Parametric Audio Coding With Exponentially Damped Sinusoids

Abstract

Abstract-Sinusoidal modeling is one of the most popular techniques for low bitrate audio coding. Usually, the sinusoidal parameters (amplitude, pulsation and phase of each sinusoidal component) are kept constant within a time segment. An alternative model, the so-called Exponentially-Damped Sinusoidal (EDS) model, includes an additional damping parameter for each sinusoidal component to better represent the signal characteristics. It was however never shown that the EDS model could be efficient for perceptual audio coding. To that aim, we propose in this paper an efficient analysis/synthesis framework with dynamic timesegmentation on transients and psychoacoustic modeling, and an asymptotically optimal entropy-constrained quantization method for the four sinusoid parameters (e.g including damping). We then apply this coding technique to real audio excerpts for a given entropy target corresponding to a low bitrate (20 kbits/s), and compare this method with a classical sinusoidal coding scheme using a constant-amplitude sinusoidal model and the perceptually weighted Matching Pursuit algorithm. Subjective listening tests show that the EDS model is more efficient on audio samples with fast transient content, and similar to the classical model for more stationary audio samples.

Modal decomposition of musical instrument sounds via optimization-based non-linear filtering

“…(34) corresponds to the extension of linear filtering with the perfect reconstruction constraint given by Eq. (13), and a solution to Eq. (13) is not modified by thex-update.…”

“…As modes contain significant information on the corresponding sound, parametric mod-eling of each mode is often considered in the context of sound synthesis. Many models have been proposed in this respect, including exponentially damped sinusoidal (EDS) model [10][11][12][13][14] and damped and delayed sinusoidal model [15][16][17][18] as an extension of EDS model. To represent complex decaying processes such as double decay and beats, recently, adaptive harmonic model (AHM) has been applied to the modeling of musical instrument sounds [19][20][21][22].…”

Effects of envelope and dictionary structure on the performance of matching pursuit

“…[14]. The example applications of non‐Gaussian function to matching pursuit include damped sinusoidal functions [15] used for parametric audio coding [16] and compression of digital fault records [17], as well as the Ricker envelope used for decomposition of seismic signals [18].…”

Robust and Efficient Harmonics Denoising in Large Dataset Based on Random SVD and Soft Thresholding

“…This is a common signal model in a wide range of engineering or research areas, e.g. audio processing [4], [5], Nuclear Magnetic Resonance (NMR) [6], biomedical applications [7], and vibration analysis [8]- [10]. These harmonic signals have an essential property, that the Hankel matrix of the noise-free signal has a rank equal to the number of the harmonics [11]- [13].…”