This paper describes a method for obtaining perceptually motivated and perfectly invertible time-frequency representation of sound signal. Based on frame theory the recent non-stationary Gabor transform, linear with resolution evolving across frequency is formulated implemented as non-uniform filterbank. To match human auditory resolution, transform uses Gaussian windows equidistantly spaced psychoacoustic "ERB" scale. Additionally, features adaptable redundancy. Simulations showed that...

Many audio applications rely on filter banks (FBs) to analyze, process, and re-synthesize sounds. For these applications, an important property of the analysis–synthesis system is reconstruction error; it has be minimized avoid audible artifacts. Other advantageous properties include stability low redundancy. To exploit some aspects auditory perception in signal chain, FBs that approximate frequency analysis performed periphery, gammatone FB being a popular example. However, current only...

This paper describes a method to obtain perceptually relevant sparse representation of sound signal. Based on matching pursuit (MP) and recent psychoacoustic data time-frequency masking measured with Gabor atoms, perceptual (PMP) algorithm is proposed. To good match between the model signal representation, dictionary atoms variable sizes chosen for MP. In proposed method, first decomposed using MP applied resulting set atoms. allows isolating masked components from residual. Experimental...

The additivity of nonsimultaneous masking was studied using Gaussian-shaped tone pulses (referred to as Gaussians) masker and target stimuli. Combinations up four temporally separated Gaussian maskers with an equivalent rectangular bandwidth 600 Hz duration 1.7 ms were tested. Each level-adjusted produce approximately 8 dB masking. Excess (exceeding linear additivity) generally stronger than reported in the literature for longer comparable levels. A model incorporating a compressive...

Peripheral compression is believed to play a major role in the masker phase effect (MPE). While almost instantaneous, activation of efferent system reduces temporally evolving manner. To study efferent-controlled MPE, experiment 1, simultaneous masking 30-ms 4-kHz tone by 40-ms Schroeder-phase harmonic complexes was measured with on- and off-frequency precursors as function curvature for two levels (60 90 dB sound pressure level). The MPE quantified threshold range [min/max difference (MMD)]...

Many audio applications perform perception-based time-frequency (TF) analysis by decomposing sounds into a set of functions with good TF localization (i.e. small essential support in the domain) using transforms and applying psychoacoustic models auditory masking to transform coefficients. To accurately predict interactions between coefficients, properties model should match those transform. This involves having data for stimuli localization. However, little is known about mathematically...

Laback et al. [(2011). J. Acoust. Soc. Am. 129, 888−897] investigated the additivity of nonsimultaneous masking using short Gaussian-shaped tones as maskers and target. The present study involved Gaussian stimuli to measure simultaneous for combinations up four spectrally separated maskers. According most basilar membrane measurements, should be processed linearly at characteristic frequency (CF) Assuming also compression target, all masker produce excess (exceeding linear additivity)....

Many audio applications rely on filter banks (FBs) to analyze, process, and re-synthesize sounds. To approximate the auditory frequency resolution in signal chain, some perceptually motivated FBs, gammatone FB being a popular example. However, most FBs only allow partial reconstruction at high redundancies and/or do not have good resistance sub-channel processing. This paper introduces an oversampled enabling perfect reconstruction, efficient design, adaptable redundancy. The filters are...

Both temporal and spectral masking have been studied extensively in the literature. Mostly, they regarded as separate phenomena. Very little is known about interaction between these two effects, i.e. time-frequency domain. Data on spread of evoked by a single Gaussian-shaped tone pulse are presented an accompanying study at same conference (Necciari et al.). The current gathers data additivity up to four, approximately equally effective Gaussian maskers (ERB=600 Hz), separated either along...

Harmonic complexes are often used as maskers for measuring the cochlear phase response. The curvature can affect masking in order of up to 20 dB, an effect known masker-phase effect. There is evidence that signals yielding peaky internal masker representations after passing filter produce minimum masking, with fast-acting compression main contributor Thus, hearing-impaired listeners showing reduced or absent compression, estimation response using method may be difficult. Here, alternative...

This study investigates auditory masking with Gaussian-windowed tones as target and masker stimuli. On the purpose of developing a time-frequency model, such stimuli minimize uncertainty. Also, proposed by van Schijndel et al. (1999), they activate single spectro-temporal observation window system. The presented here measured an ERB 600 Hz effective duration 9.6 ms. was centered at 4 kHz. Its level 60 dB SL. Four experiments were conducted. (1) Absolute thresholds for 300-ms-sinusoidal...

Stimulation of the olivocochlear efferent system can reduce gain basilar membrane (BM) response to a sound. This reduction may be accompanied by in BM compression, although behavioral data are inconclusive. In this study, effect stimulation on compression is assessed using fixed-duration masking curves (FDMCs) with precursor. pilot experiment, presenting broadband noise precursor before masker resulted large within-listener variability for shortest target durations, three-interval...