Sonus Faber Minuetto loudspeaker Measurements
Sidebar 3: Measurements
The Sonus Faber's B-weighted sensitivity, at a calculated 85dB/W/m, was moderate for a small speaker. Its impedance (fig.1) is kind to amplifiers, remaining above 6.5 ohms at low frequencies, and above 12 ohms in the treble. The port is tuned to 42Hz, the frequency of the saddle between the two low-frequency impedance peaks, though the lower peak is much lower in frequency than is usually the case.
The individual response plots for the drivers and port (fig.2) reveal the port outputthe bandpass centered on 50Hzto be significantly down in level from that of the woofer. It also has a significant peak in its output at 900Hz, though this will probably have no subjective consequences, due to its low level and the fact that the port faces the rear. The woofer's output is humped up in the upper bass, though relatively flat through the midrange and rolling off smoothly above 2kHz. The tweeter comes in slowly between 1 and 3kHz, rising to its maximum output level above 8kHz.
The integration of these drive-unit responses on the tweeter axis at a microphone distance of 45" and averaged across a 30° horizontal window is shown in fig.3. The midrange and lower midrange are shelved-up by up to 5dB, while the low treble is shelved-down by 2dB. Here is much of the reason for Corey's finding the Sonus Faber to have a "polite," over-dull balance. The relative lack of energy in the presence region (an octave below the suckout in the Audio Physic Step's mid-treble, a speaker reviewed by Jack English elsewhere in this issue) will result in a balance that will be too uninvolving for many tastes. The hump up in the midbass, however, will endow the Minuetto with impressive low frequencies, as CG commented.
It's possible that the Minuetto's limited HF dispersion also contributes to its mellow balance. Fig.4, which shows just the differences to be found off-axis to the speaker's sides, reveals that the off-axis fall-off in the Minuetto's tweeter output occurs more rapidly than is normal with a 1" dome tweeter. This will give the listening room's reverberant field an HF-less balance which will exacerbate the recessed on-axis sound quality.
Corey remarked upon the Minuetto's sensitivity to vertical listening height due to its use of first-order crossover filters. This is confirmed by its vertical-dispersion graph (fig.5). As well as the top octaves rolling off more than a few degrees off the tweeter axis, the presence-region depression is exaggerated by a crossover suckout appearing both above and below this axis. In addition, if you sit too high, the upper midrange becomes exaggerated in level; sit too low and the upper midrange starts to develop a suckout.
In general, a designer will choose to use first-order filters to give time-coherent performance. However, with conventional dynamic drive-units, this will also involve the tweeter having to be physically stepped-back somehow, due to its smaller chassis depth. This is normally achieved with a sloping baffle, as in the Spica and Thiel designs, or by staggering separate driver modules (Vandersteen and Dunlavy Audio Labs). Without either of these strategies being adopted, the speaker will not be time-coherent, as is revealed by the Minuetto's step response (fig.6). Both drivers have the same positive acoustic polarity, but the tweeter's output pulse leads that of the woofer by 300µs or so. The good news is that the Sonus Faber's cumulative spectral-decay, or waterfall, plot (fig.7) is very clean, with only some low-level hash present in the treble.
Finally, while the speaker's solid-feeling cabinet has a couple of resonant modes noticeablefig.8 shows a waterfall plot calculated from the impulse response of a simple PVDF-tape accelerometer fastened to the center of the rear panelthese are high enough in frequency that they should not have too much of a subjective effect.John Atkinson