472 Recording Studio Design 1 Amplitude 0 ­1 0 5 10 15 20 Time (ms) Figure 19.7 A step function woofer. This time non-synchronicity is due to the physical alignment of the drivers and the group delay due to the crossover frequency and slope. The manufacturers considered the subjective effect to be too small to warrant complicating the design. The Westlake shows in Figure 19.6(b) a more compact response. The tweeter can still be seen to speak in advance of the woofers, though only by about 500 microseconds. The response can be considered good. The time separation of the drivers at the crossover point also manifests itself as a kink in the acoustic source plot, Figure 19.4(b), at around 800 Hz. The Tannoy shows a better step function response, this time with only about a 200-microsecond delay, which is probably inaudible. However, for comparison, Figure 19.8 shows the exemplary step function leading edge of two Quad Electrostatic loudspeakers, which shows what can be achieved. 19.7 Power cepstra The power cepstrum responses highlight problems due principally to surface irregularities and diffraction. The power cepstrum plot of the JBL, in Figure 19.9, shows almost no evidence of reflexions, which suggests that the stereo imaging should be good due to the lack of physically-separated secondary diffraction sources. The cepstrum plot of the Westlake shows evidence of echoes at 0.2 and 0.5 ms, and these may be responsible for the slightly irregular frequency response in the mid-range, due to interference between the direct and secondary sources (see Chapter 4). The Tannoy shows evidence of reflexions at 140 and 280 microseconds (0.14 and 0.28 ms) which may be due to the discontinuities in the horn flare where the metal phasing plug meets the coil gap of the low frequency cone. Again, these could be the source of some of the on-axis mid- range frequency response irregularities.