Sidebar 3: Measurements
I used DRA Labs' MLSSA system and a calibrated DPA 4006 microphone to measure the Wilson Sabrina's frequency response in the farfield; for the nearfield frequency responses, I used an Earthworks QTC-40, which has a ¼" capsule and thus presents no significant obstacle to the sound. For logistical reasons, I measured a different sample (serial no.0485) from those auditioned by Robert Deutsch (0375 and 0376). For the measurements, I hoisted this surprisingly heavy speaker onto my Outline speaker turntable—oh, my achin' back—and to avoid the possibility of damaging the cabinet's finish, I did not remove its protective film. As RD found the Sabrina's sound to be "just a bit less open" with the grilles, and so did all of his critical listening with the grilles removed, I performed almost all of the measurements without the grille.
My estimate of the Sabrina's voltage sensitivity on the tweeter axis was 88.2dB/2.83V/m, slightly higher than the specified 87dB. Wilson specifies the Sabrina's electrical impedance as 4 ohms, with a minimum magnitude of 2.53 ohms at 139Hz. My measurement of the impedance is shown in fig.1: the value (solid trace) varies between 4 and 6 ohms between 250Hz and 4kHz, and the minimum magnitude is 2.44 ohms at 135Hz. However, there is also a combination of 4.44 ohms and a capacitive phase angle of –58° at 72Hz, a frequency where music can have considerable energy; this will place a significant demand for current on the partnering amplifier.
I suspect that the Sabrina has a relatively strong internal acoustic resonance at 242Hz. This resonance is of high Q (Quality factor), which means that it needs to be excited by continuous energy at that exact frequency to become fully developed. The frequency 242Hz also falls between the notes B-flat3 and B3 in Western music (when A4=440Hz)—so other than with drums, the resonance might well not be excited with music. I note that RD didn't comment on any lack of clarity in the Wilson's upper bass that could be due to this behavior—quite the opposite, in fact, as he wrote that recordings of bass drums or timpani "had appropriately quick onset and very little overhang."
The broad peak between 40 and 240Hz in fig.3 will be due in part to the nearfield measurement technique. But it does suggest that the Sabrina's woofer alignment is a little on the underdamped side, which will endow the speaker with subjectively powerful low frequencies, as RD found. Higher in frequency in this graph, the Sabrina's farfield output, averaged across a 30° horizontal window centered on the tweeter axis, is relatively even, with small dips at some frequencies balanced by small peaks at others. The most obvious feature is the slight excess of energy between 4 and 6kHz. This graph was taken without the speaker's grille. Repeating the measurement with the grille in place actually gave a more even output in the mid-treble, with the peak eliminated—although, as noted earlier, RD preferred the sound without the grilles.
There is a very narrow suckout in the tweeter's output. I initially thought this was due to the usual leakage at 15,625Hz from the measurement computer's video display, but the suckout occurs slightly lower in frequency, at 14,650Hz. This suckout was not affected by the grille, but as it is both very narrow and high in frequency, I can confidently predict that it will not be audible.
Fig.4 shows the Sabrina's horizontal dispersion, normalized to the response on the tweeter axis. The apparent flare off axis at the cursor position in this graph actually coincides with the slight lack of energy in the on-axis response, meaning that, in typical rooms, the speaker's presence-region balance will sound neutral. With the Sabrina's narrow baffle surrounding the tweeter, the speaker's top-octave dispersion is wider than is usual with a 1" dome, which is presumably why the Wilson design team didn't feel the need to boost the tweeter's on-axis output in this region. In the vertical plane (fig.5), there is very little change in the speaker's response over a ±5° window centered on the tweeter axis.
Turning to the time domain, the Sabrina's step response on the tweeter axis (fig.6) resembles that of the other Wilson speakers we've reviewed in recent years, with the tweeter and woofer connected in positive acoustic polarity and the midrange unit in negative polarity (footnote 1). But more important than the polarities is the fact that the decay of each unit's step smoothly blends with that of the next lower in frequency, suggesting that, in combination with the sloped-back baffle, the crossover has been optimally designed. The Sabrina's cumulative spectral-decay plot (fig.7) shows a clean initial decay, particularly in the region covered by the tweeter, but with some low-level modes developing in the low treble.
Footnote 1: See, for example, fig.8 here.
Fig.1 Wilson Sabrina, electrical impedance (solid) and phase (dashed) (2 ohms/vertical div.).
A very slight discontinuity at 240Hz in the magnitude trace (solid trace) coincides with a point of inflection in the electrical phase angle (dashed). This suggests some sort of resonant behavior in the enclosure at this frequency, though I note that RD found the Sabrina's cabinet to be extremely inert. The top part was indeed inert, but the knuckle-rap test indicated that the lower portion was slightly lively. Investigating this behavior with a plastic-tape accelerometer, I found a high-Q mode at 242Hz on the rear and side panels (fig.2), though this was relatively low in level.
Fig.2 Wilson Sabrina, cumulative spectral-decay plot calculated from output of accelerometer fastened to center of rear panel level with woofer (MLS driving voltage to speaker, 7.55V; measurement bandwidth, 2kHz).
The saddle centered between 30 and 40Hz in the impedance-magnitude trace suggests that the 3"-diameter port at the base of the rear panel is tuned in this region. However, looking at the woofer's output in the nearfield (fig.3, blue trace), this has the expected minimum-motion notch—where the back pressure from the port resonance holds the woofer cone stationary—higher in frequency, at 44Hz. The port's output, again measured in the nearfield (red trace), peaks between 40 and 70Hz, and there is a small peak in its upper-frequency output at 242Hz, the frequency of the mode I found on the rear panel, followed by a steep rolloff. There is also a sharply defined discontinuity in the woofer output at the same frequency, as well as in the sum of the midrange, woofer, and port outputs (black trace). (I didn't include the contribution of the port that loads the midrange unit.)
Fig.3 Wilson Sabrina, anechoic response on tweeter axis at 50", averaged across 30° horizontal window and corrected for microphone response, with nearfield responses of woofer (blue), port (red), and complex sum of midrange, woofer, and port responses (black), all plotted below 300Hz.
Fig.4 Wilson Sabrina, lateral response family at 50", normalized to response on tweeter axis, from back to front: differences in response 90–5° off axis, reference response, differences in response 5–90° off axis.
Fig.5 Wilson Sabrina, vertical response family at 50", normalized to response on tweeter axis, from back to front: differences in response 15–5° above axis, reference response, differences in response 5–10° below axis.
Fig.6 Wilson Sabrina, step response on tweeter axis at 50" (5ms time window, 30kHz bandwidth).
Fig.7 Wilson Sabrina, cumulative spectral-decay plot on tweeter axis at 50" (0.15ms risetime).
Overall, its measured behavior suggests that Wilson Audio's design of the Sabrina has been carefully optimized.—John Atkinson
Footnote 1: See, for example, fig.8 here.
