Rayban | Meta are the second generation of smart glasses developed by Meta and Luxottica. They are one of the first mass-market all-in-one consumer devices allowing users to record and reproduce sounds binaurally. Traditionally, binaural recording systems use two microphones, one in each hearing canal, belonging either to a person or to a dummy head. In both cases, the incoming sound reflects on the body, shoulders, and ear pinnae, thus physically encoding several binaural cues. Rayban | Meta, instead, rely on a 5-microphone array, none of which enter the ear canal, and therefore devoid of the information encoded by the pinnae. The binaural signal is obtained through a beamforming algorithm, about which nothing has been published in the literature. For this reason, we evaluated the quality of the binaural signals through impulse response measurements. Wearing a pair of Rayban | Meta and a set of DPA4560 binaural microphones, we used the exponential sine sweep method, sampling every 10°. Using the Aurora plugins, we obtained values for IACC (Inter Aural Cross Correlation), ITD (Interaural Time Difference) and ILD (Interaural Level Difference). As frequency response tests, especially regarding sound reproduction, are widely available, we focused on the binaural parameters only.
