ExaSound Audio Design is a Canadian company that designs and domestically manufactures two DACs: an e18 multichannel model and the recently released $2499 two- channel e20, the subject of this review. So just what distinguishes the e20 from the veritable flood of new DACs showing up on the market these days? In my view, the e20’s most distinctive feature is its ability to play Direct Stream Digital (DSD) files in their native format without converting them to PCM data first.
That’s a new capability recently developed by industry heavyweights dCS, Playback Designs, and Channel D to handle the DSD format. If you’re unfamiliar with DSD recording, it was the system originally designed to record music for SACDs. SACDs don’t use raw DSD files; they add a copy-protection scheme that’s been difficult, though not impossible, to crack. If you’re unfamiliar with DSD files, they usually come with file name extensions of .DSF or .DFF. For example, a DSD file could be named “Bolero.DFF.” The e20 doesn’t just play standard 2.82MHz DSD files; it also plays the 5.64MHz double- speed DSD files, which some recorders are capable of making.
“And why,” you may ask, “should I care about playing back DSD files? Do they really sound any better than 192kHz/24- bit PCM files?” They should, since DSD DACs don’t have to use the very complex oversampling digital filter needed for PCM conversion. Although DSD DACs do use a filter, it’s very simple, not the sonically damaging brickwall filter needed for PCM. Another important question to ask if you’re considering a DSD-capable DAC is: “Are there any DSD files commercially available to play?” And the answer, at least as of Winter 2012, is not many. I don’t have a complete list of sources for DSD files; two that I’ve encountered are Channel Classics (channelclassics. com) and Blue Coast Records (bluecoastrecords.com). Both sources charge higher prices for DSD files than for high- resolution PCM files, but that’s hardly surprising, since it always costs more to pursue the bleeding edge of technology. If the demand for DSD files continues to increase, competition among record companies is bound to drive prices down.
The e20 also plays Digital eXtreme Definition (DXD) files, which are 352.8kHz/24-bit PCM files used in professional recording studios. DXD files are also hard to find, but a few are available from highresaudio.com. ExaSound’s Web site includes a page dedicated to DXD files from highresaudio.
For the hardware junkies, inside the e20 is an ES9018 Sabre32 DAC chip from ESS Technologies, which accepts up to 384/32 PCM files. As I just noted, there aren’t currently many companies selling files that advanced, but at least you can consider this capability as a sort of future-proofing. Software upsampling to 384kHz is also possible. Three precision oscillators are used to minimize jitter, with a claimed ultra-low 0.13 picosecond master-clock error. Eleven power-filtering stages are used to reduce noise, jitter, and channel crosstalk. The 128dB signal-to-noise ratio illustrates how effective these measures are.
Like most modern DACs, the e20 uses an asynchronous USB interface to minimize jitter. The USB receiver is galvanically isolated from the DAC circuits to eliminate noise or ground loops from your computer. The e20 supports automatic sample- rate switching if your source does. There are custom Audio Stream Input/Output (ASIO) drivers for Windows and high- performance Macintosh OSX drivers, which support PCM at sampling rates up to 32bit/384kHz and DSD at 2.82 or 5.64MHz.
As a frequent headphone listener, I was tickled to see that the e20 had a headphone output jack. ExaSound’s Web site claims the e20’s headphone amplifier is “capable of driving the most demanding headphones.” With an output impedance of only 1 ohm, the e20 should be able to drive low-impedance headphones like those from Grado. Unlike some headphone amps, the line output is not muted when you plug in headphones. The volume control on the e20’s front panel controls both headphone level and line level.
Compared to full-size DACs, the silver (black is not available) aluminum e20 is svelte. Its 6.5″ x 2.2″ x 7.9″ chassis won’t take up much space on your equipment rack, nor will its 1.653 pounds put much strain on the shelf. The low weight results in part from an external in-line power supply. The e20 is well-built and attractive, but probably wouldn’t be characterized as audio jewelry, which makes sense to me; you’d probably pay twice as much for a full-size chassis with a thick faceplate.
On the front panel are five silver buttons, a window showing what the e20 is doing, and the 1⁄4″ headphone jack. The buttons are (from left to right) the power switch, a set-up menu, the input selector, and volume up and down. The volume control works on both the headphone amplifier and the e20’s outputs. The input selector switches among the three digital inputs: USB, S/PDIF on coaxial, and S/PDIF on TosLink (optical).
The status window has a two-line blue display. While easily readable from directly in front of the e20, there’s no way you could read the display from your listening chair across the room. (Well, I guess you could use binoculars.) The display’s top line tells you which input is selected, the type of signal (PCM or DSD, with automatic switching) coming into the input, and the sampling rate playing through that input. The bottom line tells you the input is 2-channel and shows you the volume level. When you first turn the DAC on, the volume level is set at -40dB, to prevent overloading your amplifier and speakers (and ears). You can adjust the volume setting in 0.5dB increments.
A poorly implemented digital volume control can degrade sound quality, so I asked George Klissarov, President of exaSound Audio Design, to describe the e20’s: “The volume control of the e20 is implemented by the ES9018 DAC chip. It is at the borderline between the analog and the digital domains. We send volume commands to the chip, and it creates lower voltage on the analog side. This way there is no loss of bit-depth, and we avoid the usual issues with noise when analog volume controls are used. This solution provides the ultimate flexibility.” Bottom line: The e20’s volume control won’t degrade your sound.
The back panel of the e20 is small, so it’s chock-full of input and output jacks. Besides the three digital inputs, there’s an input jack for power from the in-line power supply. Signal output is available on both balanced (XLR) and unbalanced (RCA) jacks.
A small toggle-switch lets you isolate the ground from the computer attached to the USB input. Normally in the isolated position, the switch may be set to connect the ground if there are hum or noise problems. To save space, the USB input jack is a mini-USB jack, rather than the usual USB Type B jack found on most DACs. That means many USB cables won’t fit. Fortunately, exaSound provides a USB cable with a mini-A USB connector on one end and a Type A connector (the type that goes into your computer) on the other. Such cables should be available at your local computer store, or if you order a cable from a cable manufacturer, there should be no problem getting one with the mini-USB plug on one end.
The simple remote control duplicates the e20’s input-switching and volume-control functions. That’s enough to make it feasible to use the e20 as a system preamp, if you only have digital sources. The remote looks a little like a Classic iPod, with a circular dial containing most of the controls. That’s not surprising, since it’s an Apple control. You can reprogram the e20 to use other IR remotes, if that would be useful. The remote controls volume, input, and muting.
The e20 is sold direct with a short one-year warranty, but has a thirty-day return policy (so you can try it out before you decide to keep it). For those, like me, who lament the demise of so many brick-and-mortar dealerships, a return policy like this seems like the next best thing to a dealer loan.
Setting Up and Using the e20
I placed the e20 on a shelf on my Billy Bags equipment rack and connected it to my linestage with Audience Au24 e balanced interconnects. Thanks to the e20’s compact dimensions, I had room on the same shelf for my Auraliti music player and my Sony XDR-F1HD tuner. I don’t think most people want huge audio/video racks in their listening rooms these days, and compact components like the e20 are one way to minimize shelf space requirements. I connected my Auraliti-based server to the e20 using a WireWorld Gold Starlight 6 S/PDIF cable, and my computer-based server using an AudioQuest Diamond USB cable. The e20’s power supply uses an unpolarized C7 connector instead of an IEC jack, so I used the rather flimsy-looking stock power cord.
Recommended break-in time was 50 hours, rather short for any component. So I inserted the e20 in my break-in system and gave it 200 hours through the S/PDIF coaxial input and another 200 hours through the USB input. A load was connected to the unbalanced output jacks to assure the analog output section was properly broken in. I also plugged in some headphones during the USB break-in, so the headphone amplifier section would also get broken in. If you think reviewing is glamorous fun, I hate to tell you that most of this reviewer’s time is spent breaking equipment in. I usually don’t do much listening during break- in, rendering moot the assertion that break-in really amounts to getting used to how a component sounds.
One of the tedious aspects of using a computer as your music server is adjusting settings on the server software to produce the best sound quality. ExaSound helps you do that with very detailed set-up instructions in its manual for the J. River software. I’m a Windows guy, so I use these Windows programs. I used Foobar2000 version 1.1.15 server software, which exaSound recommended for its stability along with J. River. ExaSound points out that J. River is the company’s first choice for its feature set, user interface, and customer service. Foobar2000 is simpler to use, but both offer excellent sound quality. ExaSound also has a very thorough set-up guide for Foobar2000 software available on its Web site, and I found it essential, since the set-up process was rather complex.
I printed out the guide to use at my computer, but due to the illegibility of one captured screen, I skipped part of a step, with the result that DSD files wouldn’t play. Klissarov patiently walked me through the setup, and we implemented the necessary setting, after which both DFF and DSF files played just fine. Although the e20 will automatically switch between DSD and PCM files, Foobar won’t; you have to go into Foobar’s File/ Preferences menu and change some settings. It’s not hard, but neither is it intuitive, even if you’re familiar with Foobar’s PCM playback settings.
I couldn’t really test exaSound’s claim that the e20 is “capable of driving the most demanding headphones.” Probably the most demanding headphones available today are HiFiMAN’s HE-6s, which need around five watts, but since I don’t have a set of those cans, I had to be satisfied using the higher-sensitivity (and far cheaper) HiFiMAN HE-400s. I also used more conventional stock Sennheiser HD 650 and AKG K701 headphones, which are easier to drive.
To get my footing, I started listening via the USB input to some of the PCM files I normally use for reviews. My first impression was that the e20 sounded very clean, with a slight high-frequency emphasis. It wasn’t at all peaky, but had a bit of extra sparkle. Jennifer Warnes’ recording of “The Panther,” from her CD The Well [SDR], opens with a wide assortment of percussion instruments, and the e20 portrayed their highest frequencies with considerable extension and detail.
On Jordi Savall’s CD La Folia 1490-1701 [Alia Vox], the piece “Folia Rodrigo Martinez” sounded very energetic, again with excellent high-frequency detail in the percussion instruments, including the opening cascabels. The bass drum, which extends surprisingly deep for an early music piece, had lots of impact. I’ve heard deeper bass extension, though not much.
One of the first high-resolution albums I downloaded was Seventeenth Century Music and Dance from the Viennese Court by Ars Antiqua Austria, conducted by Gunnar Letzbor (96/24 FLAC, Chesky/HDtracks). Initially, with other gear, I thought it sounded sweet but a bit bland and flat. However, with the e20, the string tone glowed with harmonic richness, and the dynamics sprung to life, infusing the album with energy and flow.
With some a cappella choral music, The Tallis Scholars’ Allegri Miserere (96/24 FLAC, Gimell), the e20 made the soundstage appear deeper than I’m used to. This recording features a main choral group at the front of the soundstage, with a smaller solo group some distance back in the church where the recording was made. Through the e20, the solo group seemed farther back than usual, although still easily understandable. Reverberant room noise from the solo group seemed somewhat reduced.
While USB connections are needed to play DSD files, perhaps you just want a moderately priced DAC to play back your server’s S/PDIF output. I thought the e20’s S/PDIF input sounded just a bit less extended in the high frequencies on the percussion instruments in “The Panther.” And the bass on “Folia Rodrigo Martinez” had very slightly less impact, although the e20 was scarcely bass-shy. Does this mean the USB input was “better” than the S/PDIF input? Well, no; remember, different sources were used and different cables connected those sources to the e20, so it’s impossible to pinpoint which component may have caused the differences, which were very slight.
So finally to the question that probably interests you most: How did the e20 sound with DSD native files? To evaluate the e20’s performance with DSD files, I selected Channel Classics’ album Super Artists on Super Audio Sampler Vol.5 and downloaded both the 192/24 FLAC version (2406.9MB) and the original DSD (DFF format, 2956.4MB) version. I listened first to several FLAC files, beginning with Dejan Lazic playing the “Allegro” from Scarlatti’s Sonata in C major. I remember thinking how clean and detailed this recording of a piano sounded. Then, hoping I would be able to detect the difference, I cued up the DSD version of the same recording. Holy flaming cow! The DSD version of the recording made the FLAC version sound flat and mechanical. With DSD, it sounded like a different piano! The DSD version had the relaxed sound I associate with analog playback. In comparison, the FLAC piano sounded bleached and harmonically threadbare; and for some reason, the DSD version of the piece also sounded distinctly more dynamic. Reach-out-and-touch-it textures further increased the impression that, with the DSD recording, I was listening to a real piano. With the FLAC file, the piano notes just splatted into existence, while the DSD file’s piano notes sounded spookily real, beginning with the hammer hitting a string, the note launching into space, and then decaying off into silence as the note ended.
Next, I listened to The Netherlands Bach Society performing the “Quoniam tu solus Sanctus” from Bach’s Mass in B Minor. The same phenomenon occurred: The DSD recording was more realistic, easier to understand, and more dynamic. When the section transitioned to the “Cum Sancto Spiritu,” the way the soundfield opened up and the effortless increase in dynamic power was just thrilling. I could cite other musical examples, but they would all lead to the same conclusion: The DSD version of the recordings sounded superior to the PCM version—and the difference was not subtle.
Thinking that my limited assortment of DSD files may have restricted my findings, I tried some sample DSD files from Norwegian recording company 2L. But here, the superiority of the DSD versions of recordings over PCM versions was even more pronounced. I sampled a recording of the “Allegro” from Mozart’s Violin Concerto in D Major, played by Marianne Thorsen and the TrondheimSolistene, led by Øyvind Gimse. The 192/24 FLAC version sounded quite rich and detailed, but the DSD version just turned up the reality meter several notches. I remember wondering briefly if the soloist was playing a different violin in the DSD version, one that sounded gorgeous. Of course, it was the same violin on both versions. Who said digital can’t reproduce sweet string tone? The spectacular 2L recordings aren’t yet commercially available in DSD format; I hope that changes soon.
I have always regarded my AKG K701 headphones as a benchmark for high-frequency linearity, but the e20’s headphone amplifier gave them a high-frequency rise that sounded a bit bright. With headphones having a bit more roll-off (the Sennheiser HD 650s and, even more so, the HiFiMAN HE- 400s) the e20’s HF response sounded much flatter. And the e20s produced deep, very well-defined bass, quite spectacular through the HE-400. The combination of the e20 and HE-400s became a favorite, sounding sweet and detailed, but not the least bit bright.
My Audio Research DAC8 cost twice as much ($4995) as the e20, and is housed in a full-sized chassis with a sturdy but not terribly thick faceplate. As with most Audio Research components, the DAC8 is available in either black or silver finishes. Like the e20, the DAC8 is totally solid-state, but nevertheless sounds sweet and beautiful and dead quiet. Since it was brought to market before the DSD playback standard was developed, it doesn’t play DSD files; however, it plays PCM files up to 192/24 through all inputs except TosLink. Both ASIO and Windows Audio Session Application Programming Interface (WASAPI) drivers are available for the asynchronous USB input. I used the USB input with ASIO drivers.
Some quick sonic impressions: on “Folia Rodrigo Martinez,” bass extended just a tad deeper, and the instruments had a slightly richer harmonic presentation. Jennifer Warnes’ “The Panther” had excellent high-frequency extension, but the e20’s extended just slightly higher. On “Allegri Miserere,” the front choral group was just as well defined through the ARC, and I could hear more room reverberation from the distant solo group, which helped placed them farther back in the recording venue.
When I played the Scarlatti “Allegro” through the DAC8, the FLAC version sounded a lot more like the DSD version did through the e20. The piano’s notes sounded more fully developed, with a hammer strike, launch of the notes, and decay that sounded much more like a real piano. And dynamics were substantially stronger, much closer to the DSD recording through the e20. Unfortunately, the Audio Research won’t play DSD files.
Is it fair to compare the e20 to a DAC that costs twice as much? Maybe not, but it shows that as good as the e20 is, it’s possible for PCM files to sound even better; all it takes is lots more money.
This review addressed two issues: 1) how the exaSound e20 sounds, and 2) how DSD music files sound. For me, listening to DSD files was a new experience, which the e20 made possible. And what an experience it was! Based on the limited number of DSD files I had available, DSD files are quite spectacular. I know lots of readers find it amusing when reviewers rave about new equipment, but for me DSD files made all the PCM files I’ve heard, even the highest-resolution PCM files, sound, well, digital. In a world where the highest accolade for digital gear is how much it sounds like analog, DSD files are off the scale. I’m going to have to think long and hard about whether I’ll ever download another high-resolution PCM file. A listening buddy and hi- fi dealer who has so far been a staunch analog holdout was so thoroughly impressed that he went away planning to implement DSD playback in his store. It’s too early to tell if DSD files will become widely available, but I hope so. Pricing them at the same level as 192/24 files might greatly speed up their acceptance.
For me, the exaSound e20 was a game-changer. It played standard PCM files extremely capably, but its greatest strength was its ability to play DSD files in their native format, which took my system to a new and higher level. The e20 is an attractive, compact, and clearly thoughtfully-designed example of a modern DAC. It produced a rich detailed sonic picture, with solid bass and extended high frequencies. I heartily recommend that if you’re in the market for a new DAC, you consider one with DSD-playback capability, like the exaSound e20.
SPECS & PRICING
Inputs: S/PDIF via RCA and TosLink, USB mini
Sample rates supported: PCM-44.k, 48k, 88.2k, 96k, 176.4k, 192k, 352.8k, 384k up to 32 bits; DSD at 2.82M and 5.64M; TosLink is limited to a maximum 96k, RCA S/PDIF to a maximum 192k
Analog output level: Unbalanced via RCA jacks, 2V RMS; balanced via XLR jacks, 4V RMS
Line output impedance: 200 ohms
Headphone output impedance: 1 ohm
Headphone output current: 250mA peak
Dimensions: 6.5″ x 2.2″ x 7.9″
Weight: 1.653 lbs.
exaSound Audio Design
3219 Yonge Street, Suite 354
Canada M4N 3S1
Speakers: Affirm Audio Lumination speakers
Amplifier: Audio Research VS 115 stereo amplifier
Preamplifier: Audio Research LS27 linestage
Sources: Hewlett Packard dv7-3188cl laptop computer running 64-bit Windows 7 Home Premium and Foobar2000 music server software version 1.1.15, Auraliti PK100 music player, Audio Research DAC8
Interconnects: Clarity Cables Organic interconnects, Audience Au24e balanced interconnects,
Speaker cables: Clarity Cables Organic loudspeaker cables
Power cords: Purist Audio Design Venustas power cords, Clarity Cables Vortex power cords, Audience powerChord e power cords
Digital: Wireworld Gold Starlight 6 S/PDIF cable and Audioquest Diamond USB cable