Harbeth HL Compact 7ES‑3 Loudspeaker

Perfecting the BBC school of design

I don’t know Alan Shaw personally, but on the basis of Harbeth, the company he has owned for many years, and the speakers he has authored, I’d guess that one of his missions in life is to preserve and perfect the BBC school of speaker design.

As early as the 1940s, the British Broadcasting Corporation engaged in systematic research of loudspeaker performance with the eventual aim of designing and manufacturing speakers for monitoring broadcasts and recordings. Flat frequency response, natural tonal balance, low coloration and distortion, and unit‑to‑unit consistency were the design parameters, with the goal to accurately reproduce voice and music in professional and domestic (as opposed to anechoic) settings. From the mid‑fifties through the seventies, most well‑regarded British speakers, among them Quads, KEFs, and B&Ws (and those of the so‑called “New England” school in America, such as Acoustic Research and KLH), basically subscribed to the BBC philosophy, and several classic products were direct offspring of its research, including the Spendor BC‑1, BC‑3, and SP1/2, and the Rogers LS3/5a. Many audiophiles would also put Alan Shaw’s Compact 7 in this group.

Introduced in 1988, the original HL Compact 7 was the first model to use Harbeth’s Radial, a patented, propriety synthetic compound—now widely used in the company’s woofers and midrange drivers—that Shaw considers far superior in resolution and low coloration to the polypropylene that Harbeth’s founder, Dudley Harwood, pioneered at the BCC. Shaw introduced a revised Compact 7 in 1999, with incremental improvements. When he turned his attention to it yet again last year, the result was to advance an already excellent design to an even higher level. How much higher? I’ll make a naked grab for your attention: between about 80Hz and 10kHz there is nearly nothing wrong with the frequency response of the 7ES‑3. A couple of absolutely miniscule glitches and anomalies remain, but they do not matter, as they are scarcely audible even with critical listening. Indeed, for sheer neutrality the 7ES‑3 rivals the Harbeth Monitor 40 (my colleague Robert Greene’s reference), and in the 2kHz– 4kHz range even trumps my reference Quads—by a very tiny margin to be sure, but a deuce trump is still a trump.

Pulling out every source by which I judge neutrality and tonal accuracy, I was soon scratching my head for adverse characteristics on which to pin adjectives. As with Quads, each and every recording sounds unto itself, not like any other. John Langstaff ’s Christmas Revels [Music Sales Corp.]—I am writing this during the holidays—is as raucously joyous a celebration of the season as you’ll hear anywhere, with the most motley collection of olde instruments since David Munrow in his prime, with sticks, swords, a sackbut, a rebec, a serpent, a lyzarden, a hurdy‑gurdy, something suspiciously reminiscent of P.D.Q. Bach’s notorious windbreaker, and a tubby tympanum that really should sound a little tubby. At the outset a lonely recorder wanders forward from deep in stage left, then crosses to disappear stage right, his mournful tune reminding us that this festive season is also darkened by the shortest days, the longest, coldest nights. But soon other revelers appear, called to attention by Room (the first character of the Mummers’ play Saint George and the Dragon), who bangs on a can, exhorting all to make merry with song and dance, his voice ricocheting off the side and rear walls. The 7ES‑3s present this spectacle so vividly as to recall the original meaning of stereo, i.e., solid, the positions and movements of the players practically diagrammable, if, that is, you can even think about such things faced with so intoxicating an entertainment.

Equally infectious is Joel Cohen’s Nonesuch classic Sing We Noel: The several voices that perform solos, duets, trios, or in small groups are clearly distinguishable one from another. On one extraordinary track (5), a speaker is firmly in the left channel, reading the Annunciation, while in the right he is answered by a soprano singing a related 13th-century carol. The antiphony is easy to reproduce; what isn’t, but which the 7E‑3 reproduces perfectly, is the rich church‑like ambience, which is clearly audible across the soundstage. Eventually instrumentalists and other singers join in to fill the space, yet they should never obliterate our sense of that enveloping acoustic.

It’s recordings like these that make me question the priorities of audiophiles who relegate accuracy of timbre to secondary status. How are the richness and color of instruments, voices, ensembles, and textures to be reproduced in all their infinite variety and beauty if a loudspeaker has less than accurate reproduction of timbre? What do dynamics, imaging, detail, transparency, and the like matter if voices and instruments don’t sound like themselves?