Modeling the Impulse Response of Higher-Order Microphone Arrays using Differentiable Feedback Delay Networks

The following audio samples are provided as supplementary material for the article Modeling the Impulse Response of Higher-Order Microphone Arrays using Differentiable Feedback Delay Networks, authored by R. Giampiccolo, A. I. Mezza, M. Pezzoli, S. Koyama, A. Bernardini, and F. Antonacci, and accepted for publication at the 28th International Conference on Digital Audio Effects (DAFx25), Ancona, Italy, 2-5 September, 2025.

The stimuli included in the perceptual test are binaural auralizations obtained starting from the Impulse Responses (IRs) of Differentiable Feedback Delay Networks (FDNs) trained with the following loss functions:

• Frequency-Independent (FI) Loss;
• Frequency-Dependent (FD) Loss;
• Spatial (SP) Loss.

Audio Samples

	Female Speech	Male Speech 1	Male Speech 2	Flute	Acoustic Guitar
Reference
FI Loss
FD Loss
SP Loss