B&W John Bowers Silver Signature loudspeaker Measurements part 3

Carrying out a spatially averaged 1/3-octave spectrum analysis for left and right speakers individually in my room and averaging the results gives the curve shown in fig.10. The unevenness between 50Hz and 300Hz is due to room effects that have not been eliminated by the spatial averaging. Superbly even in balance with, as conjectured earlier, just a slight depression in the presence region, this curve ties in neatly with my listening impressions: the Silver Signature is fundamentally a neutral, uncolored loudspeaker, with good small-signal low-frequency extension to the 32Hz band.

Fig.10 B&W Silver Signature, spatially averaged 1/3-octave response in JA's listening room.

In the time domain, the Silver Signature's performance is more mundane. Its impulse response (fig.11) is dominated by the tweeter's ultrasonic ringing. The step response (fig.12) confirms that the tweeter is connected in inverse polarity to the woofer and indicates that it leads it slightly in time. (Though the HF unit is stepped back on top of the enclosure, this is not enough to bring its output into time step with that of the woofer.) The importance of time coherence in a loudspeaker is not settled: good-sounding designs can be time-coherent or not. The B&W falls into the latter camp (footnote 2).

Fig.11 B&W Silver Signature, impulse response on tweeter axis at 45" (5ms time window, 30kHz bandwidth).

Fig.12 B&W Silver Signature, step response on tweeter axis at 45" (5ms time window, 30kHz bandwidth).

The waterfall plot calculated from the impulse response is shown in fig.13. A superbly clean decay overall is undoubtedly associated with the speaker's refined sound. Note, however, the ridge parallel to the time axis associated with the on-axis peak at 3kHz. Here's that pesky woofer-cone mode again.

Fig.13 B&W Silver Signature, cumulative spectral-decay plot at 45" (0.15ms risetime).

Footnote 2: Martin Colloms once told me that the best candidates for DSP optimization are the loudspeakers that need it least. The correction can therefore be obtained with the least compromise of dynamic range and/or resolution. A long-term project of mine will be to experiment with digital correction filters for the Silver Signature, to see if correcting for the speaker's very slight departure from on-axis flatness and rendering it time-coherent will take it to the ultimate level of performance.