NAD Masters Series M32 DirectDigital D/A integrated amplifier Page 2

After years playing musical instruments, I have finally gotten adept at tapping three beats in the bar with one hand against four in the other; on a good day with the wind behind me, I can even manage five against four. But most days, the tapping quickly degenerates into an arrhythmic mess. Similarly, good amplifiers and speakers can preserve polyrhythmic intricacies in music after lesser products have stopped making musical sense. The underlying groove of "Ascension Day," from Talk Talk's Laughing Stock (16/44.1 rip from UK CD, Verve 847-717-2), was recorded in a fairly reverberant space with the mikes quite a long way from the drums. The rhythmic pulse can get obscured with amplifiers that can't keep control of the speakers. (This is what I believe Art Dudley is talking about when he refers to the "force" of loudspeaker sound.)

Stereo imaging was superb—stable, precise, and accurate on my own recordings, such as "The Mooche," from the Jerome Harris Quintet's Rendezvous (16/44.1 WAV master file for Editor's Choice, Stereophile STPH016-2). But perhaps image depth was not quite as well developed as it is with my current reference D/A and amplifier combination: the PS Audio DAC feeding Lamm M1.2 Reference monoblocks, and a little less so with the M50.2 via Ethernet than it was via AES/EBU.

So far, I'd gotten my impressions of the M32's sound quality by listening through speakers with which it might be paired out there in the real world: KEF LS50s, TAD ME1s, Sonus Faber Guarneri Traditions, and GoldenEar Technology Triton References. But toward the end of my listening I installed in my room a pair of Wilson Alexia 2 speakers, for review in our July issue. The Alexia 2 costs $57,900/pair—it's hard to conceive of anyone using a pair with the NAD M32, but I did anyway. The Alexia 2 is an easier load than the original Alexia, which dropped to below 2 ohms in the midbass. The new speaker remains between 2.6 and 6 ohms for much of the bass and midrange, and averages 8 ohms in the treble. I set the M32's speaker matching to "4 ohms."


Well now. What I heard was not quite what I'd expected. Yes, the low frequencies in the Talk Talk track were less well-controlled, the bass guitar and kick drum sounding more like what you'd hear in a live concert than on a studio recording. But the treble remained clean and clear, and the imaging was good by any standard. I spent an enjoyable morning surfing my music library with Roon using this unlikely combination of amplifier and speakers.

The amplifiers that had shared my listening room with the NAD M32 were the pair of Lamm Industries M1.2 Reference monoblocks ($27,790/pair) I'd reviewed in April 2012, and that had returned to Brooklyn after a sojourn in Jim Austin's system. The Lamms are thoroughbreds, producing a large, rich-sounding sweep of sound, especially when driven directly by the PS Audio DirectStream DAC. I replaced them with the M32 and fed it digital data from NAD's own M50.2, with levels matched at 1kHz. The soundstage was smaller than the Lamms'—which is perhaps a bit too generous—and had less "glow." But after playing several recordings using Roon's "Radio" function, which randomly selects from the user's library tracks similar to the first one played, I realized that I wasn't enjoying the music any less. The NAD's tonal balance was lighter than the big Lamms', but its clean, clear signature let through a wealth of recorded detail. While I wouldn't suggest that owners of a pair of M1.2 References and an outboard DAC replace them with an M32, the NAD wasn't embarrassed by the comparison.

A more relevant comparison would be with another integrated amplifier, but integrateds rarely visit my listening room. As chance would have it, I still had to hand a Luxman L-509X ($9495), which I'd been measuring to accompany Ken Micallef's review of it elsewhere in this issue. A contrast to the all-digital M32, the Luxman is a traditional integrated amplifier, with no digital inputs, no digital signal processing, and no networking ability—even its two large, analog VU meters look like holdovers from an earlier age, compared with the NAD's four-color touchscreen.

I spent a week living with the Luxman, driving its balanced inputs from the PS Audio DAC. KM, whose listening to the Luxman had primarily been with vinyl, had concluded that "The L-509X is one of the most intimate-sounding, dynamic, texturally nuanced, truthful purveyors of music of my experience." The Luxman is certainly a great-sounding amplifier, but its soundstage was narrower and shallower than that thrown by the Lamm monoblocks, even with the KEF LS50s, which are imaging champs. It also had more difficulty driving the Wilson Alexia 2 speakers than I'd expected from its measured performance, the dynamics sounding somewhat restricted.


I replaced the Luxman with the NAD, fed it AES/EBU data, and again listened with levels matched at 1kHz. The M32's low frequencies were well defined, but with not quite the weight or extension offered by the Luxman. At the other end of the audioband, the M32's highs seemed to have more "air" than the L-509X. This was a benefit with After the Fall, a live recording from 1998 of Keith Jarrett's Standards Trio of bassist Gary Peacock and drummer Jack DeJohnette (16/44.1 WAV files, ECM 2590/91), reviewed by Richard Lehnert elsewhere in this issue. This album has a somewhat closed-in balance, the piano sitting atop the double bass and drums, but the NAD added some useful top-octave life to DeJohnette's cymbals.

I noted the same difference with the Linn recording of Monteverdi's Vespers 1610. Philip Hobbs has done his usual excellent job of capturing a natural-sounding balance of performers and performing space, and the M32 presented that space with a better sense of depth than the Luxman had managed. This was not what I'd expected, given that the combination of the Luxman and the PS Audio DAC costs $15,494, not including cables, and the NAD only $3999. But to send AES/EBU data to the NAD I was using the Meridian 518, which cost $1650 when last available. So I reverted to the long plastic TosLink connection from the M50.2. The sweetness of the M32's mid-treble was slightly diminished, but other than a slight reduction in depth, the excellent soundstaging remained intact.

Looking back at my "Listening" comments, I see that I neglected to mention that the NAD M32 and KEF LS50s formed a synergistic relationship. The M32, however, will give great sound with more expensive loudspeakers. For $4398 with the BluOS module, this amplifier really does punch above its price class. The other one-box solutions we've recently reviewed—Naim's Uniti Nova ($6995), in the March issue, and Bel Canto Design's Black ACI 600 ($25,000), in April—are, respectively, slightly and considerably more expensive. And the fact that the M32 is Roon Ready is a plus for those who, like me, have migrated to using this impressive app for managing their music libraries. Highly recommended.

NAD Electronics International
633 Granite Court
Pickering, Ontario L1W 3K1
(905) 831-6555

Scintilla's picture

Reviewing your test results, John, it is hard to argue this is a better solution than using an M51/C510 paired with a good-quality conventional amplifier. Your tests of the M51 noise-floor revealed approximately 21 bits of resolution and pairing that with a high-current, low noise class A-A/B power amp with a high S/N ratio looks like a significantly better solution to me. All you gain here is some convenience with BluOS and maybe the potential for in-built 2-channel Dirac processing on an MDC board. Frankly, I'd rather have my Krell delivering the current than this class-D amplifier and I think NAD has objectively taken a step backwards with this product vs. the now-discontinued M51. Booo.

Long-time listener's picture

Yes, I have used the NAD M51 with an average-quality amp, and it does sound better than the M32. When that amp went bad, I started listening to the M51 through an Onkyo receiver. Not too bad, actually. I would give the M32 a Class B rating since it just doesn't sound entirely natural to me, especially with voices, and because of a persistently 'hard' quality to the sound.

cundare's picture

There's an easy way to do this. Tell your head you're tapping in 12/8. The rhythm abruptly becomes obvious, like when a hidden Magic Eye 3D image snaps into place.

Long-time listener's picture