Stand Loudspeaker Reviews

John Atkinson wrote about the Dutch & Dutch 8C in April 2020 (Vol.43 No.4):

Kalman Rubinson very favorably reviewed the three-way, active Dutch & Dutch 8c ($12,500/pair) in the August 2019 issue of Stereophile. "The D&D 8c demonstrates that active, DSP-empowered speakers are the future," he concluded, adding, "the Dutch & Dutch 8c might be the sweet spot in the new category of active speakers. It's a truly full-range system with enough dynamic range and power for almost any domestic situation."

I was impressed by what I found when measuring the Dutch & Dutch 8c. A flat on-axis response and superbly well-controlled dispersion in both horizontal and vertical planes were combined with time-coincident output and one of the cleanest waterfall plots I have encountered. "In summing up the Dutch & Dutch 8c's measured performance, all I can say is 'Wow!'" I wrote.

As Kal's review samples were promptly returned to the Dutch & Dutch US distributor, and as I wanted to spend more time with this intriguing design, I asked for a second pair of 8cs. I received these early in the New Year—Kal's samples were serial numbers 515 and 516; mine were 947 and 948—and D&D's Martijn Mensink visited to set them up.

Ideally, these speakers should be positioned between 100mm (4") and 1.5m (59") from the wall behind them, which allows the output of the twin rear-firing subwoofers to be reinforced with the in-phase reflection from the wall. However, this wasn't possible in my room; I had placed the speakers farther out, sitting on 24" stands in the positions where the Magico M2s I reviewed in February and the Vimberg Minos I review elsewhere in this issue worked best. I connected the 8c's to my network and, using the control app at the Dutch & Dutch site lanspeaker.com, set each speaker to "free space" regarding both the rear wall and sidewalls. (This app also allows one to control volume, set each speaker to Left or Right, choose the input type—analog at –10dBV or +4dBU or digital AES/ EBU—and access the Room Matching and parametric equalizer settings.) (footnote 1)

Mensink felt my placement would be okay, then set to work optimizing the speakers' performance. A new feature added to the loudspeaker since Kal's review (the new samples were running v.1.4.38 firmware) is the ability to use John Mulcahy's Room Equalization Wizard app (REW) to correct for the room's acoustic problems (footnote 2). The 8c uses digital signal processing (DSP) to implement the crossover, drive-unit time alignment, and equalization; with REW, DSP can now also be used to apply digital correction filters created with REW. Mensink got out his laptop, plugged its analog output into the speakers' analog inputs (set to –10dBV), logged into my Wi-Fi network, and ran REW.

REW located the loudspeakers on the network; Mensink had already combined them into a group. With the aid of a USB microphone, he used the app to measure the response of each at several points centered on the listening position. REW overlaid the responses with a target curve, then, with that approved, calculated the appropriate filters and uploaded them to the speakers' DSP. Mensink remeasured the listening-position response and decided that, while it was close to the target, it was appropriate to manually add a couple of low-Q, low-amplitude parametric equalizer settings, one in the midrange and one in the bass, to the filters employed by REW. That done, we listened to several kinds of music—rock, classical orchestral, choral, EDM, and piano—and Mensink declared himself satisfied with the setup.

The next day, I started my critical listening using the same ancillaries: a Roon Nucleus+ server sending data to a PS Audio DirectStream D/A processor via either Ethernet or USB to the Chord Hugo M Scaler I reviewed in March, and then on to the PS Audio via S/PDIF. The PS Audio's own volume control was effectively bypassed by setting it to "100," and the DAC's balanced analog outputs were connected to the D&Ds with AudioQuest Wild Blue Yonder cables. The D&Ds got their AC from an AudioQuest Niagara 5000.

Without the REW correction filters active, the 8c's sounded more neutral in the low treble than the KEF LS50s I purchased following my review in 2012, but with more top-octave energy. Listening to the 1/3-octave warble tones on my Editor's Choice CD (Stereophile STPH016-2), the low bass sounded exaggerated. My placement was not sympathetic to the tuning of the 8c's subwoofers, as the lowest-frequency mode in my room was being maximally excited.

Using the lanspeaker.com app, I activated the REW correction filters for each loudspeaker and continued listening. The midrange still sounded neutral, with excellent transparency. Joni Mitchell's husky contralto on the title track to her 2000 orchestral album Both Sides, Now (26/96 ALAC file, ripped from DVD-A, Reprise) raised goose bumps, so well-formed was its presentation on the 8c's. One-third–octave warble tones were reproduced in full measure down to the 40Hz band, with the 80Hz band slightly lower in level. The 32Hz tone was still boosted by the lowest-frequency mode in my room, though not nearly as much as it had been without the REW correction. The 25Hz warble was readily audible, and I could just hear the 20Hz tone at my normal listening level. The lower-frequency tones sounded very clean, meaning that distortion must have been very low. When I played my 2014 recording of Jonas Nordwall's performance of Widor's Organ Symphony No.5 (24/88.2 AIFF file), even with the REW filters active it sounded as if the wall of my listening room were shuddering in sympathy with the lowest-frequency organ pipes. The same thing happened with the dropped bass line in the second half of Trentemoller's "Night Walker" from The Last Resort (16/44.1k ALAC rip from CD, Poker Flat 827170113022). The D&Ds can output serious low frequencies!

Played on the 8c's, the Widor recording transported me back to Portland, Oregon's First United Methodist Church, with the image of the organ's ranks of pipes spread unambiguously across the stage. The D&D speakers did indeed excel when it came to stereo imaging. Images were tightly focused and impressively stable at all frequencies. This could have been because the responses of the two speakers were very closely matched but also because their lateral radiation pattern is narrower than usual. The reflections from the sidewalls, which might smear the stereo image, would therefore be lower in level than with a conventional design (footnote 3).

Kal had remarked in his review that the Dutch & Dutch speakers' soundstage "began at the plane described by the speakers' front baffles and extended deep behind them." I also noted the excellent soundstage depth. Acoustic objects that were balanced forward in the mix remained to the front of the soundstage, but more depth was apparent with objects that had some reverberation associated with them. When, for example, Jonas Nordwall selects a rank of higher-frequency pipes that are farther away for the ostinato arpeggios in the quiet passage halfway through the Widor, "farther away" is what I heard.

Having got a handle on the Dutch & Dutch 8c's sound quality, I got out my test gear to examine the effect of the filters calculated by REW on their nearfield responses (fig.1). The black traces in this complicated-looking graph show the original responses of the woofers and subwoofers; the blue and green traces respectively show the equalized outputs of the woofer and subwoofers of S/N 947; the red and magenta traces show the equalized woofer and subwoofer outputs of S/N 948. You can see that, with the exception of the filter just below 30Hz, which applied a 12dB cut, the correction filters generated by REW are relatively mild in their action, with boosts around 3dB and cuts between 3dB and 5dB. The sloped-down responses above 400Hz in this graph are an artifact of the nearfield measurement technique.

Fig.2 shows the spatially averaged response (footnote 4) in my room without any room correction (green trace) and with the REW-generated filters active (red trace). The first thing to be seen in this graph is that the D&Ds' output rolls off sharply above 20kHz, confirming my conjecture in the original review that the 8c's internal digital signal processing operates with a sample rate of 48kHz. I asked Martijn Mensink about this when he visited. He confirmed that this was the case and explained that this was because the speaker's DSP could operate with greater precision at this sample rate. An additional benefit would be that there would be no energy present to excite the fundamental dome resonance of the 8c's 1" aluminum-magnesium–alloy tweeter, which lies around 28kHz.

The second thing to note in fig.2 is that the huge peak between 15Hz and 50Hz in the uncorrected response has been largely eliminated by the correction filter. Nevertheless, the in-room response still extends to 20Hz, though there is also now a slight lack of energy in the 50–100Hz octave.

Third, the corrected D&Ds' spatially averaged output (red trace) is extremely flat. The response falls within ±1dB limits between 740Hz and 10.3kHz, and while there is a raised plateau in the midrange, this remains at or less than +2.1dB. (Mensink had originally chosen a reduction of level of 3.5dB with the low-Q, 720Hz filter rather than 2dB, and it appears from my measurements that –3.5dB would have been a better choice.)

The red trace in fig.3 repeats the REW-corrected D&Ds' spatially averaged response in my room. For reference, the blue trace shows the spatially averaged response of my KEF LS50s, measured in an identical manner. To plot the KEF LS50s' in-room response, I normalized it to the D&Ds' response in the midrange. With speakers that are not perfectly flat, choosing which frequency to use as the reference will depend on the music played. A speaker with a peak in the midrange will sound as if it has insufficient treble with music whose midrange is perceived as being correctly balanced. On the other hand, if the treble balance sounds correct in level with some music, the midrange will sound exaggerated.

My choice of making the two pairs of speakers equal in the midrange makes it appear that the KEFs have more low-treble energy than the D&Ds but slightly less energy in the top two octaves. This is pretty much what I heard as the difference between the two speakers. And as the KEF is a minimonitor with limited low-frequency extension, it steps out of the way of the lowest-frequency resonance in my room, while even with the REW-generated correction filters, this mode still makes its presence known with the D&Ds.

Fig.4 compares the D&Ds' spatially averaged response (again the red trace) with the two conventional and considerably more expensive high-performance loudspeakers that spent time in my listening room before and after them: the Magico M2s (green trace) and Vimberg Minos (blue trace). (Both passive loudspeakers were driven by Lamm M1.2 monoblocks for these measurements.) I have expanded the vertical scale in this somewhat busy graph compared with figs.1 and 2 to make the differences more apparent. (Each horizontal division is 5dB.)

The in-room response of all three pairs of speakers slopes down in the top two audio octaves; this is due both to the increasing absorption of the room furnishings at high frequencies and to the increasing directivity of their tweeters in this region. The Magicos start to roll off a little earlier than the other two speakers, and the D&Ds have more energy in-room above 8kHz than either. Because the 8c's Treble control tilts the high-frequency response up or down, I could experiment with using this control to reduce the speaker's top-octave output (not shown).

The Vimberg's in-room response is as flat on average throughout the midrange and mid-treble as the 8c's, but the peaks and dips are greater, requiring ±3dB limits. All three loudspeakers excite the lowest-frequency mode in my room, though the sealed-box M2 does so to a lesser extent than the REW-corrected 8c, the reflex-loaded Mino to a greater degree.

I was impressed both by the sound quality and the quality of engineering exhibited by the Dutch & Dutch 8c. Even without REW, the 8c's response extends at full level to 20Hz. But with the REW integration and the ability for a pair of these speakers to act as a Roon endpoint, which will be available by the time you read these words (footnote 5), I echo Kal's conclusion that the D&D 8c demonstrates that active, DSP-empowered speakers are the future. And I also echo what I wrote in our original review: "Wow!"—John Atkinson

---

Footnote 1: The network connection is only necessary for setting the controls; the 8c is fully operational without it.

Footnote 2: REW is freeware but donations are welcome: www.roomeqwizard.com/.

Footnote 3: See fig.3 in the original review.

Footnote 4: I create these response traces by averaging 20 1/6-octave–smoothed spectra, taken for the left and right speakers individually using an EarthWorks QTC-40 microphone and a Metric Halo MIO2882 interface operating at a 96kHz sample rate, in a vertical rectangular grid 36" wide 18" high and centered on the positions of my ears. This grid is larger than that used by Martijn Mensink for his measurements with REW.

Footnote 5: I experimented with the Roon integration at the end of the review period. Roon identified the Group rather than the individual loudspeakers, running RAAT v.1.1.31 and labeled "Uncertified."