The role of short-time intensity and envelope power for speech intelligibility and psychoacoustic masking

Abstract

The generalized power spectrum model [GPSM; Biberger and Ewert (2016). J. Acoust. Soc. Am. 140, 1023-1038], combining the "classical" concept of the power-spectrum model (PSM) and the envelope power spectrum-model (EPSM), was demonstrated to account for several psychoacoustic and speech intelligibility (SI) experiments. The PSM path of the model uses long-time power signal-to-noise ratios (SNRs), while the EPSM path uses short-time envelope power SNRs. A systematic comparison of existing SI models for several spectro-temporal manipulations of speech maskers and gender combinations of target and masker speakers [Schubotz et al. (2016). J. Acoust. Soc. Am. 140, 524-540] showed the importance of short-time power features. Conversely, Jørgensen et al. [(2013). J. Acoust. Soc. Am. 134, 436-446] demonstrated a higher predictive power of short-time envelope power SNRs than power SNRs using reverberation and spectral subtraction. Here the GPSM was extended to utilize short-time power SNRs and was shown to account for all psychoacoustic and SI data of the three mentioned studies. The best processing strategy was to exclusively use either power or envelope-power SNRs, depending on the experimental task. By analyzing both domains, the suggested model might provide a useful tool for clarifying the contribution of amplitude modulation masking and energetic masking.

Development of temporal auditory processing in childhood: Changes in efficiency rather than temporal-modulation selectivity

“…These studies on AM perception have provided support for the development of computational models of AM processing based on the concept of a modulation filterbank (Dau et al, 1997a). These models are used to predict the perception of AM cues and have recently accounted for a wide range of detection, discrimination and identification data in adults (e.g., Biberger and Ewert, 2017).…”

Instrumental Quality Predictions and Analysis of Auditory Cues for Algorithms in Modern Headphone Technology

“…The GPSM q ( Biberger et al., 2018 ) represents an audio quality extension of the GPSM, which has been demonstrated to predict the results of many psychoacoustic and speech intelligibility experiments ( Biberger & Ewert, 2016 , 2017 ). GPSM q applies a linear, fourth-order gammatone filterbank with bandwidth equal to the equivalent rectangular bandwidth of the auditory filter (ERB N ; Glasberg & Moore, 1990 ; Moore & Glasberg, 1983 ) that simulates the behavior of the basilar membrane, followed by calculating the low-pass filtered Hilbert envelope (cutoff frequency of 150 Hz) to account for decreased modulation sensitivity at high modulation frequencies.…”

“…Biberger and Ewert combined the Power Spectrum Model (PSM; Fletcher, 1940 ; Patterson & Moore, 1986 ) and Envelope Power Spectrum Model (EPSM; Ewert & Dau, 2000 ) with multiresolution analysis as suggested by Jørgensen et al. (2013) , denoted the Generalized Power Spectrum Model (GPSM), which has been demonstrated to predict the results of several experiments on psychoacoustic masking and speech intelligibility ( Biberger & Ewert, 2016 , 2017 ). Recently, Biberger and colleagues suggested the Generalized Power Spectrum Model for quality (GPSM q ; Biberger et al., 2018 ) that has been shown to predict the perception of a large variety of monaural distortions.…”

A two‐path model of auditory modulation detection using temporal fine structure and envelope cues

“…Viemeister, ; Dau et al ., ; Ewert & Dau, ) and the most recent attempts successfully described a wide range of behavioural data ( e.g . Piechowiak et al ., ; Jepsen et al ., ; Dau et al ., ; Biberger & Ewert, , ). In comparison, very few modelling attempts have been made to model for FM sensitivity (Hartmann & Klein, ; Moore & Sek, ; Ernst & Moore, ; Paraouty et al ., ).…”