PAN-AR: A Multimodal Dataset of Higher-Order Ambisonics Room Impulse Responses, Ambient Noise and Spherical Picture

This work presents PAN-AR, a dataset of spatial room impulse responses (SRIRs) encoded in the higher-order Ambisonics (HOA) format. The dataset encompasses measurements obtained in four distinct environments: a printer room, a meeting room, a classroom, and an underground parking area. The SRIRs have been recorded for different configurations of the emitter and receiver positions, including up to six emitter positions and two receiver positions. The dataset also includes samples of the ambient noise and spherical pictures captured at the receiver positions, along with other metadata. These characteristics render PAN-AR well-suited for extended reality applications, as it enables the dynamic simulation of virtual sound sources in the space surrounding the listener. Moreover, the dataset favors the creation of ecological virtual auditory environments, as it encompasses scenarios that are commonly encountered in the everyday lives of workers and students. In addition, the ambient noise recordings and the spherical pictures provide a more immersive and realistic experience. Finally, an analysis of the acoustical properties of the measured SRIRs is provided in the paper.