Sidebar 3: Measurements
I used DRA Labs' MLSSA system and a calibrated DPA 4006 microphone to measure the PSB Imagine T3's frequency response in the farfield, and an Earthworks QTC-40, which has a ¼" capsule and thus doesn't present a significant obstacle to the sound, for the nearfield frequency responses. I also used the Earthworks mike for the in-room response.
My estimate of the Imagine T3's voltage sensitivity was 89.2dB/2.83V/m, confirming the specified figure. The PSB's electrical impedance magnitude and phase are shown in fig.1. Other than in the regions between 28 and 40Hz and between 75 and 200Hz, the magnitude (solid trace) remains above 4 ohms at all frequencies. The minimum value is 3.5 ohms at 85Hz, and the phase angle (dotted trace) is generally low. Any amplifier that is comfortable driving 4 ohms will have no difficulty with the Imagine T3.
The complicated-looking fig.2 shows the nearfield responses of the T3's midrange unit (black trace), upper woofer and port (blue), middle woofer and port (green), and bottom woofer and port (red). The three woofers do all behave somewhat differently, the bottommost having a slightly higher Quality factor (Q), and both it and the port that loads it have a sharp rolloff above 80Hz. To my surprise, it was the middle woofer that extended high enough in frequency to cross over to the midrange unit, while the top woofer rolled off above 500Hz. This is the opposite of what is written in the T3's press materials, but the two samples were the same in this respect; perhaps this is to arrange for the acoustic center of the PSB's lower-frequency output to coincide with the physical center of the three drivers. However, what concerned me much more in this graph is the high-level resonant peak at 650Hz that can be seen in the outputs of the upper two ports. The ports face away from the listener, and my listening suggested that this behavior looks worse than it sounds. But I know from his earlier designs that Paul Barton has always worked hard to eliminate port misbehavior in the midrange.
Fig.1 PSB Imagine T3, electrical impedance (solid) and phase (dashed) (5 ohms/vertical div.).
The impedance traces are free from the discontinuities that would hint at the presence of resonance. My usual accelerometer was suffering from a broken preamplifier when I did the testing, but listening to the enclosure with a stethoscope, I could hear some narrowband liveliness between 550 and 750Hz, particularly on the sidewalls level with the middle woofer.
Fig.2 PSB Imagine T3, nearfield responses of: midrange unit (black), top woofer and port (blue), middle woofer and port (green), bottom woofer and port (red), all plotted in the ratios of their radiating diameters.
Fig.3 shows the Imagine T3's acoustic crossover, with the summed outputs of the three ports (red trace) and of the three woofers (blue). The handover from the woofers to the midrange unit (green trace) occurs closer to 800Hz than the specified 450Hz, but the high- and low-pass rolloffs are steep and symmetrical. The apparent rise in the upper bass is entirely an artifact of the nearfield measurement technique, and the output of the woofers is flat and even below the crossover point. Similarly, the outputs of the midrange unit and tweeter are generally even, though with slight excesses of energy in the presence region and the top audio octave. The same trends can be seen in the overall farfield response, averaged across a 30° horizontal window centered on the tweeter axis (fig.4).
Fig.3 PSB Imagine T3, acoustic crossover on tweeter axis, corrected for microphone response, with farfield responses of midrange-and-tweeter and woofers, with summed nearfield responses of ports and woofers.
Fig.4 PSB Imagine T3, anechoic response on tweeter axis at 50", averaged across 30° horizontal window and corrected for microphone response, with complex sum of nearfield responses plotted below 300Hz.
The Imagine T3's plot of lateral dispersion, normalized to the tweeter-axis response (fig.5), offers evenly spaced contour lines with the usual narrowing of the radiation pattern in the top two octaves of the tweeter's passband. The vertical dispersion, again referenced to the response on the tweeter axis (fig.6), which is 37" from the floor, suggests that the T3's balance doesn't appreciably change over quite a wide window. This does not correlate with what I heard, however. A suckout just below 900Hz appears 10° below the tweeter axis, which again suggests that the crossover between the woofers and the midrange unit occurs higher than the specified 450Hz.
Fig.5 PSB Imagine T3, lateral response family at 50", normalized to response on tweeter axis, from back to front: differences in response 90–5° off axis, reference response, differences in response 5–90° off axis.
Fig.6 PSB Imagine T3, vertical response family at 50", normalized to response on tweeter axis, from back to front: differences in response 15–5° above axis, reference response, differences in response 5–10° below axis.
The Imagine T3's spatially averaged response at the listening position in my room (fig.7, red trace) is extremely even from the midbass through to the mid-treble, above which it shelves down due to the increasing absorptivity of the room's furnishings. The bottom port of both speakers was blocked for this measurement, but there is still a residual peak evident between 25 and 35Hz, which is due to the lowest-frequency diagonal room mode. This peak can also be seen in the spatially averaged response of the Vandersteen Model Seven Mk.II, which I reviewed in the May issue. With its powered woofers, the Vandersteen extends lower in frequency, while having less upper-bass energy but more midrange energy in-room than the PSB. The two speakers match closely in the treble, however.
Fig.7 PSB Imagine T3, spatially averaged, 1/6-octave response in JA's listening room (red) and of Vandersteen Model 7 Mk.II (blue).
In the time domain, the Imagine T3's step response on the tweeter axis (fig.8) shows that all five drive-units are connected in positive acoustic polarity, and that the crossover filters have been optimally designed. The T3's cumulative spectral-decay plot (fig.9) demonstrates a clean decay in the region covered by the tweeter.
Fig.8 PSB Imagine T3, step response on tweeter axis at 50" (5ms time window, 30kHz bandwidth).
Fig.9 PSB Imagine T3, cumulative spectral-decay plot on tweeter axis at 50" (0.15ms risetime).
As always with a Paul Barton design, PSB's Imagine T3 demonstrates excellent speaker engineering, though I was bothered by the midrange port resonance and the disparity between the measured and the specified crossover frequencies.— John Atkinson
