Download Differentiable Time–frequency Scattering on GPU Joint time–frequency scattering (JTFS) is a convolutional operator in the time–frequency domain which extracts spectrotemporal modulations at various rates and scales. It offers an idealized model of spectrotemporal receptive fields (STRF) in the primary auditory cortex, and thus may serve as a biological plausible surrogate for human perceptual judgments at the scale of isolated audio events. Yet, prior implementations of JTFS and STRF have remained outside of the standard toolkit of perceptual similarity measures and evaluation methods for audio generation. We trace this issue down to three limitations: differentiability, speed, and flexibility. In this paper, we present an implementation of time–frequency scattering in Python. Unlike prior implementations, ours accommodates NumPy, PyTorch, and TensorFlow as backends and is thus portable on both CPU and GPU. We demonstrate the usefulness of JTFS via three applications: unsupervised manifold learning of spectrotemporal modulations, supervised classification of musical instruments, and texture resynthesis of bioacoustic sounds.
Download On the control of the phase of resonant filters with applications to percussive sound modeling Source-filter models are widely used in numerous audio processing fields, from speech processing to percussive/contact sound synthesis. The design of filters for these models—be it from scratch or from spectral analysis—usually involves tuning frequency and damping parameters and/or providing an all-pole model of the resonant part of the filter. In this context, and for the modelling of percussive (non-sustained) sounds, a source signal can be estimated from a filtered sound through a time-domain deconvolution process. The result can be plagued with artifacts when resonances exhibit very low bandwidth and lie very close in frequency. We propose in this paper a method that noticeably reduces the artifacts of the deconvolution process through an inter-resonance phase synchronization. Results show that the proposed method is able to design filters inducing fewer artifacts at the expense of a higher dynamic range.