Opera Callas loudspeaker Measurements
John Marks was impressed by the Callas stand-mounted loudspeaker from Italian manufacturer Opera Loudspeakers. The Callas is unusual in that it has tweeters mounted above and below its single SEAS magnesium-cone woofer. The conventional wisdom is that the higher the frequency, the closer together you need to position drive-units covering the same passband if you want to avoid unwanted off-axis lobing. Not only does the Callas break that rule, it also has three rear-firing tweeters mounted vertically in-line on its rear panel, below a pair of reflex ports.
Nevertheless, JM was impressed by what he heard. Driving the Callases with Unison Research's S6 amplifier, he wrote that he "was pleasantly surprised by both the dynamic capability and the bass extension of the Callas-S6 combo. Unlike with many two-way speakers, I never got the sense during most normal listening (as distinct from playing very loud to impress myself or friends) that there was 'almost' enough bassthere really was enough bass. . . . There was never a sense of too much treble unless the recording itself was too hot; the treble and midrange were very well integrated."
Intrigued, I asked JM to ship the speakers to me when he was done with them, so that I could run them through my test protocol. I used DRA Labs' MLSSA system (www.mlssa.com) and a calibrated DPA 4006 microphone to measure the Opera Callas's frequency response in the farfield, and an Earthworks QTC-40 for the nearfield responses. The Callas is not very sensitive, its estimated voltage sensitivity a low 84.3dB(B)/2.83V/m. However, it is a relatively easy load for the partnering amplifier to drive, its impedance magnitude (fig.1, solid trace) dropping below 6 ohms only in the low midrange and high treble, and reaching a minimum value of 3.7 ohms at 165Hz.
The traces in fig.1 are free from the small wrinkles that would imply the existence of cabinet resonances. However, when I investigated the vibrational behavior of the enclosure, I found a fairly strong resonance at 586Hz on the sidewalls (fig.2). This is probably high enough in frequency to have no negative effect on sound quality. There was also a lower-level mode at 350Hz on the top of the cabinet, but again, this is unlikely to affect the speaker's sound, given the small radiating area affected by the resonance.
The saddle centered on 42Hz in the impedance-magnitude trace in fig.1 suggests that this is the tuning frequency of the two small ports on the cabinet's rear panel. Measured in the nearfield (fig.3, blue trace), the woofer output's minimum-motion notch (ie, where the back pressure from the port resonance holds the cone stationary) is indeed at 42Hz. The ports' output (red trace) peaks in textbook fashion between 30 and 65Hz, but there is also a vicious-looking peak visible at 700Hz. I could hear this peak as a slight whistle imposed on the sound of pink noise when I stood behind the speaker, but it's fair to note that JM didn't remark on any coloration in the upper midrange that could have resulted from this port behavior.
Higher in frequency in fig.3, the Callas's farfield response is impressively flat from the midrange through the mid-treble, though with a vestigial peak at the approximate frequency of the ports' midrange resonance. Though JM didn't feel that the speaker's treble sounded exaggerated, there is actually a significant excess of energy between 8 and 24kHz. To some extent, this will be compensated for by the speaker's increasing directivity in the horizontal plane in this region (fig.4). However, the room's reverberant field will be boosted at high frequencies by the fact that the farfield response at the speaker's rear (fig.5) is also boosted in the top octaves, thanks to the three rear-firing tweeters. An upper-midrange peak can be seen in this graph, though peculiarly, it is a little lower in frequency than the peak in the ports' nearfield output.
In the vertical plane (fig.6), there is the inevitable pattern of off-axis peaks and dips that result from the spaced tweeters. This speaker definitely needs to be listened to on the woofer axis to get the most even balance treble balance, though moving slightly above or below that axis will pull down the top octave a little.
The Opera's step response on the woofer axis (fig.7) indicates that all three drive-units on the front baffle are connected with positive acoustic polarity, and that the outputs of the tweeters lead that of the woofer by a small amount. Other than a very small amount of delayed energy at 3.2kHz, probably due to a residual mode in the woofer's cone, the Callas's cumulative spectral-decay plot is superbly clean (fig.8).
Again from JM's review: "To sum up the Opera Callas: luscious midrange, sweet treble, large soundstage, surprising bass, eminently listenable; Class B (Restricted Extreme Low Frequencies)." Other than that resonant mode in the ports' output and the excess of high-treble energy, I would say that its measured performance confirms JM's high opinion of the Opera Callas's sound quality.John Atkinson