Review: KEF LS50 speaker

Measurements

Frequency response
46 Hz to 20 kHz ±3.2 dB on-axis, ±3.3 dB avg 0°-30°

Impedance (minimum/average)
3.7/8 ohms

Sensivitity (1 watt/1 meter)
84.0 dB

Bass limits
93.5 dB at 40 Hz

To measure the LS50’s quasi-anechoic frequency response, I set it atop a 2-meter stand and placed the microphone at a distance of 1 meter. (Quasi-anechoic measurements eliminate reflections from surrounding objects to simulate measuring in an anechoic chamber.) The microphone was placed on the same axis as the tweeter. I ran a ground-plane measurement at 1 meter to get the bass response. The curves in the graph here shows the bass response spliced at 200 Hz to the 0° on-axis measurement and the average of quasi-anechoic measurements taken at 0°, ±10°, ±20°, and ±30°. I used a Clio FW analyzer in MLS mode for the quasi-anechoic measurements and log chirp mode for ground plane, powering the LS50 with an Outlaw Model 2200 amplifier. The quasi-anechoic measurements were smoothed to 1/12^th octave. The blue trace shows the 0° on-axis response, while the green trace shows the averaged response.

On-axis, the LS50’s frequency response measurement is very good. Its off-axis results are the best I can ever remember seeing. Even way out at ±60°, the treble roll-off is mild and there are no dips in the midrange. (It’s so good I added an extra graph to show it.) Of course, all of KEF’s Uni-Q speakers have great off-axis response, but I’ve measured lots of them over the years and none come close to the LS50’s off-axis performance. In fact, its off-axis response is so different from what I’m used to seeing that I’m not 100% certain what the implications would be for acoustical treatment, speaker positioning, etc. In my treated listening room, I’m sure much of this off-axis treble energy was absorbed or diffused, so I didn’t hear anything particularly unusual in the highs, but I wonder what the LS50 might sound like in a highly reflective room.

I measured the LS50’s sensitivity by taking the average quasi-anechoic output with a 2.83-volt (1 watt @ 8 ohms) signal from 300 Hz to 10 kHz. The 84 dB result is fairly average; expect another +3 dB or so in real-world, in-room applications. Impedance averages 8 ohms, and drops to a low of 3.7 ohms at 240 Hz/+5° phase. In short, the LS50 might be a little tough for a low-powered, single-ended tube amp to drive, but I doubt any other amp would prove inadequate to the task.

Bass limits were measured using CEA-2010 technique: ground plane at 3 meters, scaled up to a 1-meter equivalent. With decent output even down to 40 Hz, the LS50 can easily mate with any subwoofer.

Bottom line

There’s no perfect speaker for everybody. Many listeners would like more bass than the LS50 can muster; they’d probably prefer an affordable tower speaker. Some listeners might want a more spacious sound; they’d probably prefer a panel speaker like a Magnepan or MartinLogan. But those who want clean midrange without annoying colorations probably can’t do better at this price than the LS50.