Bookshelf Speaker Reviews

Bottom Line

RSL has a very, very good small speaker system here. It looks like any number of other similar small speaker systems, but it performs, without fuss, at a level higher than most at this price point. The absence of upper-bass excess is the key to the RSL’s success, which makes it one of the few pint-size-sat systems that can play in the big sandbox without apology. At $2,075 for the complete system, the price is a good deal more than many similarly petite competitors, but disregard physical size; for this kind of performance, it’s a bargain.

Test Bench

Frequency response

• satellite 88 Hz to 20 kHz ±9.5 dB
• center 85 Hz to 20 kHz ±3.9 dB
• subwoofer 30 to 120 kHz ±3 dB

Sensitivity (SPL at 1 meter/1 watt)
• satellite 84.5 dB
• center 86.9 dB

Impedance (minimum/nominal)
• satellite 6.2/10 ohms
• center 3.6/10 ohms

Bass output, subwoofer (CEA-2010 standard)
• Ultra-low bass (20-31.5 Hz): 108.0 dB
• Low bass (40-63 Hz): 117.9 dB

Bass limits
• satellite 80 Hz at 81 dB
• center 80 Hz at 84 dB

I measured the CG4 satellite and the CG24 center without grilles at a distance of 1 meter, which was adequate with these small speakers to incorporate the contributions of all drivers and cabinet diffraction. Both sat atop a 6-foot stand to give quasi-anechoic results down to 260 Hz. Because it’s designed to be used both horizontally and vertically, I measured the CG24 positioned both ways; the vertical position measurement is shown in the graph.

The curves in the graph show an averaged response from 0° to 30°, smoothed to 1/12th of an octave. I close-miked the woofers and ports of both speakers, then scaled and summed the results to get each speaker’s bass response. I then spliced the bass responses to the averaged quasi-anechoic responses to produce the curves you see here.

To measure the subwoofer’s frequency response, I used a ground-plane measurement at 2 meters.

The satellite and center responses are normalized to 0 dB at 1 kHz, the subwoofer normalized so that its peak response shows as +3 dB.

Up to 10 kHz, both of the main speakers measure beautifully: ±2.6 dB for the CG24 center and ±3.1 dB for the CG4 satellite. When you take the full frequency response up to 20 kHz into account, though, the CG4 shows much more treble rolloff: It’s ±9.5 dB versus ±3.9 dB for the CG24. This is surprising because it seems that both speakers would measure roughly the same above 10 kHz; I confirmed the result by re-measuring the first CG4 sample and then measuring a second sample. (I wonder if those tweeters aren’t as alike as they look.) Both speakers have an unusual dip right in the middle of the woofer’s range, at 560 Hz. It’s –3.8 dB for the CG24, –5.9 dB for the CG4. Effects of the grilles were minor: a decrease of about 1 dB in the treble from 5 to 9 kHz, and a narrow dip of about –4 dB centered at 15 kHz.

Used horizontally, the CG24 measures unusually, but well. There’s more treble rolloff than expected in the averaged 0° to 30° response: Compared with the vertical response, it’s –1 to –3 dB from 1 kHz to 10 kHz, –2 to –5 dB above 10 kHz. A –20-dB dip appears at 1.8 kHz at 45° off-axis, but the dip mostly smoothes out at 60°. For a two-way, woofer-tweeter-woofer center speaker, this is very good on- and off-axis performance; most show a lot of interference between the woofers off-axis.

Impedance of the RSL speakers runs high enough that the speakers can be used safely with practically any receiver. Both the CG24 and the CG4 have a nominal measured impedance of 10 ohms, which is fairly flat across the audio band. The CG24 drops to a minimum impedance of 3.6 ohms at 184 Hz, and the CG4 to 6.2 ohms at 290 Hz. (Strangely, the CG24 is rated at 4 ohms nominal.) Impedance phase shift is a little more extreme than average, but the impedance at maximum phase shift is so high that it won’t matter; max phase shift for the CG24 is +60° at 940 Hz/12 ohms, and for the CG4 it’s +48° at 940 Hz/16 ohms.

Average sensitivity from 200 Hz to 10 kHz at 1 watt (2.83 volts) at 1 meter is okay for the CG24 at 86.9 dB but a little low for the CG4 at 84.5 dB. I’d recommend using a receiver rated to deliver at least 70 or 80 watts per channel with the CG4; use whatever you want with the CG24.

I measured the output of the Speedwoofer 10 using the CEA-2010 method. Measurements were made on the ground at 2 meters; 6 dB was added to the results to approximate measurements at 1 meter. This Speedwoofer 10 is fairly powerful for a modestly sized 10-inch sub, delivering an average of 108.0 dB from 20 to 31.5 Hz and 117.9 dB from 40 to 63 Hz. Max output is 119.7 dB at 63 Hz. Output drops dramatically below 25 Hz, but it’s still usable at 20 Hz, where it hits 101.2 dB. However, the noise from the front slot port is high at all frequencies from 20 to 40 Hz. CEA-2010 test tones are particularly revealing of port noise, though, and I expect this noise will seldom be noticeable with most program material. With the crossover set to 100 Hz (one of the few marked settings on the crossover frequency knob), response was down 5.7 dB at 100 Hz and combined low-pass rolloff function was –16 dB/octave. — Test Bench by Brent Butterworth