The MLS technique is widely used for measurement in architectural acoustics and in audio. The MLS technique requires that the system under test is time-invariant, and it has been known for years that small changes are a source of error in such measurements, in particular if averaging is used to increase the signal-to-noise ratio. Time variance can, for example, be caused by a gradually changing temperature or motion of the air in a room. The effect of time-variance on MLS measurements has been studied through computer simulations and experiments. A new indicator analogous to the coherence function in measurements with dual channel FFT analysers has been suggested and examined experimentally in a reverberation room.
|Period||15/04/1998 → 15/04/1999|