Audience ClairAudient The One loudspeaker Measurements
Sidebar 3: Measurements
For reasons of consistency, I measured the Audience ClairAudient The One in free space, mounted on a tall stand, as I do all bookshelf loudspeakers that pass through my lab. In use, however, the Ones will be in close proximity to one surfacethe top of the user's deskand perhaps another (the wall, if any, behind the desk). This difference between measurement conditions and those of actual use should be kept in mind throughout this sidebar. However, I note that Bob Reina did do some of his listening with the speakers mounted on 24"-high stands, and far enough away from the room's walls that the latter would give no boundary reinforcement of the lower frequencies. I used DRA Labs' MLSSA system and a calibrated DPA 4006 microphone to measure the Audience One's frequency response in the farfield; and, for the nearfield responses, an Earthworks QTC-40 mike.
My estimate of the Audience One's voltage sensitivity was 83dB(B)/2.83V/m, which is well below the norm. With a specified maximum power handling of 25W, you would have thought this low sensitivity would confine the One to nearfield listening, as on a desktop. However, again note that BJR successfully used the speakers for conventional farfield listening. The Audience's plot of impedance magnitude and electrical phase against frequency is shown in fig.1. The speaker offers basically a 6 ohm load to the partnering amplifier, and without a crossover, the phase angle is low throughout the midrange and treble. Although the phase angle does become increasingly positive in the top two octaves in this graph, due to the single drive-unit's voice-coil inductance, the One is basically still an easy load.
There is a slight bump between 200 and 300Hz in the impedance-magnitude trace in fig.1, but investigating the enclosure wall's vibrational behavior uncovered nothing of concern in this frequency region. However, there was a strong resonant mode at 746Hz, as well as some pumping of the panels at the tuning frequency of the rear-facing passive radiator (fig.2; note that the driving voltage used to create this graph was 5dB lower than usual, due to the One's limited power handling). A resonance at 746Hz is sufficiently high in frequency that it should have no audible consequences.
The saddle just above 40Hz in the impedance-magnitude trace suggests that the tuning frequency of the passive radiator lies close to that frequency. The red trace in fig.3 shows the radiator's output measured in the nearfield; the output peaks a little lower in frequency than suggested by fig.1, at around 35Hz, and the corresponding notch in the drive-unit's output lies at 38Hz. This is a low tuning frequency for such a small speaker and correlates with BJR's finding the speaker's bass surprisingly extended. Still, the danger of such a design is that when the speaker is used in free space, its bass might sound a little disconnected from the lower midrange. It's fair to note that BJR didn't note hearing any such disconnect.
Higher in frequency, the One's response slopes up in the upper midrange, and the treble region is balanced 35dB too high in level compared with the midrange. Using the speakers close to boundaries, as on a desktop, will bring up the lower-midrange and bass regions, to give a more neutral balance. In free space, however, this sloped-up response will indeed give the impression of superb retrieval of recorded detail, as BJR heard. Fig.3 was taken with the grille removed; adding it raised the upper-midrange peak by about 1dB, but lowered the output between 5 and 12kHz by up to 1dB (not shown).
The Audience's lateral dispersion, normalized to its on-axis response (which therefore appears as a straight line) so that the differences are revealed, is shown in fig.4. What appears to be an off-axis flare at the cursor position (3.6kHz) in this graph is actually due to the on-axis suckout centered on that frequency filling in to the speaker's sides. The One's lateral dispersion is actually wide and even from the lower midrange through to the mid-treble. However, as expected from the drive-unit's 2.5" radiating diameter, its output becomes quite directional in the top two audio octaves, which will counteract the high-treble peakiness in the speaker's on-axis output. The picture in the vertical plane (fig.5) is very similar, which is to be expected given the design's symmetry.
In the time domain, the use of a single drive-unit without a crossover inherently endows the One with a time-coincident step response (fig.6). However, what should be a smooth decay of the right-triangleshaped response is broken by what appears to be a reflection about 300µs after the initial arrival of the pulse. The cumulative spectral-decay plot (fig.7) is clean in the midrange through to the mid-treble, but the twin peaks in the top octave are associated with ridges of delayed resonant energy. The upper peak is too high in frequency to be audible, but the lower-frequency peak, centered on 16.25kHz, might bother younger listeners. (Ignore the black ridge that visually breaks up this mode; this is due to the inevitable interference from the computer's video circuitry.)
The measured performance of Audience's ClairAudient The One suggests that it will offer the most neutral balance when used in close proximity to a boundary, as on a desktop. Bob Reina was also impressed by its sound when he listened to it in a conventional freefield situation, though its sloped-up response will make the choices of matching components rather critical.
Before writing these comments, I took a listen the Audience speakers on my desk in the office, playing files with iTunes and driving the speakers with a Yamaha desktop receiver. I was pleasantly surprised by what I heard. Yes, there was a touch of "cardboardy" coloration on vocals, a somewhat peaky top octave, and the low frequencies did occasionally sound a little "gruff," but the overall balance was musically pleasing. And as BJR noted, the One's upper-frequency clarity was impressive.John Atkinson