Music in the Round

Before receiving the Alpha 3, I looked on PassMark Software's Cpubenchmark.com site to find test results for the Alpha 3's six-core i7-8700 processor, and for the four-core i7-7700K processor in my Baetis ProdigyX server, and found them to be respectively 15,156 and 12,045—a more than 25% increase in overall speed in favor of the Alpha 3. Later, I ran the JMark tests built into JRiver Media Center and was surprised to discover that the four-core i7-7700K was marginally faster in its overall JMark score, and in the subcategories of image and database processing, but that the six-core i7-8700 was much faster in the critical category of math processing. In practice, there was little to choose between them: both tackled my test task of playing multichannel DSD256 files, converting them on the fly to 24-bit/192kHz PCM, applying Dirac's Live room EQ, and adding bass management—all without pops, ticks, or signal interruptions. The only difference was that CPU usage was consistently lower with the Alpha 3's i7-8700, indicating that it had more headroom, even when doing the toughest tasks.

The Flux Capacitor's USB output was subtly but consistently superior to the motherboard's USB outputs in terms of increased precision. There was greater apparent detail from the treble to the bass, while the overall sound seemed more relaxed and less aggressive.

Overall, I was impressed with the sound and performance of Wolf Audio Systems' Alpha 3 HFAS+ Apex. It may seem to be an off-the-shelf PC, but its secret sauce is a mix of Wolf's optimization of its physical components, skillful and clean construction, and integration of software applications—and I think that combination is beyond what most aficionados of multichannel sound can achieve on their own. In addition, Wolf's live/VPN tech support ensures that even a newcomer will be able to get this system up and running quickly and flawlessly. As a package, the Alpha 3 is a great choice for playing hi-rez files of multichannel music.

NAD Masters Series M17 V2 surround-sound preamplifier-processor
In my review of NAD's Masters Series M17 surround-sound preamplifier-processor ($5499) in the January 2015 issue, I wrote that I was "extremely pleased" by its performance. I hoped to hold on to the review sample to see if NAD, taking further advantage of their Modular Design Construction (MDC), would add new capabilities to it. Well, it turned out I didn't keep it—but now NAD has released the M17 V2 ($5999., with facilities claimed to enhance both sound and image.

I refer you to my review of the M17 for a description of the appearance and operation of the M17 V2—which, from the front panel, looks identical to the original model. Also retained is the soft-touch On/Standby switch, with which I had problems before; either it's been improved or I've become more adept, but I now find it delightfully easy to use. The M17 V2's front-panel display and interface remain models for all competitors in terms of visibility, clarity, and simplicity, except for a continued insistence on the use of the right shift instead of Enter to advance through the menu structure. But the regular user will readily adapt—this is a problem only for those of us who switch willy-nilly among AVRs and pre-pros.

It's become common for manufacturers not to include a printed user manual, but while NAD did include docs—a Quick Setup Guide, Internet Update Guidelines, a BluOS Kit Quick Start Guide, and for installing Dirac Live—there's no electronic copy of the User's Manual, or a link to it on the elegant, leather-sheathed USB drive that's also included. Of course, the manual can be downloaded from NAD's website, and you can easily get the M17 V2 up and running with the included guides—assuming you have only HDMI sources, an HDMI display (footnote 2), and a subwoofer!

Such carping aside, all of the M17 V2's audio inputs and outputs are clearly labeled, whether you're using the RCA or XLR jacks. The new AM17 board adds four more channels to the basic 7.1 for Atmos, in the form of mini-XLRs (adapters to standard XLR are provided). Inside, the original M17's Audyssey XT32 digital room correction (footnote 3) is replaced with Dirac Live. Another I/O enhancement is in the implementation of BluOS. The original M17 required some fussy handshaking by WiFi/Bluetooth to implement network connections between the NAD and other network devices. But while the M17 V2 comes with a little USB hub, cord, WiFi dongle, and Bluetooth adapter, I didn't need them. As it's my practice to attach an Ethernet LAN cable to any device that has one, I simply selected BluOS (a preinstalled source on the M17 V2), directed it to my NAS, and began selecting files from my library. The icing on the cake was downloading the BluOS app to my iPad and iPhone, to let me also control BluOS from those devices. Immediately, I was streaming files of resolutions up to 24/192.

Back to the main event: multichannel sound. The only way to get discrete multichannel into the M17 V2 is via HDMI. There are no discrete analog multichannel inputs, and the USB input, though it can handle the digital input from a microphone for Dirac Live calibration, is not set up to accept any other audio source, stereo or multichannel. I'm not specifically targeting NAD with this criticism; with rare exceptions, no multichannel AVRs or pre-pros accept multichannel audio via USB. I connected my Mac mini server (running Windows 7 and JRiver under BootCamp) for file playback, and my Oppo UDP-205 for disc playback, to two of the M17 V2's HDMI inputs. In both cases, there were no problems with any PCM source up to 24/192.

I set about running Dirac Live (footnote 4). For those who've done this before and/or who've read about using Dirac Live, the procedure is unchanged. I won't repeat what's been published, and will say only that I followed the onscreen instructions. NAD permits the use of either the supplied microphone or any other calibrated USB mike. I used the same miniDSP UMIK-1 I've used before with my Windows 10 laptop; a mobile app for Dirac Live was promised for sometime in late 2018.

Because I've measured the same speakers—three Monitor Audio Silver 8s in front, two Silver 2s in the rear corners, and a mixed trio of subwoofers—in the same room with the same mike so many times, the results were unremarkable. To correct the familiar foibles of my system and room, I accepted Dirac Live's default curve, which has served me well: It has a mild LF boost that slopes slowly down through the rest of the spectrum. However, the M17 V2 has three virtual filter storage slots; after many trials, I ended up with three correction filters: 1) full-range correction for all speakers, 2) correction below 300Hz for all speakers, and 3) correction below the zero-crossing point nearest the nominal 500Hz woofer/mid crossover for the L/C/R speakers, and full-range correction for the rear speakers.

What a delightful and satisfying result! Filter 1 did all the right things, but with a bit too detailed upper MF and HF. Filter 2 dealt only with sounds below the room's Schroeder frequency, but there was a lack of presence. Filter 3 was just right: lively but open, and remarkably free of room coloration. You need only a single good set of measurements to experiment with target curves and windows and, hopefully, be led to the kind of felicitous results I found.

I liked the M17 V2's sound through Filter 3 more than I did anything I ever heard from the Marantz AV8805 with the Audyssey app, even though Dirac Live limits the resolution to 48 (footnote 5). Unfortunately, I have not yet been able to apply Filter 3's target parameters to Audyssey. On the other hand, I did run Dirac Live on my server with those parameters, and with a stack of three Mytek Brooklyn multichannel DACs through the Marantz's analog pass-through. It's both a more cumbersome and a more flexible arrangement than just using the NAD, but in terms of sound quality, it was even better.

NAD's Masters Series M17 V2 is a superb-sounding pre-pro and DAC right out of the box, but Dirac Live elevates it to something special. Instrumental timbres seemed more realistic, soundstages were nearly holographic, and my listening sessions grew longer from sheer musical enjoyment. Of course, with the M17 V2, I no longer needed to run Dirac Live on my server; but if, somehow, NAD could empower BluOS to support multichannel files at higher resolutions, I'd no longer need even the server. Heck, for that, I'd even buy a new plasma display!

---

Footnote 2: My screen is a fairly ancient Fujitsu P50XHA40U plasma display whose maximum resolution is only 1080i. So while real video is accepted and displayed just fine, I couldn't get the M17 V2's onscreen display (OSD) to appear. As I learned, the "NAD OSD only supports 1080p at 24/50/60Hz frame rates." NAD's otherwise excellent app is not as comprehensive for system setup as the OSD, and I had to resort to using an Oppo UDP-103 as a video converter.

Footnote 3: Although Audyssey is not included in the M17 V2, my sample still bore the Audyssey logo; and on my iPad, under "DSP Options," NAD's very useful control app also lists Audyssey.

Footnote 4: In June 2018, when the M17 V2 was delivered, a bug in its implementation of Dirac Live routed some test signals to the wrong speaker channels; NAD quickly provided a patch via download. By the time you read this, it's likely the patch will have been incorporated into production units of the M17 V2 for shipping. This experience underlines NAD's and Dirac's admonitions to update all firmware before investing time and effort in the EQ calibration.

Footnote 5: The M17 V2 supports Dirac Live in stereo and multichannel, but while the pre-pro itself supports resolutions up to 24/192, limitations are imposed when Dirac Live is engaged: "When the output is more than 2 channels the Dirac filtering is limited to 48k. With 2 channel output the filtering is limited to 96k. (Note that the sub counts as an output channel so if you have the sub enabled you will be limited to 48k.)"