found on a remote control about the size of an iPad mini (8.9" wide by 2.2" high by 5.1" deep) but thicker and heavier. For a variety of good and sufficient reasons, the remote is connected to the mainframe with a multi-conductor cable with 15-pin D-sub connectors at either end (these have nonstandard pin assignments). The stock cable is 25' long because, in a recording or mastering studio, the mainframe will often be bolted into an equipment rack some distance from the mastering position. Grace will provide a shorter cable at no additional charge if the request is made at the time of the original order. They're also exploring the practical possibility of controlling the m905 via an iPhone/iPad app. The remote comes with a detachable base, from which the control panel can be tilted up about 30°.

The remote's front panel is dominated by, on the right, a large, finely machined volume knob and multifunction switch, and on the left by a color graphical LCD display (not a touchscreen) with adjustable brightness. The screen's image is divided into left and right halves, each of which has a prominent volume-setting readout: for the selected monitor speakers (left) and the headphone outputs (right). Smaller readouts cover such system information as the selected monitors, whether or not a subwoofer is engaged, and the digital input signal's sampling rate, bit depth, and lock status.

For me, the most fascinating readout on the LCD screen was the real-time display of sound-pressure level at the mastering position, courtesy a microphone built into the remote (!). The particulars of SPL metering can be selected via the Setup menu, including canceling the SPL display. Note that the remote displays SPL, not loudness. Professional loudness meters cost as much as or more than the entire m905.

If you have separate control and tracking rooms, the same microphone that feeds the SPL display can be used for Talkback. Even if you don't, it can be fun to hold down the large Talkback button (to the right of the Volume knob), watch it light up red, and tell Ella what a fine job she's just done.

Grace Design claims that the remote's design replicates the solid feel of traditional hardware, while adding the greater information and functionality of a graphical user interface. I applaud them for putting all of the m905's important functions under mechanical control. I think that a first-class rotary volume control is far more intuitive and precise, and more likely to hold up under heavy use, than a piece of glass you slide your fingertip over. And the remote and its base have a textured, fingerprint-resistant silver finish that fairly shouts Authentic German Engineering—even though the m905 and all Grace products are designed and built in Lyons, Colorado (footnote 1).

The m905's remote offers three forms of mechanical control. The speed-sensitive volume knob toggles at a tap between controlling (in 0.5dB steps) the monitor speakers selected and the two headphone jacks (one on the mainframe, one at the rear edge of the remote). Eight small, black buttons, four each above and below the LCD screen, select the active input. Illuminated pushbuttons on either side of and above the volume knob select among three possible monitor-speaker outputs; enable frequently used control functions such as solo, subwoofer control, sum-to-mono; and diminish the output by 20dB or mute it entirely.

All of that might seem daunting; even more daunting is the very complete owner's manual. However, setting up and getting sound out of the m905 was not difficult at all. The remote's cable connects to the mainframe's rear panel. Connect a digital source (in my case, Parasound's excellent Halo CD 1) and connect balanced cables to the mainframe's analog outputs. Those balanced cables connect to your amplifier, with adapters if need be. Grace provides a sturdy power cord. Press the power button on the mainframe's front panel; it should softly glow a comforting light green.

As part of its "Please allow us to make your life as foolproof as possible" mission statement, every time the m905 is powered up, it resets the volume controls for both the monitor speakers and the headphones to "0." Yay! That'll prevent some "whoopsie" moments, especially if the previous night's work got a little loud. And so you don't have to crank the volume control all the way up to a usable level every time, in the Setup menu you can arrange it so that when you hold down the volume control, the output level moves to your predetermined level (in my case, 70dB).

After setup, all you have to do is select a source and rotate the volume knob (or press and hold, if you've entered a default level) until you get sound. The silky-smooth feel of the volume knob and its silky-smooth detented action strike me as more of those little touches that indicate that Michael Grace was strongly influenced by the time he spent at the Jeff Rowland Design Group. Grace's products not only represent exceptional value for money; they also manage to make a lot of workaday professional components look like mud fences.

The m905, as described above, is pretty much plug'n'play right out of the box. However, what I've described is the simplest, plainest-vanilla setup. There is a wealth of options for customizing setup and use. The two that will likely make the biggest differences for audiophiles are the abilities to rename the input-selection buttons' legends on the LCD display and to individually adjust the inputs' levels.

Features that home audiophiles aren't likely ever to need include eight channels of optical ADAT ins and outs (if you don't know what ADAT is, you absolutely don't have to worry about it), and up to 10 channels of computer audio via USB. Those are for the serious home-studio musician, or a professional with a digital audio workstation (DAW) program such as Pro Tools.

While the m905 lacks a physical balance control, the balance can be adjusted on each loudspeaker pair's Setup screen. There is no polarity-invert function, owing to the added complication of implementing that for headphones. However, signal polarity can often be inverted at the digital source, or in DAW software. Like the m903, the m905 offers selectable headphone cross-feed.

There's not space here to get as deeply as I might wish into the m905's inner workings. Suffice it to say that the m905 is based on cutting-edge technology, seamlessly implemented. Its DAC chip is a Burr-Brown PCM1798 whose interpolation filter 8x-oversamples an S/PDIF input from a "Red Book" source. (Sample rates higher than "Red Book"'s 44.1kHz are oversampled 4x or 2x.) The delta/sigma modulator then converts the 352.8kHz PCM signal to what Michael Grace believes is a 6-bit/2.88MHz signal. (Texas Instruments' current data sheet for the PCM1798 does not specify bit depth.)