Floorstanding Loudspeakers

All measurements were taken using DRA Labs' excellent MLSSA system. However, the bulk and weight of the Dynaudio Evidence Temptation made it impossible for me to raise it off the floor for the acoustic measurements. As a result, the anechoic time window was not as long as I usually enjoy, due to the presence of an early reflection from the ground, which will reduce the resolution of the measurement in the midrange. However, the amplitude of this reflection was not as great as I was expecting, due, I imagine, to the speaker's controlled vertical dispersion, with the radiating length decreasing with increasing frequency.

The Temptation goes very loud with only a small input signal—I estimated its voltage sensitivity as a high 91dB(B)/2.83V/m. However, its impedance (fig.1) is moderately difficult, with a minimum value of 3.1 ohms at 120Hz and a demanding combination of 4 ohms and a 42 degrees capacitive phase angle at 80Hz.

Fig.1 Dynaudio Temptation, electrical impedance (solid) and phase (dashed). (2 ohms/vertical div.)

The saddle at 28Hz in this graph's magnitude trace indicates the tuning frequency of the twin large-diameter ports. Note, however, that the traces are free from the small wrinkles and discontinuities that would otherwise indicate the presence of cabinet resonances. Fig.2, a waterfall plot derived from the output of a piezoelectric accelerometer fastened to the side of the midrange enclosure, confirms that the Temptation's cabinet is acoustically inert. I detected no other modes of any significance.

Fig.2 Dynaudio Temptation, cumulative spectral-decay plot of output of accelerometer fastened to midrange enclosure side panel. (MLS driving voltage to speaker, 7.55V; measurement bandwidth, 2kHz.)

Fig.3 shows the individual outputs of the midrange units, woofers, and ports, all measured in the nearfield; ie, with the microphone capsule (a Mitey Mike II) almost touching the diaphragms or the boundary between the port and the outside world. The port output is the broad bandpass centered on the 20-40Hz octave; while there is some output visible both around 160Hz and in the middle of the midrange, this is well suppressed. The sum of the woofer responses has its minimum-motion point at the port tuning frequency of 28Hz, as expected from the impedance plot, and it appears to roll off smoothly above 200Hz, with a first-order slope. The sum of the midrange units' response also appears to roll off below 200Hz with, again, a first-order, 6dB/octave slope. There are some small peaks and dips evident in what would otherwise be a smooth curve; as it is not possible to drive the woofers and midrange units separately, it's possible that these effects are due to interference from the output of the woofers.

Fig.3 Dynaudio Temptation, summed nearfield responses of the midrange units, woofers, and ports.

The complex sum of the individual responses is shown in the left portion of fig.4. The speaker's bass response is down by 6dB at the port tuning frequency, with a 24dB/octave rolloff below that point. Higher in frequency, the midrange and treble regions—assessed in the farfield with a calibrated B&K microphone and averaged across a 30 degrees horizontal window on an axis midway between the twin tweeters—are basically flat, though with a couple of small suckouts apparent in the low treble.

Fig.4 Dynaudio Temptation, anechoic response on axis midway between the two tweeters at 50", averaged across 30 degrees horizontal window and corrected for microphone response, with the complex sum of the nearfield midrange, woofer, and port responses plotted below 300Hz.