DALI Rubicon 8 loudspeaker Measurements

Sidebar 3: Measurements

I used DRA Labs' MLSSA system and a calibrated DPA 4006 microphone to measure the DALI Rubicon 8's frequency response in the farfield, and an Earthworks QTC-40 for the nearfield and spatially averaged room responses. My estimate of the Rubicon 8's voltage sensitivity was 89.8dB(B)/2.83V/m, which is both within experimental error of the specified 90.5dB and significantly higher than average. Though this is 2dB lower than the Triangle Delta Signature's sensitivity, the Rubicon 8 will still play loudly with low-powered amplifiers. Those amplifiers, however, should be comfortable with driving low impedances: the DALI's impedance is specified as 4 ohms, though the solid trace in fig.1 reveals that its impedance magnitude actually remains above 4 ohms at all audio frequencies, and varies very little above 100Hz. Other than a combination of 6 ohms and –38.5° at 81Hz, the electrical phase angle (fig.1, dotted trace) remains low above 100Hz.

Fig.1 DALI Rubicon 8, electrical impedance (solid) and phase (dashed) (2 ohms/vertical div.).

Though the traces in fig.1 are smoothly continuous in the midrange, implying an absence of enclosure-panel resonances, I did find on the sidewalls, in places level with the uppermost woofer, modes at 355Hz, and between 600 and 700Hz (fig.2). The lower-frequency mode was present on all enclosure surfaces, and strongest on the small area of the gently curved baffle below the bottom woofer (fig.3). I heard no congestion in the midrange that correlated with this behavior, though when I listened to the panels with a stethoscope, I heard all of these modes. The upper-frequency resonances may be associated with the port resonance noted below, and with the slightly clangy coloration I occasionally heard with piano recordings.

Fig.2 DALI Rubicon 8, cumulative spectral-decay plot calculated from output of accelerometer fastened to center of sidewall level with top woofer (MLS driving voltage to speaker, 7.55V; measurement bandwidth, 2kHz).

Fig.3 DALI Rubicon 8, cumulative spectral-decay plot calculated from output of accelerometer fastened to center of the baffle below the bottom woofer (MLS driving voltage to speaker, 7.55V; measurement bandwidth, 2kHz).

The saddle between 30 and 40Hz in the impedance-magnitude trace suggests that the three ports on the rear panel are tuned in this region, and indeed, looking at the nearfield outputs of the three woofers does reveal that each has a sharply defined minimum-motion notch at 33Hz (fig.4, red trace). This is the frequency where the back pressure from the port resonance holds the cone still, all the acoustic output coming from the port. Though the woofers were tuned identically in this respect, the bottommost one begins to roll off above 200Hz and the middle one above 800Hz, leaving the top woofer to cross over to the tweeters (blue trace) at the specified 2.5kHz. The ports have very similar outputs, peaking between 20 and 60Hz, but there is a sharply defined peak at 700Hz in their summed response (green trace). This could be heard from behind the speaker with pink noise, but fortunately, the audibility of this resonance will be much reduced by the fact that the ports fire to the Rubicon 8's rear.

Fig.4 DALI Rubicon 8, acoustic crossover on supertweeter axis at 50", with summed nearfield responses of woofers (red) and ports (green).

Higher in frequency in fig.4, the crossover appears to have asymmetrical slopes, with the top woofer rolling off at an overall slower rate than the tweeters' 18dB/octave. The output of the tweeters continues to rise through their passband, which can also be seen in the graph of the Rubicon 8's farfield response on the supertweeter axis, averaged across a 30° horizontal window (fig.5, black trace above 300Hz). The black trace below 300Hz in fig.5 shows the complex sum of the woofer and port outputs, taking into account both acoustic phase angle and the different distances of the ports and woofers from a nominal farfield microphone position. The peak in the midbass will be due in part to the nearfield measurement technique, which assumes that the speaker is firing into a hemisphere rather a full sphere. But fig.5 also suggests that the Rubicon 8's low frequencies are balanced to sound rich, perhaps to balance the excess of top-octave energy seen in this graph.

Fig.5 DALI Rubicon 8, anechoic response on supertweeter axis at 50", averaged across 30° horizontal window and corrected for microphone response, with complex sum of nearfield responses plotted below 300Hz (black); anechoic response 20° to side of supertweeter axis at 50", corrected for microphone response (blue).

The blue trace in fig.5 shows the Rubicon's farfield response 20° to the speaker's side, which is the axis I sit on with the speakers firing straight ahead, as DALI recommends. Other than the increased rolloff above 18kHz, it is surprisingly similar to the black trace in this graph, confirming that the speaker does have wide horizontal dispersion.

Fig.6 shows the Rubicon 8's response plotted from 90° off axis on one side of the supertweeter axis to 90° off axis on the other side; fig.7 is similar except that it shows only the differences in response to the speaker's sides, which means that the on-axis response is plotted as a horizontal line. These graphs confirm that the Rubicon 8 has a horizontal dispersion that is both wide and well controlled, the latter evident in the even spacing of the contour lines. The slight off-axis flare in fig.7, indicated by the cursor position, is actually due to the small suckout centered on 6.5kHz in the on-axis response filling in to the speaker's sides. In the vertical plane (fig.8), the top-octave output shelves down 5° below the supertweeter axis. As the ribbon unit is 40" from the floor and the ear height of the average seated listener is 36", this will work against the speaker's top octave sounding exaggerated, as I found in my auditioning.

Fig.6 DALI Rubicon 8, lateral response family at 50", from back to front: responses 90–5° off axis, reference response, responses 5–90° off axis.

Fig.7 DALI Rubicon 8, lateral response family at 50", normalized to response on supertweeter axis, from back to front: differences in response 90–5° off axis, reference response, differences in response 5–90° off axis.

Fig.8 DALI Rubicon 8, vertical response family at 50", normalized to response on supertweeter axis, from back to front: differences in response 15–5° above axis, reference response, differences in response 5–10° below axis.

Fig.9 compares the Rubicon 8's spatially averaged, 1/6-octave response at the listening position in my room (red trace) with that of the KEF LS50 that preceded it in that room (see my Follow-Up review in the January 2015 issue, p.137). The two speakers' responses are very similar in the midrange and low treble. However, the DALI has significantly more energy in-room between 3kHz and the current limit of my hearing, just below 15kHz. That I was only occasionally aware of this excess is explained by what happens in the bass, where the Rubicon 8 has more energy apparent below 200Hz, which will tend to balance the speaker's upper-octaves output. Note also that the close coincidence between the resonant frequency of the reflex ports and the frequency of the lowest-frequency mode in my room gives rise to a significant boost between 20 and 40Hz that is not ameliorated by the spatial averaging.

Fig.9 DALI Rubicon 8, spatially averaged, 1?6-octave response in JA's listening room (red); and of KEF LS50 (blue).

In the time domain, the Rubicon 8's step response on the supertweeter axis (fig.10) suggests that its supertweeter and woofers are connected in inverted acoustic polarity, its tweeter in positive polarity. However, the decay of the tweeter's step smoothly blends at the 4ms mark with the negative-going start of the woofers' step, implying optimal crossover design. Correlating with DALI's "low-loss" concept, the Rubicon 8's cumulative spectral-decay plot (fig.11) is one of the cleanest I have seen, with no ridges of delayed energy visible above the midrange.

Fig.10 DALI Rubicon 8, step response on supertweeter axis at 50" (5ms time window, 30kHz bandwidth).

Fig.11 DALI Rubicon 8, cumulative spectral-decay plot on supertweeter axis at 50" (0.15ms risetime).

This loudspeaker's measured performance indicates a careful balance of its design parameters to produce a seductive yet clean full-range sound—as I heard in my listening room.—John Atkinson

US distributor: The Sound Organisation
1009 Oakmead Dr.
Arlington, TX 76011
(972) 234-0182

dalethorn's picture

It's fascinating what changes are possible with advancing materials technology. Someday we'll be 3-D printing speakers at home from patterns we buy off the Internet, much like we buy music tracks and e-books now. One of the greatest things future listeners will enjoy is the fact that since physical speakers today are an inventory item and expense, there's only so much flexibility possible in getting an 'ideal' speaker to perform in a variety of rooms. In the future, the patterns to print will offer numerous options to tweak the speakers ahead of printing them, or doing modifications afterward.

I mention this only after reading the article and seeing how DALI is using new technology in their magnet designs etc.

SNI's picture

To me it seems that all the drivers are connected in noninverted acoustic polarity, but the speaker as a whole was conected in inverted polarity during the measurement.
I suggest that the step response should på turned upside down to paint the real picture.
Anyways this is impressive time domain behavior, I do not recall any speaker matching this.
In combination with the frequency response, the impedance and phase, this could prove to be a very neutral reproducer.
Wonder what their top of the line Epicon can do.