Bookshelf Speaker Reviews

Extended Test Bench

Frequency Response
• satellite 124 Hz to 20 kHz ±9.6 dB avg. 0°-30°, ±9.9 dB on-axis
• center 115 Hz to 20 kHz ±7.2 dB avg. 0°-30°, ±6.3 dB on-axis
• subwoofer 41 Hz to 285 kHz ±3 dB

Sensitivity (SPL at 1 meter/1 watt)
• satellite 79.6 dB
• center 81.5 dB

Impedance (minimum/nominal)
• satellite 3.8/6 ohms
• center 3.2/6 ohms

Bass output, subwoofer (CEA-2010A standard)
• Ultra-low bass (20-31.5 Hz) average: NA
20 Hz       NA
25 Hz       NA
31.5 Hz    90.1 dB
• Low bass (40-63 Hz) average: 109.8 dB
40 Hz      101.1 dB
50 Hz      109.2 dB L
63 Hz      114.4 dB L

Bass Limits
• satellite 91.6 dB at 80 Hz
• center 86.5 dB at 50 Hz

I measured the frequency response of the S 35 system using quasi-anechoic technique to remove the effects of reflections from nearby objects. I placed the satellite and center speakers atop a 2-meter-high stand, with the microphone placed at a distance of 1 meter. With the center speaker, I positioned the microphone directly in front of the tweeter. With the satellite, I experimented with different microphone heights to get the flattest response; what you see here was measured with the mike roughly even vertically with the center of the midrange driver. The curves you see in the chart represent the average of responses at 0°, ±10°, ±20°, and ±30°. Quasi-anechoic measurements were made with the grille for the satellite; the center speaker has no grille. All were smoothed to 1/12^th octave. Bass response of both speakers was measured using close-miking technique, with the mike positioned about 6mm from each midrange driver. These measurements were scaled appropriately, then spliced to the quasi-anechoic measurements at 300 Hz. Frequency response of the Sub 800 was measured using close-miking of the woofer and passive radiators; the results of all three measurements were summed to get the curve you see here. Results are normalized to 0 dB at 1 kHz for the speakers, +3 dB peak output for the subwoofer. All frequency response measurements were made with a Clio FW audio analyzer (in MLS mode for quasi-anechoic and log chirp mode for close-miking) then imported into a LinearX LMS analyzer for post-processing.

With the satellite, the interference and reflections at high frequencies caused by the unusual driver arrangement and speaker design make the response far from flat. The most obvious anomaly is that deep dip centered at 4 kHz. However, the response is pretty flat up to about 3 kHz, so in the most important part of the audio range this speaker might sound reasonably neutral. Off-axis response is consistent up to 10 kHz at angles out to ±45°. Adding the grille has negligible effect below 8 kHz, and boosts response between 8 and 12 kHz by about +3.5 dB.

The center measures better. On-axis, it’s ±3.2 dB from 400 Hz to 13 kHz, which is great. Off-axis, the expected interference effects between the two midranges start to appear, resulting in a dip centered at 1.4 kHz but spanning the spectrum from 1 kHz to 3.5 kHz that deepens as you move further off-axis.

For the satellite, minimum impedance is 3.8 ohms at 401 Hz with a phase angle of -7°. For the center, it’s 3.2 ohms at 241 Hz/-13°. Although the phase angle at the points of lowest impedance isn’t extreme, the relatively low nominal impedance (especially in the midrange) and the unusually low sensitivity of both speakers (measured on-axis outdoors, average output from 300 Hz to 10 kHz at 1 meter with a 2.83-volt RMS signal) suggests that this system should be combined with an A/V receiver offering decent power and not be used as replacement speakers for a home-theater-in-a-box system.

Although neither the satellite nor the center has much measured bass response, the bass power output measurements are surprisingly high for such tiny midwoofers, indicating that the speakers may blend better with the sub than their frequency response measurements suggest.

CEA-2010A output measurements for the Sub 800 were taken at 1 meter. This contradicts the CEA-2010A requirement to measure at 3 meters, but I’ve found that the output of small subwoofers like the Sub 800 at 3 meters is inadequate to achieve a usable signal-to-noise ratio for outdoor measurements. Averages are calculated in pascals. A “L” marks results where the number was determined by the sub’s internal limiter, not by exceeding the CEA-2010A max distortion threshold. Overall, output of the Sub 800 is comparable to that of the small subwoofers included with typical 2.1-channel soundbar systems. Combined low-pass function of the driver and crossover is -21 dB/octave.—Brent Butterworth