On the variability of room acoustical parameters: Reproducibility and statistical validity

Abstract

For many years acoustic quality has been evaluated using a set of objective parameters. Although there is now a growing concensus about the kind and number of parameters to be used, problems still occur when trying to characterise the acoustics of a hall using these parameters. The first part of this work focusses on the specific problems which occur while measuring objective parameters in halls. The influence of each element of the measuring system is considered; it is shown for instance that many parameters vary strongly with the sound source. The spatial variability of parameters is then discussed. The effects of small spatial variations on each parameter are evaluated and compared to reproducibility and within hall variations. It is found that some parameters, including the well known reverberation time, are unable to describe spatial variations within a hall. Lastly, the critical problem of hall to hall comparisons is considered. It is shown that in some cases a comparison between two different halls is impossible because the parameter values overlap strongly. Statistical comparison between halls usually assumes that parameter values within halls reject this assumption for many parameters. Stability of the parameters distribution is then evaluated.