Stephan M. A. Ernst

Frequency difference limens at high frequencies: Evidence for a transition from a temporal to a place code

It is commonly believed that difference limens for frequency (DLFs) for pure tones depend on a temporal mechanism (phase locking) for frequencies up to 4-5 kHz and a place mechanism at higher frequencies. The DLFs predicted from a place mechanism, expressed as a proportion of center frequency (Δf/f), should be approximately invariant with frequency at medium to high frequencies. If there is a transition from a temporal to a place mechanism, Δf/f should increase with increasing center frequency until the transition occurs, and then reach a plateau. Published data do not show such an effect. In this study, DLFs were measured for center frequencies from 2 to 14 kHz, using earphones designed to produce a flat response at the eardrum. The level of every tone was varied over a range of ±4 dB, to reduce loudness cues. The value of Δf/f increased progressively from 2 to 8 kHz, but did not change significantly for frequencies from 8 to 14 kHz. The results are consistent with the idea that there is a transition from a temporal to a place mechanism at about 8 kHz, rather than at 4-5 kHz, as is commonly assumed.

Spatial dissociation of changes of level and signal-to-noise ratio in auditory cortex for tones in noise.

Functional magnetic resonance imaging has been used to investigate the signal representation in human auditory cortex for a sinusoidal signal in the presence of a noise masker. This paradigm is widely used in auditory research to study auditory processing. Five-note tonal melodies were presented in a masking noise for signal-to-noise ratios (S/N) from -18 dB to+24 dB in 6 dB-steps. For small S/N (-18 dB, -12 dB, -6 dB) the overall level of the sound is nearly constant, but the audibility of the tone varies with S/N. For S/N of 0 dB and above, the tone is always clearly audible, and the perceived change is mainly the increase in overall level. This interaction between S/N, overall level and perception is reflected by a spatial dissociation of the respective activation in auditory cortex. Brain regions mainly sensitive to level changes were found in various parts of the superior temporal lobes, including primary auditory cortex and Planum temporale, while those regions mainly sensitive to S/N changes were located at or close to lateral Heschls gyrus. The overlap between these two regions is small. The results are interpreted as indicating that the coding of overall level and, thus, loudness is different from the coding of audibility of a periodic signal. The S/N-sensitive region largely overlaps with the pitch-sensitive regions in lateral Heschls gyrus found in previous studies. The results from the present study further suggest that the audibility of a tone in noise is related to the overall pitch strength.

Role of suppression and retro-cochlear processes in comodulation masking release

Recent physiological studies suggest that comodulation masking release (CMR) could be a consequence of wideband inhibition at the level of the cochlear nucleus. The present study investigates whether the existence region of psychophysical CMR is comparable to the inhibitory areas of units showing a physiological correlate of CMR. Since the inhibitory areas are similar to suppressive regions at the level of the basilar membrane, the amount of CMR that can be accounted for by suppression was determined by predicting the data with a model incorporating a peripheral nonlinearity. A CMR of up to 6 dB could still be experimentally observed for a flanking band (FB) four octaves below the on-frequency masker (OFM). For FB frequencies below the OFM, the suggested model predicts CMR equal to the measured CMR for high levels of the FB. The model underestimates the magnitude of CMR for midlevels of the FB, indicating that suppression alone cannot account for CMR. The data are consistent with the hypothesis that wideband inhibition plays a role in CMR.

The role of time and place cues in the detection of frequency modulation by hearing-impaired listeners.

Frequency modulation detection limens (FMDLs) were measured for five hearing-impaired (HI) subjects for carrier frequencies f(c) = 1000, 4000, and 6000 Hz, using modulation frequencies f(m) = 2 and 10 Hz and levels of 20 dB sensation level and 90 dB SPL. FMDLs were smaller for f(m) = 10 than for f(m) = 2 Hz for the two higher f(c), but not for f(c) = 1000 Hz. FMDLs were also determined with additional random amplitude modulation (AM), to disrupt excitation-pattern cues. The disruptive effect was larger for f(m) = 10 than for f(m) = 2 Hz. The smallest disruption occurred for f(m) = 2 Hz and f(c) = 1000 Hz. AM detection thresholds for normal-hearing and HI subjects were measured for the same f(c) and f(m) values. Performance was better for the HI subjects for both f(m). AM detection was much better for f(m)  = 10 than for f(m) = 2 Hz. Additional tests showed that most HI subjects could discriminate temporal fine structure (TFS) at 800 Hz. The results are consistent with the idea that, for f(m) = 2 Hz and f(c) = 1000 Hz, frequency modulation (FM) detection was partly based on the use of TFS information. For higher carrier frequencies and for all carrier frequencies with f(m) = 10 Hz, FM detection was probably based on place cues.

Peripheral and central aspects of auditory across-frequency processing.

Many natural sounds such as, e.g., speech show common level fluctuations across frequency. It is generally assumed that the auditory system uses this spectro-temporal information to group the frequency components into auditory objects although the exact physiological mechanism is still not fully understood. The aim of the present study is to disentangle the relative contribution of peripheral and central aspects of this across-frequency processing using psychophysical experiments and modelling. The study focuses on two different psychophysical phenomena which are thought to be related to the ability to compare information across frequency: comodulation masking release (CMR), i.e., a release from masking of a sinusoidal signal due to the addition of a comodulated off-frequency masker component to the masker component at the signal frequency, and comodulation detection difference (CDD), i.e., the reduced ability of the auditory system to detect a masked signal if masker and signal share the same envelope. The comparison between model predictions and experimental results indicates that a considerable amount of these effects can be accounted for by peripheral processing alone. This is confirmed by experimental results with confounding across-frequency information about the grouping of the different frequencies into auditory objects.

Mechanisms underlying the detection of frequency modulation

Frequency modulation detection limens (FMDLs) were measured for carrier frequencies (f(c)) of 1000, 4000, and 6000 Hz, using modulation frequencies (f(m)) of 2 and 10 Hz and levels of 20 and 60 dB sensation level (SL), both with and without random amplitude modulation (AM), applied in all intervals of a forced-choice trial. The AM was intended to disrupt excitation-pattern cues. At 60 dB SL, the deleterious effect of the AM was smaller for f(m) = 2 than for f(m) = 10 Hz for f(c) = 1000 and 4000 Hz, respectively, while for f(c) = 6000 Hz the deleterious effect was large and similar for the two values of f(m). This is consistent with the idea that, for f(c) below about 5000 Hz and f(m) = 2 Hz, frequency modulation can be detected via changes in phase locking over time. However, at 20 dB SL, the deleterious effect of the added AM for f(c) = 1000 and 4000 Hz was similar for the two values of f(m), while for f(c) = 6000 Hz, the deleterious effect of the AM was greater for f(m) = 10 than for f(m) = 2 Hz. It is suggested that, at low SLs, the auditory filters become relatively sharp and phase locking weakens, so that excitation-pattern cues influence FMDLs even for low f(c) and low f(m).

Comparing Binaural Pre-processing Strategies I: Instrumental Evaluation

In a collaborative research project, several monaural and binaural noise reduction algorithms have been comprehensively evaluated. In this article, eight selected noise reduction algorithms were assessed using instrumental measures, with a focus on the instrumental evaluation of speech intelligibility. Four distinct, reverberant scenarios were created to reflect everyday listening situations: a stationary speech-shaped noise, a multitalker babble noise, a single interfering talker, and a realistic cafeteria noise. Three instrumental measures were employed to assess predicted speech intelligibility and predicted sound quality: the intelligibility-weighted signal-to-noise ratio, the short-time objective intelligibility measure, and the perceptual evaluation of speech quality. The results show substantial improvements in predicted speech intelligibility as well as sound quality for the proposed algorithms. The evaluated coherence-based noise reduction algorithm was able to provide improvements in predicted audio signal quality. For the tested single-channel noise reduction algorithm, improvements in intelligibility-weighted signal-to-noise ratio were observed in all but the nonstationary cafeteria ambient noise scenario. Binaural minimum variance distortionless response beamforming algorithms performed particularly well in all noise scenarios.

Cortical representation of release from auditory masking

The aim of the present study was to find a functional MRI correlate in human auditory cortex of the psychoacoustical effect of release from masking, using amplitude-modulated noise stimuli. A sinusoidal target signal was embedded in a band-limited white noise, which was either unmodulated or (co)modulated. Psychoacoustical thresholds were measured for the target signals in both types of masking noise, using an adaptive procedure. The mean threshold difference between the unmodulated and the comodulated condition, i.e., the release from masking, was 15 dB. The same listeners then participated in an fMRI experiment, recording activation of auditory cortex in response to tones in the presence of modulated and unmodulated noise maskers at five different signal-to-noise ratios. In general, a spatial dissociation of changes of overall level and signal-to-noise ratio in auditory cortex was found, replicating a previous fMRI study on pure-tone masking. The comparison of the fMRI activation maps for a signal presented in modulated and in unmodulated noise reveals that those regions in the antero-lateral part of Heschls gyrus previously shown to represent the audibility of a tonal target (rather than overall level) exhibit a stronger activation for the modulated than for the unmodulated conditions. This result is interpreted as a physiological correlate of the psychoacoustical effect of comodulation masking release at the level of the auditory cortex.

Suppression and comodulation masking release in normal-hearing and hearing-impaired listeners

The detectability of a sinusoidal signal embedded in a masker at the signal frequency can be improved by simultaneously presenting additional maskers in off-frequency regions if the additional maskers and the on-frequency masker component have the same temporal envelope. This effect is commonly referred to as comodulation masking release (CMR). Recently, it was hypothesized that peripheral nonlinear processes such as suppression may play a role in CMR over several octaves when the level of the off-frequency masker component is higher than the level of the on-frequency masker component. The aim of the present study was to test this hypothesis by measuring suppression and CMR within the same subjects for various frequency-level combinations of the off-frequency masker component. Experimental data for normal-hearing listeners show a large overlap between the existence regions for suppression and CMR. Hearing-impaired subjects with a sensorineural hearing loss show, on average, negligible suppression and CMR. The data support the hypothesis that part of the CMR in experiments with large spectral distances and large level differences between the masker components is due to the nonlinear processing at the level of the cochlea.

Comodulation masking release for regular and irregular modulators

The present study investigates whether the difference in comodulation masking release (CMR) for different modulator types is due to the different degrees of modulator regularity, as suggested in the literature, or results from different envelope distributions. Thresholds of a sinusoidal signal are measured in the presence of a noise masker which was either broadband or narrow band. A square-wave modulator with different degrees of regularity is used that preserves the envelope distribution. The measured CMR does not decrease as the regularity decreases. This finding argues against the hypothesis that the difference in CMR for different modulator types reported in the literature is due to differences in regularity. The data for the narrow-band conditions which either mimic the auditory frequency selectivity or preserve the modulation spectrum indicate that most of the CMR of the present study is due to within-channel cues. In agreement with this finding, within-channel models using either a peripheral nonlinearity or a modulation filterbank predict a CMR of a similar size. In contrast to the model predictions and the findings for the narrow-band conditions, the CMR for the broadband masker increases as the regularity decreases. This suggests that the CMR is not solely determined by the envelope distributions.