Floor Loudspeaker Reviews

I used DRA Labs' MLSSA system and a calibrated DPA 4006 microphone to measure the Göbel Divin Marquis's frequency response in the farfield, and an Earthworks QTC-40 mike for the nearfield and in-room responses.

Usually, I measure loudspeakers in our backyard, weather permitting, or in our living room with the furniture pushed to the sides. This eliminates or moves back in time the reflections of the speaker's output. However, as this 330lb loudspeaker was too massive for me to move outside or upstairs, I had to do the quasi-anechoic measurements in my listening room. I slid one of the speakers forward so that it was aimed across the room's diagonal and was as distant as possible from the nearest sidewall. However, the proximity of room boundaries, the floor in particular, meant that even though I measured the Göbel's quasi-anechoic farfield behavior at 1m rather than my usual 50", I still had to aggressively window the time-domain data. This reduces the measurements' resolution in the midrange.

Göbel specifies the Divin Marquis's sensitivity as 92dB/W/m; my estimate was a little lower, at a still-high 89.5dB(B)/2.83V/m. The Divin Marquis's impedance is specified as 4 ohms with a minimum value of 3.4 ohms at 95Hz. The impedance magnitude (fig.1, solid trace) remains between 4 and 6 ohms for almost the entire audioband, with a minimum value of 2.9 ohms between 83Hz and 99Hz. The electrical phase angle (dashed trace) is generally low, but there is also a current-hungry combination of 5 ohms and a phase angle of –45° at 26Hz. I used the formula in a 1994 JAES paper by Eric Benjamin to calculate what UK writer Keith Howard has called the "equivalent peak dissipation resistance" (EPDR, footnote 1). The Divin Marquis has minimum EPDRs of 1.77 ohms at 25Hz and 1.53 ohms between 53Hz and 57Hz. Though the EPDR is close to 4 ohms in the midrange and treble, this Göbel loudspeaker will work best with amplifiers that are comfortable driving loads below 4 ohms.

The Divin Marquis seemed extremely inert to the "knuckle rap" test. When I investigated the enclosure's vibrational behavior with a plastic-tape accelerometer, the only resonant modes I found were extremely low in level. Fig.2 shows the only one I found on the top panel, at 602Hz. Any modes on the side panels were even lower in level.

The saddle centered at 25Hz in the impedance magnitude trace implies that this is the tuning frequency of the four ports, and the resultant minimum-motion notch in the woofer's nearfield output (fig.3, blue trace) lies at that frequency. The nearfield response of the ports (red trace) peaks between 20Hz and 60Hz, though I suspect that the measured output is contaminated by some crosstalk from the woofer. The ports' upper-frequency rolloff is clean overall, though two low-level peaks are visible between 400Hz and 600Hz. These peaks are also present in the woofer's high-frequency rolloff, and I could just hear them with the noise-like MLSSA signal when I drove the woofer and ports by themselves. There is the usual upper-bass boost in both the woofer and port outputs in fig.3, which are due to the nearfield measurement technique, which assumes the baffle extends to infinity in both lateral and vertical planes.

The woofer crosses over to the midrange unit (fig.3, green trace) at the specified 140Hz; the farfield response of the midrange unit and tweeter, averaged across a 30° horizontal window (fig.3, black trace above 300Hz), is impressively even up to the top of the audioband. A small suckout is visible between 800Hz and 1.2kHz, as well as some small ripples in the response higher in frequency. This implies interference between the midrange unit's output and the reflections of its output. (The woofer was not connected for this measurement.) The geometry of the measurement setup meant that the reflection of the midrange driver's output from the floor—the closest boundary—arrived at the microphone approximately 3ms after the direct sound. The suckout must therefore be due to a reflection occurring earlier in time, perhaps from the edges of the wide baffle. But, given the evenness of the farfield response, this is probably of academic interest only.

Fig.4 shows the Göbel's horizontal dispersion, normalized to the response on the tweeter axis, which thus appears as a straight line. The geometrical limitations of my listening room meant that I could only plot the Divin Marquis's off-axis responses to 45° instead of my usual 90°. The horn-loaded tweeter offers wide dispersion up to 10kHz, and the contour lines in this graph below that frequency are relatively evenly spaced, which correlates with the stable stereo imaging I noted in my auditioning. In the vertical plane (fig.5), with the off-axis response again normalized to the tweeter-axis response, the Divin Marquis's balance doesn't change appreciably up to 10° above and below the axis. A suckout develops 15° above the tweeter axis at the upper crossover frequency of 1.6kHz, but this will only be heard by a listener standing close to the speaker.

Fig.6 shows the Divin Marquises' spatially averaged response in my room when they were driven by the Parasound amplifiers with AudioQuest cables. It is generated by averaging 20 1/6-octave–smoothed spectra, taken for the left and right speakers individually using a 96kHz sample rate, in a vertical rectangular grid 36" wide × 18" high and centered on the positions of my ears. This tends to average out the peaks and dips below 400Hz that are due to the room's resonant modes. Even so, the Göbels still excite the lowest-frequency modes in my room, and a slight excess of energy can be seen between 500Hz and 800Hz. The in-room response is otherwise superbly even from the midrange through the mid-treble and smoothly slopes down above 6kHz, this due to the increasing absorption of the room's furnishings in this region. While performing these measurements, I noticed both that the responses at the listening position of the two Divin Marquises were closely matched in the midrange and treble and that the horn-loaded tweeter did indeed offer wide dispersion.

The Göbels' in-room response is shown as the red trace in fig.7 but is overlaid with the spatially averaged responses of the two pairs of speakers that had preceded them in my room: the GoldenEar BRXes that I reviewed in September 2020 (blue trace) and the Vimberg Minos that I reviewed in April 2020 (green trace). While the small BRXes have a little more upper-bass energy than the two floorstanding speakers, their low frequencies roll off much faster, of course. The GoldenEars also have a similar peak in the upper midrange to the Divin Marquises, and both the BRXes and Minos have a little more mid-treble energy than the Göbels. Compared with the Divin Marquises, the GoldenEars have a little too much top-octave output in-room, the Vimbergs not quite enough.

In the time domain, the Divin Marquis's step response on the tweeter axis (fig.8) indicates that the tweeter and midrange unit are connected in positive acoustic polarity. The decay of the tweeter's step smoothly blends with the positive-going start of the midrange unit's step, which implies optimal crossover implementation. The woofer is connected in inverted acoustic polarity. Its output arrives at the microphone after that of the midrange unit, but the smooth blend of its step with the decay of the midrange unit's step is disturbed by a reflection 1.5ms after the midrange unit's step. As above, this reflection is too early to be due to a floor bounce, but because of the presence of this reflection, interpreting the Göbel Divin Marquis's cumulative spectral-decay plot (fig.9) is difficult. However, this graph is relatively clean in the region covered by the tweeter.

The late Spencer Hughes, founder of Spendor, used to say that "big speakers can have big problems." The Göbel Divin Marquis may be one of the biggest loudspeakers I have had in my listening room, but its measured performance reveals that if it has any problems, they are minimal.—John Atkinson