Floor Loudspeaker Reviews

I used DRA Labs' MLSSA system and a calibrated DPA 4006 microphone to measure the Magico's impedance and farfield frequency response, and an Earthworks QTC-40 for the nearfield and spatially averaged room responses. The sheer bulk of the Q5—it weighs almost 400 lbs—precluded my being able to place it on my Outline turntable for the acoustic measurements. I therefore performed the quasi-anechoic measurements with the speaker sitting on a dolly in Michael Fremer's driveway (scroll down the page). The inevitable reflection of the speaker's output from the ground between it and the microphone will therefore reduce the resolution of the measurements in the midrange; it was also not possible to do a full set of lateral-dispersion measurements, due to the need to keep to a minimum the time the speaker was left standing in direct sunlight.

The Q5 has a rated sensitivity of 87dB. However, my estimate was lower than this, at an estimated 84dB(B)/2.83V/m, which is also lower than average. The speaker is also fairly difficult to drive, with an impedance that drops below 4 ohms in the high treble, the lower midrange, and the upper bass (fig.1). As well as minimum values of 2.75 ohms at 56Hz, 3 ohms at 200Hz, and 2.8 ohms at 40kHz, there is an amplifier-crushing combination of 3.85 ohms and a –56° capacitive phase angle at 45Hz. This speaker really does need to be used with powerful solid-state amplifiers to sound its best, I feel, such as Michael Fremer's Musical Fidelity Titan. The traces in fig.1 are free from any wrinkles that would indicate the presence of cabinet vibrational resonances; listening to the cabinet walls with a stethoscope while I swept a sinewave tone up and down in frequency, I could detect only a small degree of liveliness at 418Hz. The cabinet's heroic, all-aluminum construction is obviously effective at minimizing vibrational resonances.

Fig.1 Magico Q5, electrical impedance (solid) and phase (dashed). (2 ohms/vertical div.)

The single impedance peak in the bass in fig.1 indicates a sealed-box alignment for the bottom two woofers (which behave identically) tuned to a low 29Hz. The red trace in fig.2 shows the output of these drivers, measured in the nearfield. It peaks between 30 and 150Hz, crossing over to the third, topmost woofer (blue trace) above that range and rolling off with a steep slope broken only by a well-suppressed peak in the midrange. The top-most woofer crosses over to the midrange unit (green trace) at 500Hz, meaning that the lower unit is responsible for handling most of the fundamental tones of the male human voice, with the midrange unit reproducing harmonics. The midrange unit's response is flat within its passband, but the black trace in fig.2 implies that the beryllium-dome tweeter is balanced a little high in level. Its output is flat, however, and extends at full level to the upper limit of this graph, at 30kHz.

Fig.2 Magico Q5, anechoic response on tweeter axis at 50", averaged across 30° horizontal window and corrected for microphone response, with nearfield responses of midrange unit (green), upper woofer (blue), and lower woofers (red) plotted below 750Hz, 1kHz, and 550Hz, respectively.

In the vertical plane (fig.3), the Magico's response doesn't change significantly over a ±5° window centered on the tweeter axis, which is 40" from the floor. Though I didn't measure the Q5's lateral dispersion, its use of a fairly wide baffle means that the speaker's top-octave output does drop off to the speaker's sides. This could be seen in the individual measurements taken to produce the Q5's spatially averaged response in MF's listening room (fig.4, red trace). Nevertheless, there is a little too much mid-treble energy apparent in-room, especially when compared with MF's Wilson MAXX 3s (fig.4, blue trace). As Michael noted, the Q5's tweeter was "not shy or polite" and the presence region was slightly forward. While the Magico speakers don't excite the low-frequency modes in Michael's room quite as much as the Wilsons do, they do have a slight lack of energy between 200 and 400Hz, which is where the Wilsons have an excess. Could this explain why the Q5s didn't cause Michael's stomach to churn as much as he was used to with cello recordings?

Fig.3 Magico Q5, vertical response family at 50", normalized to response on tweeter axis, from back to front: differences in response 15–5° above axis, reference response, difference in response 5° below axis.

Fig.4 Magico Q5, spatially averaged, 1/6-octave response in MF's listening room (red), and of Wilson MAXX 3 (blue).

Turning to the time domain, the Magico's step response on its tweeter axis is shown in fig.5. All five drive-units are connected with positive acoustic polarity, and the decay of each unit's step blends smoothly into the start of that of the next lower in frequency, suggesting optimal crossover design. To generate the Q5's farfield cumulative spectral-decay plot (fig.6), I had to aggressively window the impulse response to eliminate the first reflection of the speaker's output from the ground, which occurred just after the 7.5ms limit of fig.5. That this graph doesn't show as much detail as usual is indicated by its dotted region. But other than a slight amount of low-level hash at the top of the midrange unit's passband, the decay is superbly clean overall, correlating with MF's finding the Q5's treble to sound free from any grain or edge.

Fig.5 Magico Q5, step response on tweeter axis at 50" (5ms time window, 30kHz bandwidth).

Fig.6 Magico Q5, cumulative spectral-decay plot on tweeter axis at 50" (0.15ms risetime).

Given that its bulk limited the lower-frequency resolution of the measurements, the Magico Q5 measured very well indeed.—John Atkinson