Measuring long impulse responses with pseudorandom sequences and sweep signals

In architectural acoustics, background noise, loudspeaker nonlinearities, and time variances are the most common disturbances that can compromise a measurement. The effects of such disturbances on measurement of long impulse responses with pseudorandom sequences (maximum-length sequences (MLS) and inverse repeated sequences (IRS)) and with linear and logarithmic sweep signals have been examined. The results reveal that the sweep method can provide a significant reduction of the effect of distortion compared with MLS/IRS techniques but, unlike what is claimed in the literature, sweep signals cannot reject all distortion artifacts from the causal part of the estimated impulse response. In all cases, an improvement of 3 dB in signal-to-noise ratio can be achieved if the measurement time is doubled either by averaging over two excitations or by doubling the length of the excitation signal. The time-frequency selectivity performed by the sweep technique is found to yield higher immunity to impulsive noise and time variances.