This thesis explores the means of objectively measuring diffuse sound fields contained within music performance auditoria. Although the diffuse field is considered to be an important component of the reverberant sound field there is currently no widely accepted method for its measurement.
A review of methods shows that attempts to characterize the field may be divided into those methods that seek to directly measure the state of the field and those that indirectly indicate the existence of the state.
The primary focus of this thesis is the application of Ambisonic techniques to capture the spatial aspects of the sound field.
Initial work explores the rotation of a directional microphone in three measurement spaces. The results and modeling in idealised simulated sound fields indicate that the method may have some efficacy.
The method is extended through the application of signal processing to the output of an Ambisonic microphone array. The method is tested firstly in a reverberation room that is modified progressively to produce a series of room states with incrementally increasing reverberation time.
The extents of the measurement system were tested by measuring the degrees of diffusivity reached in a reverberation room. Diffusing panels were progressively added in the expectation the increases in diffusivity would be detected. The measurement was carried out in conjunction with standard absorption coefficient measurements outlined in Appendix A of ISO 354. Comparison was made between the measured field and the standard method for achieving a diffuse field in a reverberation room test facility.
The final stage attempts to find correlation between physical measures of diffuse fields and listener’s subjective assessment of those fields. To that end a paired comparison test was conducted where listeners were presented music samples rendered through simulated halls where the scattering coefficients and consequently the sound field diffusivity was varied. Subjects were asked to choose which pair they preferred.