A Comparison Of Late Lateral Energy (GLL) and Lateral Energy Fraction (LF) Measurements Using a Spherical Microphone Array and Conventional Methods

Abstract

Late Lateral Energy Level (GLL) and Lateral Energy Fraction (LF) are two room acoustics measures that have both been shown to correlate with certain aspects of the spatial impression of a listening space. In order to obtain these quantities, the lateral energy must be measured, which is typically carried out using microphones with a figure-of-eight (figure-8) polar pickup pattern. However, most commercially available figure-8 microphones are intended for use in audio recording applications, and are not laboratory-grade or designed for room acoustics impulse response (IR) measurements. Such microphones may suffer from non-ideal frequency response and/or directivity patterns. This study compares measurements that were taken in a 2500 seat auditorium using an omni-directional and studio-grade figure-8 microphone pair versus the omni-directional and dipole components extracted from a 32 element spherical microphone array. The results show that the two measurement methods agree in the 500 Hz and higher octave bands, but differ at low frequencies due to differences in the directivity patterns. The difference of the LF average from 125 Hz to 1 kHz for the two methods was between 0.59 and 1.81 just noticeable differences (JNDs) at the six receiver locations. The difference of the GLL average from 125 Hz to 1 kHz for the two methods was between 0.02 and 0.48 JNDs (applying the JND for strength of 1 dB). It was also found that repeatability error was present at one of the six receiver locations for the LF measurements, but was very small for the GLL measurements.