MartinLogan CLS loudspeaker Page 2

Without a doubt, the CLS is the most revealingly transparent loudspeaker I have used. The Wilson Audio Beethoven violin sonata, for example, was presented spatially the best I have heard, the soloist being obviously around four feet in front of the piano, and the piano image naturally sized.

Tonally, the midrange is neutral, voices being rendered without undue coloration, but the lower midrange is slightly depressed. The exact amount of upper midrange energy seems to depend very much on listening height: my chair puts my ears about two-thirds of the way up the panels; lower than that and the midrange depresses, leaving the low treble a little exposed. The bass was a little lightweight, there being a lack of upper bass (though there is a slight bump in the response that makes male voices a bit chesty). The midbass was fine, but there was then a hole before the relatively sparse low bass came in. (In-room, spatially-averaged measurements using pink noise showed that the speakers rolled off rapidly below 60Hz.)

The low frequencies are also a little "slow": on the "Nature Boy" track on the Norwegian Radka Toneef album, which features a Scandinavian ice maiden, grand piano, and air-conditioning—all immaculately recorded—the piano appeared almost to anticipate the beat, the bass being definitely a little sluggish compared with the treble, which is quite "fast."

Above the midrange and bass, we now start to get to an area of tonal performance where I was less happy. It both sounds and measures as if there is a broad plateau between 1kHz and 4kHz, which adds to the feeling of transparency but gives the speaker a merciless quality. (I also felt that there was something weird going on at very high frequencies, around 14kHz.) The sound is detailed, yes; transparent, certainly; but there is a feeling of "glare" which, no matter how effective at revealing recorded detail, detracts from musical enjoyment. It is almost as if you are reading a book with fine print by the light of a 500W bulb. You are made aware of every detail, including the intricacies of the typeface and the texture of the paper, but you become a little fatigued. You become so aware of the "how" that you lose interest in the "why."

The longer I used the speakers, the less this character was apparent, and changing to the D-250 considerably improved things in this respect. Listening to the CLSes from a greater distance than the 6-9' I used will also help. But there was still some residual forwardness that gives the sound an overall "cold" balance. Pianos have too much "snarl," rosin noise on strings is accentuated, and recorded sibilance, as featured heavily on my favorite Clannad album on RCA, can often become intolerable.

Another aspect of the CLS sound bothered me more in the long term: a lack of dynamic range in the lower mids and upper bass. We are at 7000' in Santa Fe, so I was prepared to accept some compromise on ultimate loudness. The diaphragms have to move further to generate high SPLs in the thin New Mexican air, and run out of excursion sooner than at sea level. Nevertheless, I found that measured peak levels in the high 90s would cause random ticking noises from the diaphragms, particularly with recordings having a predominance of lower midrange energy (footnote 1).

The Cyprien Katsaris recording of the Liszt "Eroica" transcription has all its energy centered between 200Hz and 800Hz, and levels above 95dB (flat, average) caused the CLSes to misbehave. The dynamic range limitation seemed less of a problem when I switched to the Mk.II Krell KSA-50 from the Robertson 4010, but it was disappointing to not achieve the admittedly high levels at which I occasionally like to listen.

Could the CLS's load impedance explain what I was hearing, electrostatics in general having a bad name for being pigs to drive? Rated by M-L as a nominal 6-ohm speaker, the CLS actually varies considerably, reaching a maximum of 40 ohms at 1kHz but dipping below 5 ohms both in the mid and low bass, and above 4kHz. The impedance minimum is 2 ohms at 16kHz. This might go some way to explaining the subjective problems in the bass and high treble, particularly if the nature of the impedance at these minima is highly capacitive. I would have thought, however, that all the amplifiers I used would have no problem supplying the necessary current.

After a couple of weeks of living with the speaker, I had no option but to call M-L's Gayle Sanders to discuss what I was hearing from the CLSes. He nodded his head sagely when I mentioned the bass quality: "The adhesive we use can soften in transit, leading to too much midbass and a corresponding dip in the upper bass." He counseled running a heat gun up and down the edges of the panels, which should both even out the upper-bass response and increase the speaker's capacity for producing high sound levels. We then got on to the subject of the residual brightness that was bothering me. Again he nodded his head, and mentioned that the latest version of the electronics module had a modification to ameliorate this character. Apparently the original version I had could introduce amplifier instability at ultrasonic frequencies, due to the capacitive nature of the load up there in the stratosphere. A new set of modules would leave Kansas via Federal Express as soon as the Munchkins had finished applying the shrink-wrap.

The new bits arrived—serial numbers were 1870/1871 compared with the original modules' 1562/1563—and I duly set the speakers back up, having first played my hair-drier up and down the edges of the panels.

Footnote 1: As a reference (for the thin New Mexico air), the Quad ESL-63s "crowbarred"—or arced, when the protection circuitry wasn't working effectively—predictably on 96-97dB peaks, depending on the frequencies being reproduced (the lower the worse). The CLSes did 2-3dB better, again with problems showing up sooner at low frequencies.—Larry Archibald