NHT Xd active loudspeaker system John Atkinson January 2006
When Kalman Rubinson reviewed NHT's $6000 satellites+subwoofer loudspeaker system in the November 2005 issue (pp.105–115), he was very impressed, concluding that "The NHT Xd is the best thing to come down the pike in a long time....Because the Xd's performance is competitive at the highest levels, anyone interested in high-quality music reproduction must hear it." I was also impressed enough by the Xd's measured performance that I decided I would do a Follow-Up.5
A Follow-Up would also allow me to experiment with different crossover filters that NHT's Jay Doherty had e-mailed me after the November issue had gone to press, and which were claimed to endow the system with greater dynamic range.
The heart of the Xd system is the XdA, a four-channel power amplifier that uses PowerPhysics class-D amplifier modules and includes a digital-domain crossover from Australian company DEQX for up to two XdS satellites and two XdW powered bipole subwoofers. The XdA's equalization and filter parameters are held in nonvolatile memory, and can be updated via a USB link using a Windows XP program (provided you have Microsoft's .NET extension installed).
The three filters NHT sent me were:
1) "flatterhighend.flt"—similar to the original filter set, but, according to Jay Doherty's enclosed notes, "Flatter above 10kHz; crossover point between tweeter and midwoofer moved up a hair (to 2.4kHz from 2.3kHz) to increase power handling; reversed sub phase to try for better sub/satellite integration";
2) "150Hzcrossover.flt"—intended for dual subwoofer use; keeps the inverted subwoofer phase but moves the crossover between the satellite and subs from 110Hz to 150Hz to enable the system to play louder; and
3) "150Hzxovereqto20Hz.flt"—equalizes the subwoofers to be –3dB at 20Hz instead of 26Hz.
There will be more filters to come. According to Doherty, "We are working on a linear-phase filter to use between the satellite and subwoofer. LP filters require firmware changes that we should be able to implement by the time the first room EQ software package is complete. However, this will increase total system processing time to about 18ms (from 7ms), so it might not be usable with video."
Kal had used only a single XdW subwoofer, but as I intended to try the filter sets with the higher crossover frequency, I asked NHT to send me a second subwoofer. This increases the system price to $7200.
Sound: As KR had described, setting up the Xd system was quick and easy. Unlike Kal, I had no ground-loop problems, but this may have been due to the fact that I used balanced connections.
Preamp was first the Mark Levinson No.326S, connected to the NHT XdA power amplifier with 15' runs of Madrigal interconnects; then an NHT Passive Volume Control, connected with 6' runs of Canare interconnects (the only balanced cables I had with the necessary TRS connectors). Digital source was either a Mark Levinson No.31.5 CD transport or a Technics DVD-A10 DVD player, hooked up to my ML No.30.6 D/A processor via Kimber Kable Orchid AES/EBU or AudioQuest SVD-4 S/PDIF datalinks, respectively. The No.30.6 was connected to the preamp with balanced 1m lengths of AudioQuest Cheetah. I also used an Olive Symphony media server to play back uncompressed 16-bit AIF files via a TosLink connection to the Levinson DAC, and an Ayre C-5xe universal player connected to the preamps with 15' lengths of balanced Crystal interconnects.
When he'd first set up the Xd, Kal had remarked on a lack of integration between the satellites and subwoofer: "I measured an in-room response that dipped several dB in the upper-bass range centered on 130Hz. This detracted from the general impression of weight, warmth, and richness of sound. While the crossover slope between the XdS and XdW may be a very steep 48dB/octave at 110Hz, there is still significant signal overlap between the woofer and satellites; the positioning of the woofer is important. A lower frequency, of course, would compromise the power-handling limits of the XdS's 5.25" driver. I found that moving the XdW forward so that it was the same distance from the listener as the satellites filled in the integrated response, as confirmed by instrument and ear."
With first one subwoofer, then two, and using the same filter set Kal had used, I couldn't eliminate the upper-bass discontinuity no matter how I experimented with subwoofer positioning. The bass region was rich and deep, but didn't integrate sufficiently well with the satellites. In addition, while midrange tonalities were reproduced with a delightful lack of coloration, there was a somewhat reticent quality to the highest frequencies. Whether or not these problems bothered me was very dependent on the music played.
I have written before about how the choice of playback equipment can change the choice of music played. The discontinuity between the satellites and subwoofers was more audible with rock music, with its ubiquitous four-in-the-bar kick drum, than it was with classical recordings. For the three weeks I used the Xd system with the original crossover filters, I found myself playing a lot more symphonies and concertos than I had done with other speakers that have recently occupied my listening room.
The Xd's extended low frequencies were a much-appreciated benefit with this kind of music, and the system's somewhat veiled highs were much less of an impediment to musical enjoyment. I dug out discs that I hadn't played much, such as Michael Tilson Thomas' reading of Mahler's Symphony 3 (SACD, San Francisco Symphony 821936-0003-2) and Seiji Ozawa's of Tchaikovsky's Symphony 6 (SACD, Pentatone PTC 5186 107). The double basses in the former's final movement had glorious weight through the Xd system, though I found the system's midrange resolution very revealing of the wayward intonation of the occasional sustained brass chord in the latter. But when I played Eric Johnson's "Desert Rose," from Live from Austin TX (CD, New West NW6084, recorded for the Austin City Limits TV program), the low bass just grumbled along with the higher frequencies.
There was also something I noticed with the toneburst track on my Editor's Choice compilation (CD, Stereophile STPH016-2). This track comprises equal-length tonebursts that move from 32Hz up to 3.2kHz in half steps, then back down again. I created this signal to investigate room and speaker-cabinet resonant problems, but when I played it over the Xd system, I was puzzled to hear what sounded like very faint "ghosts" accompanying the sinewave bursts, almost but not entirely like modulation noise. A puzzle, though I did wonder if this phenomenon had something to do with the veiling I had noticed on music.
Time to install the new filter set. I began with "flatterhighend.flt," but then changed to "150Hzcrossover.flt" because the system was still having difficulty handling music with extreme amounts of upper-bass energy. Stanley Clarke's double-bass solo on Airto Moreira's "Nevermind," from our Test CD 3 (Stereophile STPH006-2), managed to shut down the left-channel satellite-woofer amplifier at the volume I was finding appropriate for the music. (The image lurched to the right and a red light came illuminated on the XdA amplifier's front panel. Turning off the amplifier and turning it on again fixed the problem.)
Repeating this torture test with the 150Hz crossover filters didn't result in any shutdowns with the Clarke solo, and the chord that begins "In the Flesh," from Pink Floyd's Is There Anybody Out There? (The Wall Live) (CD, Columbia C2K 62058), sounded positively cataclysmic. The low-frequency crash of the wall being demolished on this album gave the XdW subwoofers one heck of a workout. However, the drum-and-unvoiced-guitar intro of Eric Johnson's "East Wes," from his live Austin City Limits CD, still managed to shut down the left XdS woofer amplifier if I wasn't careful with the volume control. (Setting the NHT PVC's control to –6dB, equivalent to an in-room SPL of around 108dB, could trigger an amplifier shutdown with this track.)
I listened again to the half-step–spaced tonebursts on Editor's Choice. Despite the revised filters, I could still hear the slight modulation noise accompanying the tonebursts. Probably not too much should be made of this phenomenon; I mention it only because I had never heard it before with conventional loudspeakers.
What about the high frequencies? The veiling that had bothered me was much reduced, the top octaves having more air apparent. Perhaps more significant was the minimizing of a slight mid-treble "shoutiness" that had limited maximum volume with the original filters. The presentation was first-rate in its lack of coloration, superbly stable and detailed imaging, excellent retrieval of subtle reverberation tails, and overall top-to-bottom integration. Wow!
Summing Up: Art Dudley recently wrote that the true test of a component under review is how you feel when it comes time to return it to the manufacturer. In the case of NHT's Xd, I felt bad about having to send it back. It did almost all of what I want an amp and speakers to do, with very little downside. However, I have five pairs of speakers on deck for review and I have to move on.
In the meantime, the revised crossover software, in conjunction with stereo subwoofers, has taken a relatively affordable, excellent-sounding loudspeaker system and made it almost beyond criticism. It's true that the NHT Xd system still lacks ultimate dynamic range, but there's only so much you can ask in terms of ultimate loudness from a pair of drive-units with a radiating diameter of just 3". Highly recommended.—John Atkinson