Loudspeaker–Room Response Equalization Using a Smartphone Microphone

Abstract

The thesis investigates the topic of loudspeaker-room response equalization (LRRE) and tries to find a suitable method for integration into a pair of compact studio monitors. LRRE is a method for improving a loudspeaker system’s performance in a room by measuring an impulse response in the listening position, based on which compensation filters are computed and applied to the loudspeaker system. The aim is also to assess whether such impulse response measurements can be performed using today’s smartphones with their built-in microphone. The whole procedure, from impulse response measurement to the computation of the compensation filter, is described, and techniques that correct only the magnitude response (minimum-phase compensation) and those correcting both the magnitude and phase responses (mixed-phase compensation) are compared. Three particular equalization functions are implemented in Python: a minimum-phase FIR filter, a mixed-phase FIR filter obtained by the x-filtered normalized least mean square algorithm, and a minimum-phase IIR filter comprising a cascade of biquad peak filters. It is found that the latter performs the best and is the most suitable option for integration into compact studio monitors. A theoretical performance of a smartphone used as the measurement device is dis cussed by presenting the expected behaviour of the present MEMS microphone, and the application of digital signal processing by the manufacturer. The actual behaviour is then evaluated by measuring the frequency responses and sensitivi ties of 30 smartphones in an anechoic chamber and assessing their SNR. 84% of the tested models emerged as suitable for measuring the loudspeaker-room impulse response