Wilson Audio Specialties WATT/Puppy System 8 loudspeaker Measurements

Sidebar 3: Measurements

I couldn't lift the 170-lb WATT/Puppy 8 onto a tall stand for the acoustic measurements, so the farfield response measurements have less frequency resolution in the midrange than is usual. I don't believe this negatively impacted the speaker's measured performance to any significant extent. In its various generations, the WATT/Puppy has always had a voltage sensitivity that is much higher than average, and the System 8 continues that tradition, my estimate coming in at 92dB(B)/2.83V/m. However, and also like its predecessors, the WATT/Puppy 8's impedance characteristic demands the use of an amplifier that can deliver generous amounts of current. There is a drop to 2.2 ohms at 77Hz (fig.1), though the phase angle in the bass is not quite as extreme as it was in the System 7. Other than that, the impedance curve is very similar to that of the earlier version.

Fig.1 Wilson WATT/Puppy 8, electrical impedance (solid) and phase (dashed). (2 ohms/vertical div.)

Other than the small blip at 20kHz, which indicates the presence of the inverted-dome tweeter's diaphragm resonance, the impedance traces are free from the small wrinkles and discontinuities that would imply the presence of cabinet-panel resonances. Even so, I found a couple of low-level modes present on the WATT's sidewalls, at 332Hz and 406Hz (fig.2). These are higher in frequency than the modes I found with the System 7 WATT, which will lower the audibility of any effect they might have. The Puppy cabinet was extraordinarily inert.

Fig.2 Wilson WATT/Puppy 8, cumulative spectral-decay plot calculated from the output of an accelerometer fastened to the WATT cabinet's side panel (MLS driving voltage to speaker, 7.55V; measurement bandwidth, 2kHz).

With the speaker driven directly, instead of via the woofer's Puppy Tail, the WATT 8's impedance trace (fig.3) indicates that the small port on its rear panel is tuned to a low 41Hz.

Fig.3 Wilson WATT, driven direct, electrical impedance (solid) and phase (dashed). (2 ohms/vertical div.)

The nearfield responses of the WATT midrange/woofer when driven direct (fig.4, blue trace) has the expected minimum-motion notch at the same frequency, while the port (red trace) broadly peaks between 30 and 100Hz. While there is a small peak in the latter's output at 400Hz, the sum of these nearfield responses, taking into account acoustic phase and the different distances of the two radiators from a nominal farfield microphone position, is basically flat (fig.4, black trace), which suggests a somewhat overdamped alignment.

Fig.4 Wilson WATT 8, driven directly, nearfield responses of the port (red), midrange/woofer (blue) and their complex sum (black), scaled in the ratio of the radiating diameters.

When the WATT 8 is driven from the Puppy via the Puppy Tail, however, the picture changes somewhat (fig.5). It looks as if a high-pass filter is used, which drastically reduces the output of the WATT's port (fig.5, red trace) and modifies the shape of both the WATT's midrange/woofer (blue trace) and the complex sum of the two outputs (black trace). With respect to the small peak at 70Hz and the lack of energy between 100Hz and 300Hz (both of which I believe are due to the interaction between the filter's series impedance and that of the drive-unit), the WATT 8 behaves very similarly to the WATT 7.

Fig.5 Wilson WATT 8, driven from the Puppy via the Puppy Tail, nearfield responses of the port (red), midrange/woofer (blue), and their complex sum (black), scaled in the ratio of the radiating diameters.

The black trace above 300Hz in fig.6 shows the WATT 8's farfield response, averaged across a 30° horizontal window centered on the tweeter axis. Below 300Hz, it repeats the black trace in fig.5. The red trace in this graph shows the output of the Puppy's port, the blue trace that of the Puppy's woofers. These have a sharply defined minimum-motion notch in their response at 30Hz, though the port's output actually peaks a little lower in frequency, with a small shelf apparent in its output between 40 and 70Hz. In this respect, it appears to behave identically to the port in the System 7 Puppy. The complex sum of the Puppy's nearfield response is shown as the green trace in fig.6: it peaks strongly between 120Hz and 40Hz, which coincides to some extent with the peak in the WATT's output.

Fig.6 Wilson WATT/Puppy 8, anechoic response without grille on tweeter axis at 50", averaged across 30° horizontal window and corrected for microphone response, with the complex sum of the WATT's nearfield port and woofer responses (black), the Puppy's port (red), woofers (blue), and their complex sum (green), plotted below 300Hz, 230Hz, 1kHz, and 300Hz, respectively, and scaled in the ratio of the square roots of the radiating areas.

Higher in frequency, the WATT/Puppy 8's response is broadly similar to that of the System 7, though with slightly more mid-treble energy and a little less high-treble output evident. But, as in the earlier design, the Focal tweeter's dome-resonance peak lies at 20kHz, which is a little close to the audioband for comfort.

The Wilson speaker's lateral-dispersion plot (fig.7) indicates a broadly even radiation pattern, with an off-axis notch developing around 4kHz, where the on-axis response has a small peak. In the vertical plane (fig.8), the treble region doesn't change much for listening axes on or just below the tweeter. However, a large suckout develops at 3.3kHz above the tweeter axis, which I assume is the frequency at which the WATT's tweeter crosses over to its woofer. As with earlier versions of the WATT/Puppy, it is important to use the correct spikes for the chosen speaker distance to tilt the System 8 toward the listener's ears.

Fig.7 Wilson WATT/Puppy 8, 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.8 Wilson WATT/Puppy 8, 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.

I measured the W/P8's spatially averaged response at the listening position in WP's room (Wes' preferred position rather than Peter McGrath's) using an Earthworks omni mike, a Metric Halo ULN-2 mike preamp-A/D converter, and SMUG Software's Fuzzmeasure 2.0 running on my Apple PowerBook. I took 10 response measurements for each speaker individually in a vertical grid centered on the position of Wes's head in his listening chair, and averaged them to reduce the effect of position-dependent room effects. The result is shown in fig.8. The in-room responses for the Dynaudio Confidence C4 and Canton Vento Reference One DC, taken under identical conditions in Wes's room, can be found at here. If you look at those curves, you can see that the big peak between 60 and 100Hz and the narrow suckout between 30 and 50Hz are symptomatic of the acoustics of Wes's room, and have not been much reduced by the spatial averaging. The lows extend at full level almost to 20Hz in this graph. But note the lack of energy in this curve between 100 and 200Hz, which I believe is due to the Puppy's upper-bass output not extending sufficiently high in frequency to compensate for the WATT's lack of energy in the same region.

If you compare fig.8 with fig.5, you can see that there is an unfortunate coincidence between the behavior of Wes's room in the midbass and the tuning frequency of both the WATT and the Puppy. Wes said that he found the speaker's bass balance to be both "warmer and ever-so-slightly smeared" in comparison with the Dynaudio speaker, and I noticed the same thing when I auditioned the W/P8s in his room. But I also felt the stereo imaging to be stable and well-defined and the sound superbly detailed, though the small presence-region peak visible in fig.8 might have had something to do with the latter aspect. The gentle downward slope of the speaker's in-room treble balance in fig.8 is due to the room furnishings' increasing absorption in this frequency region; the WATT/Puppy 8's treble sounded both natural and very clean, I thought.

Fig.9 Wilson WATT/Puppy 8, 1/6-octave, spatially averaged response in WP's listening room.

In the time domain, the W/P8's step response on the tweeter axis (fig.10) is basically identical to that of the W/P7, with the tweeter and Puppy woofers connected in positive acoustic polarity, the WATT midrange/woofer in inverted acoustic polarity. Each drive-unit's step smoothly hands over to that of the next lower in frequency, which correlates with the good frequency-domain integration seen in fig.5. The WATT/Puppy 8's cumulative spectral-decay plot, taken on the tweeter axis (fig.11), is broadly similar to that of the W/P7, but cleaner both at the top of the WATT midrange/woofer's passband and in the tweeter's top octave. Despite my caution about the tweeter's dome resonance being close to the audioband, very little delayed energy seems to be associated with it, which will minimize any audible problems. The latest version of this Focal tweeter is definitely better than the one used in the W/P7.

Fig.10 Wilson WATT/Puppy 8, step response on tweeter axis at 50" (5ms time window, 30kHz bandwidth).

Fig.11 Wilson WATT/Puppy 8, cumulative spectral-decay plot at 50" (0.15ms risetime).

Measuring the performance of a speaker as complex as the WATT/Puppy 8, in which the outputs of spaced drive-units overlap in some frequency bands, is an exercise in difficulty. But I believe I have correctly characterized how it behaves. It is definitely a difficult load for an amplifier to drive, which leaves me puzzled as to why Wes had no problems driving it with the tubed Cayin amplifier. The Wilson did sound superb driven by Ayre MX-R monoblocks on both of the occasions I listened seriously to it in Wes's room, but its idiosyncratic low-frequency behavior compared with more conventional designs will make matching it to the listening room tricky—a task best left to the Wilson-trained dealer who will be installing the WATT/Puppy 8s in the customer's room.—John Atkinson

Wilson Audio Specialties
2233 Mountain Vista Lane
Provo, UT 84606
(801) 377-2233