Spica Angelus loudspeaker Angelus blind test 1993

The Angelus was reviewed in a blind panel test in September 1993 (Vol.16 No.9):

Equipment & Listening Panel
As in our past tests, the playback system was of better high-end quality than would usually be the case with loudspeakers in this price range. It was chosen to give the loudspeakers their best shots. A C.E.C. TL 1 transport and Mark Levinson No.35 D/A converter (linked by a Kimber KCAG digital interconnect) was the program source, feeding a Krell KSA-250 power amplifier. As already noted, the preamp was the Rowland Consummate. The processor-preamp link was TARA Labs Master RSC, and the pre-power-amp connection was Cardas Hexlink unbalanced. The loudspeaker cable was Straight Wire Virtuoso Gold. The loudspeakers were bi-wired if their design permitted it.

None of the panel members attended all four sessions. John Atkinson, J. Gordon Holt, and Guy Lemcoe each attended three, Robert Harley and Dick Olsher two, and myself one. A local audiophile, Michael Mandel, was invited to participate the third day to fill out the panel. I changed the loudspeakers around on the first three days; RH took over the test-running chores on the fourth day so I could participate as a scorer. JA, who missed Day Three, also sat in on the final day.—Thomas J. Norton

All the loudspeakers were broken-in for at least 50 hours with moderate-level pink noise. They were concealed, one pair at a time, behind a grillecloth that completely concealed the short end of the 20' by 15.5' by 9' Stereophile listening room. Only the identity of the reference Vandersteens was revealed to the listeners, who left the room between sessions so that the loudspeakers could be changed. Except for the Vandersteens, the grilles on each loudspeaker were removed to eliminate what would have been a double layer of grillecloth between drivers and listeners.

Two loudspeaker locations behind the grillecloth screen were chosen, one close to the wall and one farther into the room. Which position to use and whether to toe-in the units were determined by the owner's manuals and by measurements made prior to the sessions by JA. One would normally prefer to measure after listening, but several measurements were vital for proper setup of the loudspeakers. Those included horizontal dispersion (for toe-in), location of the vertical listening axis, and sensitivity.

The Spica Angeluses were placed nearer the rear wall. There are inevitable placement compromises in panel sessions because only one, possibly two, listeners can sit in the best listening seats. There were three rows of listeners (one in the first row, two in the second, and three in the third), and those who participated more than one day occupied the same seats each day. Still, the "environment" around each listener varied somewhat on different days, depending on whether nearby chairs were occupied. And only two listeners could occupy center seats—which inevitably affected soundstaging impressions. For the record, JGH always occupied the front-row seat, JA the center-rear.

Proper vertical listening axis is a much ignored setup criterion, but it is critically important for many loudspeakers. We found most of our candidate loudspeakers to be too short for a typical seated listener. The director's chairs we used for the sessions placed a listener's ears 43" from the ground, which is slightly higher than the domestic average (about 38"). But even with lower chairs, most of our loudspeakers would have had to be raised or tilted back for proper orientation. Tilt-back is not a practical alternative with three rows of listeners—the geometry with tilt-back would be correct at only one distance from the loudspeakers. So combinations of heavy bases, MDF shims, spikes, and cones were used to raise the loudspeakers. The Spica required the least adjustment—it would be reasonably correct at a normal sitting height. (Changing the woofer's proximity to the floor does alter a loudspeaker's bass balance somewhat. We considered this a less serious problem than the response deviations almost always encountered on the wrong vertical listening axis.)

Prior to the sessions, I chose an optimal listening level for each program selection. The Spicas were used for this "level referencing" because our experience indicates that this speaker is sensitive to bass overloading, something we especially wanted to avoid. Bass overloading would perhaps have given a clue to the Spicas' identity, or it might have damaged them, putting a serious crimp in the tests. I converted the chosen level for each musical selection to an equivalent level for each of the other loudspeakers, using JA's measurements for B-weighted sensitivity. The level adjustments during each session were made with the Rowland Consummate preamplifier, which is ideally suited to this task because of its unique front-panel reference readout and its subtle, repeatable (and previously measured) level change with each step. The Consummate and its less expensive sibling, the Consonance, are the only pieces of consumer electronics equipment known to me that can make such adjustments on anything like a real-time basis.

Discussing impressions was not allowed, and we advised the panelists to avoid any verbal or bodily clues that might betray their reactions to the sound and influence others.

I organized and ran the tests, compiled the data and comments, and wrote up the results. JA conducted and wrote up the measurements, and Robert Harley provided the introductory descriptions of each loudspeaker.—Thomas J. Norton

The envelope, please
The data were compiled in a number of different ways. Panelists scored each piece of program material heard over each unknown loudspeaker from 0 to 10, 0 indicating the worst reproduction imaginable and 10 the best possible, the real thing. The Vandersteen 2Ce was arbitrarily specified to be a "5" across the board to establish a plane of reference. Space on the score sheets was also provided for written comments. The results for each loudspeaker are given below, just prior to the panel members' comments. The four scores in parentheses are for each of the four days. Keep in mind that there were four panelists on Day One, five on Day Two, four on Day Three, and two on Day Four. Because of this, the overall score—the figure not in parentheses—is not the average of the total score for each of the four days, but rather the average of the 15 total individual scores compiled over four days. Since the order in which a loudspeaker is auditioned may affect its score—speaker A, auditioned following bass-light speaker B, will be rated higher for bass performance than when it follows bass-heavy speaker C—our order was shuffled each day.

Whenever I refer to a listener having rated a model either above or below average, I mean relative to that individual's personal average score, not the group average.—Thomas J. Norton

Spica Angelus: $1275/pair
In production since late 1987, the Angelus has been the subject of much discussion and praise in Stereophile. The loudspeaker uses an 8" plastic-cone woofer and 1" soft-dome tweeter in an unusually shaped enclosure. The cabinet is a pair of trapezoids, with the top one inverted in relation to the bottom one. Together, the two sections form a vertical mirror image. This geometry reportedly has many sonic advantages over a conventional box design. The top section houses the drivers, and the baffle is covered by thick felt padding to reduce diffraction effects. The 2.5-ft3 sealed enclosure is made from 3/4" MDF and is braced internally.

The computer-derived crossover uses a fourth-order network on the woofer and a first-order on the tweeter circuit, said to give a time-coherent impulse response and a flat frequency response. Both drivers are connected with positive polarity, and the crossover point is a rather high 3.4kHz. Film capacitors are used extensively in the crossover. A single pair of five-way binding posts provides signal connection. Finish is natural oak or walnut veneer on the front and sides, with a black grille covering the top section's baffle.—Robert Harley

Blind listening-panel score: 4.20 (4.78, 3.91, 4.18, 3.83)
JA had some serious misgivings about the Spica Angelus on all three days, downgrading it accordingly. He found the Duruflé "lean-sounding, bass lightweight overall but some organ notes stick out." He also commented on a "chalky upper midrange." He commented a number of times on a "lispy" quality, noting a mid-treble emphasis. He thought the drum solo had "good clarity," but found a "hooty" coloration to the snare drums. On the Albeniz, he found the Spica "lean'n'mean, attempts to reproduce bass drum but fails." Surprisingly, given the Spica's reputation, he found the image depth rather restricted.

GL echoed these sentiments, but did have positive comments. He found the drum solo "fast and dynamic," with "good control and fine spatial rendering." And he liked the reproduction of the Fairfield Four, with the "voices delineated across the stage [and] articulate, [with the] words easily understood." On the Duruflé, he found the "vocal separation quite nice, [with] good soundstaging." But he also found the sound tipped-up. On the Duruflé, a "lack of impact in the low bass" kept him from rating the Spicas higher. He also disliked the highs, finding them "tipped-up" and "splashy" on the Jennifer Warnes, "aggressive yet not that involving" on the Eddy Louiss, and "blatty [and] raspy" on Luke and the Locomotives.

JGH, however, was much more positive. He gave the Spicas a strong endorsement on the first day, finding the drum solo "tight, detailed, and articulate," with "good impact and good timbres." He thought the Fairfield Four "almost real, [with a] wide stage [and] good imaging." He called the Jennifer Warnes "magical!" and "hard to fault." And he felt the Britten had "good depth, breadth, articulation, impact, almost goose-bumpy." His ratings dropped off considerably on Days Two and Three, but his feelings were strong enough on Day One to push the Spicas to the top of his ratings pile.

RH rated the Spicas somewhat below average. He thought the treble "aggressive" and "raspy" on the drum solo, sibilant on the vocals, and "bleached and thin," in need of more midbass weight. He noted "good extension" on the Eddy Louiss, but expressed "relief" when it was over. While he thought the Albeniz had a "nice soundstage," he thought the Spicas had "trouble on [the] bass-drum whacks."

DO placed the Spicas almost in the middle of his dance card, but his written remarks were mostly critical. He thought that the "balance serves [the] female voice well" on the Warnes and noted a "good drive," but remarked that the Lesley was "devoid of [the] proper sheen." He remarked on several occasions on the "coarse" and "peaky" treble, finding the "tonal balance too thin." On the Albeniz, he heard a "decent orchestral drive," but felt the balance to be "all wrong."

MM—though he thought the sound "somewhat bright"—felt that the music "came through very well." He preferred the Spicas to the Vandersteens on several cuts.

I rated the Spicas just below average, noting a "soft-dome" grain in its high-frequency performance, "slightly toppish" overall "but not pushy in the low- to mid-treble." I remarked on several occasions on a "slightly boxy" quality to the sound, and was also bothered by the "slightly lean" and "up-tilted" balance that lacked the weight needed to fully flesh out instruments and voices.

In summary, the Spicas were criticized for their lean, rather bright balance. This suggests that a subwoofer—if carefully matched—might well restore the balance. Or perhaps the Spicas might be better suited to a room in which their lean balance won't seriously excite the low-frequency room modes. Of course, it also might be true that the basic design of the Angelus—now six years old with some intermediate crossover modification—is due for a freshening-up.—Thomas J. Norton