Asymmetry of masking between noise and iterated rippled noise: evidence for time-interval processing in the auditory system

Ripple density resolution dependence on ripple width

The goal of the study was to investigate how variations in ripple width influence the ripple density resolution. The influence of the ripple width was investigated with two experimental paradigms: (i) discrimination between a rippled test signal and a rippled reference signal with opposite ripple phases and (ii) discrimination between a rippled test signal and a flat reference signal. The ripple density resolution depended on the ripple width: the narrower the width, the higher the resolution. For distinguishing between two rippled signals, the resolution varied from 15.1 ripples/oct at a ripple width of 9% of the ripple frequency spacing to 8.1 ripples/oct at 64%. For distinguishing between a rippled test signal and a non-rippled reference signal, the resolution varied from 85 ripples/oct at a ripple width of 9% to 9.3 ripples/oct at a ripple width of 64%. For distinguishing between two rippled signals, the result can be explained by the increased ripple depth in the excitation pattern due to the widening of the inter-ripple gaps. For distinguishing between a rippled test signal and a non-rippled reference signal, the result can be explained by the increased ratio between the autocorrelated and uncorrelated components of the input signal.

Noise thresholds in harmonic series maskers

The presence of noise is a salient cue to the perception of breathiness and aspiration in speech sounds. The detection of noise within harmonic series (maskers) composed of unresolved components was found to depend on the fundamental frequency (f_o) and the overall level of the masker [Gockel, Moore, and Patterson (2002). J. Acoust. Soc. Am., 111 (6), 2759-2770]. In the present study, noise detection thresholds were measured as a function of the frequency range, the f_o, and the overall level of harmonic maskers. Frequency range was specified in equivalent rectangular bandwidth (ERB) units (3-13, 13-23, 23-33, or 3-33 ERBs). The results were consistent with the idea that listeners rely on spectral cues when maskers comprise only resolved components (3-13 ERBs), and on temporal (dip listening) cues when maskers contain only unresolved components (23-33 ERBs). Noise detection thresholds were generally lower when masker level was high (70 dBA) than when it was low (50 dBA). Masker f_o affected thresholds only when listeners relied on spectral cues for noise detection. With the wideband (3-33 ERBs) masker, listeners likely detected noise by focusing on the frequency band (23-33 ERBs) with the most advantageous noise-to-harmonic ratio.

Estimates of Ripple-Density Resolution Based on the Discrimination From Rippled and Nonrippled Reference Signals

Rippled-spectrum stimuli are used to evaluate the resolution of the spectro-temporal structure of sounds. Measurements of spectrum-pattern resolution imply the discrimination between the test and reference stimuli. Therefore, estimates of rippled-pattern resolution could depend on both the test stimulus and the reference stimulus type. In this study, the ripple-density resolution was measured using combinations of two test stimuli and two reference stimuli. The test stimuli were rippled-spectrum signals with constant phase or rippled-spectrum signals with ripple-phase reversals. The reference stimuli were rippled-spectrum signals with opposite ripple phase to the test or nonrippled signals. The spectra were centered at 2 kHz and had an equivalent rectangular bandwidth of 1 oct and a level of 70 dB sound pressure level. A three-alternative forced-choice procedure was combined with an adaptive procedure. With rippled reference stimuli, the mean ripple-density resolution limits were 8.9 ripples/oct (phase-reversals test stimulus) or 7.7 ripples/oct (constant-phase test stimulus). With nonrippled reference stimuli, the mean resolution limits were 26.1 ripples/oct (phase-reversals test stimulus) or 22.2 ripples/oct (constant-phase test stimulus). Different contributions of excitation-pattern and temporal-processing mechanisms are assumed for measurements with rippled and nonrippled reference stimuli: The excitation-pattern mechanism is more effective for the discrimination of rippled stimuli that differ in their ripple-phase patterns, whereas the temporal-processing mechanism is more effective for the discrimination of rippled and nonrippled stimuli.

Rippled-spectrum resolution dependence on frequency: Estimates obtained by discrimination from rippled and nonrippled reference signals

The resolution of spectral ripples is a useful test for the spectral resolution of hearing. However, the use of different measurement paradigms might yield diverging results because of a paradigm-dependent contribution of excitation-pattern and temporal-processing mechanisms. In the present study, ripple-density resolution was measured in normal-hearing listeners for several frequency bands (centered at 0.5, 1, 2, and 4 kHz), using two paradigms: (i) discrimination of a rippled-spectrum test signal from a rippled reference signal differing by the ripple phase pattern, and (ii) discrimination of a rippled-spectrum test signal from a nonrippled reference signal. For the rippled reference signals, the resolution slightly depended on signal frequency. For the nonrippled reference signals, the resolution depended on the signal frequency; it varied from 8.8 ripples/oct at 0.5 kHz to 34.2 ripples/oct at 4 kHz. Excitation-pattern and temporal-processing models of spectral analysis were considered. Predictions of the excitation-pattern model agreed with the data obtained with the rippled reference signals. In contrast, predictions of the temporal-processing model agreed with the data obtained with the nonrippled reference signals. Thus, depending on the used reference signal type, the ripple-density resolution estimates characterize the discrimination abilities of the corresponding mechanisms.

Hearing sensitivity to gliding rippled spectrum patterns

The sensitivity of human hearing to gliding rippled spectrum patterns of sound was investigated. The test signal was 2-oct wide rippled noise with the ripples gliding along the frequency scale. Both ripple density and gliding velocity were frequency-proportional across the signal band; i.e., the density was specified in ripples/oct and the velocity was specified in oct/s and ripple/s. The listener was required to discriminate between a test signal with gliding ripples and a non-rippled reference signal. Limits of gliding velocity were measured as a function of ripple density. The ripple gliding velocity limit decreased with an increasing ripple density: from 388.9 oct/s (388.9 ripple/s) at a ripple density of 1 ripple/oct to 11.3 oct/s (79.1 ripple/s) at a density of 7 ripple/oct. These tendencies could be approximated by log/log regression functions with slopes of 1.71 for the velocity expressed in oct/s and 0.71 for the velocity expressed in ripple/s. A qualitative model based on combined action of the excitation-pattern and the temporal-processing mechanism is suggested to explain the results.

Hearing sensitivity to shifts of rippled-spectrum patterns

The sensitivity of human hearing to shifts in rippled spectrum patterns of sound was investigated. The test signal was band-limited rippled noise with spectrum ripples of various frequency spacing and bandwidth; this type of sound may be considered as a quantitatively controlled imitation of complex natural sounds. The listener was required to detect a shift in the spectrum ripple phase while keeping the other parameters of the noise constant. For cosine-shaped ripples, the lowest threshold (1.1%) was found at a ripple frequency of 3.5 ripples per octave (rpo), which corresponds to a ripple spacing of 20% of the center frequency. The threshold increased for both lower and higher ripple densities. Qualitatively similar patterns of threshold dependence on ripple density were observed for center frequencies from 1 to 4 kHz. Making the ripples narrower than cosine decreased the thresholds to 0.7%-0.75% for ripple densities of 2-5 rpo. Keeping the ripple width constant at 3.5%-7.5% of the frequency resulted in a monotonic threshold dependence on ripple density: The threshold decreased with decreasing density (down to 0.7%). An excitation-pattern model explains qualitatively the observed dependence of the ripple-phase shift threshold on ripple pattern parameters.

Loudness perception of low tones undergoing partial masking by higher tones in orchestral music in concert halls

Objective acoustical parameters for halls are often measured in 1-octave bands with mid-frequencies from 125 to 4000 Hz. In reality, the frequency range of musical instruments is much wider than that, and the fundamentals of the lower notes of bass instruments are contained in 31.5 or 63 Hz bands. Overtones of fundamentals in these bands fall in 125 Hz band. This report presents subjective experiments designed to determine to what extent the overtones in 125 Hz band and higher bands influence the loudness sensation of the components in 63 Hz band. In the experiments, the 125 Hz and higher components of the musical tone are used to act as a masker against the lower component used as a maskee. The threshold of the difference between G(125 Hz) and G(lower band) that just enables one to hear the fundamental tones in the lower band is determined. Masked loudness of 63 Hz sinusoidal tone caused by partial masking noise with higher frequencies was determined based on a similar procedure to the masked loudness-matching function. The result indicates that the difference in loudness of low tone will not be noticeable even if G changed by ±2.5 to ±3 dB, at least when there are other accompanying instruments.

Asymmetry of masking in the European starling: behavioural auditory thresholds

Psychophysical studies of simultaneous masking with human observers exhibit an asymmetry in the amount of masking that depends on the relative bandwidths of signals and maskers. For noise bands up to the bandwidth of one auditory filter, masked auditory thresholds are considerably lower when the bandwidth of the signal exceeds that of the masker compared to the reversed condition. We investigate asymmetry of masking in an animal model, that will allow to study the mechanisms associated with the asymmetry of masking effect. European starlings (Sturnus vulgaris) were trained in a Go/NoGo paradigm to report the detection of a 500 ms noise signal centred in a 700 ms noise masker. Signals and maskers with centre frequencies of 2 kHz had bandwidths of 4 Hz or 256 Hz. Thresholds for detecting the 256 Hz wide-band signal in a 4 Hz narrow-band masker were considerably lower compared to detecting the 4 Hz narrow-band signal in a 256 Hz wide masker and compared to all other conditions. The asymmetry of masking in starlings was on average 15 and 17 dB for 40 and 70 dB SPL overall masker level, respectively. Our animal model thus proved perceptual abilities similar to human subjects. The results are discussed with respect to the importance of both intensity and temporal cues for signal detection.

Some problems in the measurement of the frequency-resolving ability of hearing

Despite the detailed development of masking methods for measurement of the frequency selectivity of hearing, these measurements are hardly used for diagnostic purposes because they are time-consuming and because of the uncertain extrapolation of the results to the perception of complex spectral patterns. A method for the direct measurement of the spectral resolving ability of hearing using test signals with rippled spectra is proposed. These measurements showed 1) that the resolving ability of the auditory system in terms of discriminating complex spectra is greater than that suggested by the acuity of auditory frequency filters; 2) that changes in the acuity of frequency auditory filters associated with sound intensity hardly affect the ability to resolve complex spectra; 3) that the effects of interference on frequency-resolving ability do not lead to decreases in the spectral contrast of signals due to superimposition of noise.

The effect of cross-channel synchrony on the perception of temporal regularity

Temporal models of pitch are based on the assumption that the auditory system measures the time intervals between neural events, and that pitch corresponds to the most common time interval. The current experiments were designed to test whether time intervals are analyzed independently in each peripheral channel, or whether the time-interval analysis in one channel is affected by synchronous activity in other channels. Regular and irregular click trains were filtered into narrow frequency bands to produce target and flanker stimuli. The threshold for discriminating a regular target from an irregular distracter click train was measured in the presence of an irregular masker click train in the target band, as a function of the frequency separation between the target band and a flanker band. The flanker click train was either regular or irregular. The threshold for detecting the regular target was 5-7 dB lower when the flanker was regular. The data indicate that the detection of temporal regularity (and thus, pitch) involves cross-channel processes that can operate over widely separated channels. Model simulations suggest that these cross-channel processes occur after the time-interval extraction stage and that they depend on the similarity, or consistency, of the time-interval patterns in the relevant channels.

Instantaneous frequency decomposition: an application to spectrally sparse sounds with fast frequency modulations

Classical time-frequency analysis is based on the amplitude responses of bandpass filters, discarding phase information. Instantaneous frequency analysis, in contrast, is based on the derivatives of these phases. This method of frequency calculation is of interest for its high precision and for reasons of similarity to cochlear encoding of sound. This article describes a methodology for high resolution analysis of sparse sounds, based on instantaneous frequencies. In this method, a comparison between tonotopic and instantaneous frequency information is introduced to select filter positions that are well matched to the signal. Second, a cross-check that compares frequency estimates from neighboring channels is used to optimize filter bandwidth, and to signal the quality of the analysis. These cross-checks lead to an optimal time-frequency representation without requiring any prior information about the signal. When applied to a signal that is sufficiently sparse, the method decomposes the signal into separate time-frequency contours that are tracked with high precision. Alternatively, if the signal is spectrally too dense, neighboring channels generate inconsistent estimates-a feature that allows the method to assess its own validity in particular contexts. Similar optimization principles may be present in cochlear encoding.

MEG responses to rippled noise and Huggins pitch reveal similar cortical representations

The onset of pitch within an ongoing noise signal evokes a particular brain activity, the pitch onset response (POR). Using whole-head MEG, PORs to iterated rippled noise (IRN) and Huggins pitch (HP), representing prototypical pitch-in-noise signals, were measured in twenty subjects during a pitch identification task (333 Hz, 400 Hz, randomized). HP and IRN yielded similar responses, lateralized to the left hemisphere and peaking about 180 ms after pitch onset. The initial phase (140 ms) showed stronger activations to 400 than to 333 Hz whereas later stages (200-300 ms) showed target vs nontarget effects. These results suggest, first, that different pitches converge into a common cortical representation and, second, that the POR encompasses various successive processing stages.

Time course and hemispheric lateralization effects of complex pitch processing: evoked magnetic fields in response to rippled noise stimuli

To delineate the time course and processing stages of pitch encoding at the level of the supratemporal plane, the present study recorded evoked magnetic fields in response to rippled noise (RN) stimuli. RN largely masks simple tonotopic representations and addresses pitch processing within the temporal domain (periodicity encoding). Four dichotic stimulus types (111 or 133 Hz RN at one ear, white noise to the other one) were applied in randomized order during either visual distraction or selective auditory attention. Strictly periodic signals, noise-like events, and mixtures of both signals served as control conditions. (1) Attention-dependent ear x hemisphere interactions were observed within the time domain of the M50 field, indicating early streaming of auditory information. (2) M100 responses to strictly periodic stimuli were found lateralized to the right hemisphere. Furthermore, the higher-pitched stimuli yielded enhanced activation as compared to the lower-pitch signals (pitch scaling), conceivably reflecting sensory memory operations. (3) Besides right-hemisphere pitch scaling, the relatively late M100 component in association with the RN condition (latency = 136 ms) showed significantly stronger field strengths over the left hemisphere. Control experiments revealed this lateralization effect to be related to noise rather than pitch processing. Furthermore, subtle noise variations interacted with signal periodicity. Obviously, thus, complex task demands such as RN encoding give rise to functional segregation of auditory processing across the two hemispheres (left hemisphere: noise, right hemisphere: periodicity representation). The observed noise/periodicity interactions, furthermore, might reflect pitch-synchronous spectral evaluation at the level of the left supratemporal plane, triggered by right-hemisphere representation of signal periodicity.

Transient and phase-locked evoked magnetic fields in response to periodic acoustic signals

Using whole-head MEG, time course and hemispheric lateralization effects of phase-locked brain responses to complex periodic acoustic signals (stimulus frequency 13, 22, 40, 67, or 111 Hz) were determined based on a dipole analysis approach. Apart from systematic rate-induced changes in amplitude and shape of the transient evoked magnetic fields (M50, M100), phase-locked brain activity emerged, being more pronounced over the right as compared to the left hemisphere. Furthermore, this MEG component showed a consistent phase angle across subjects, indicating active synchronization mechanisms within auditory cortex that operate upon afferent input. Conceivably, these early side-differences in periodicity encoding contribute to or even snowball into hemispheric lateralization effects of higher-order aspects of central-auditory processing such as melody perception.

Rippled-spectrum resolution dependence on level

Rippled-density resolution of a rippled sound spectrum (probe band) in both the presence and absence of another band (masker) was studied as a function of sound level in normal listeners. The resolvable ripple density in the probe band was measured by finding the highest ripple density at which an interchange of ripple peak and valley positions was detectable (the phase-reversal test). Probe bands were 0.5 oct wide with center frequencies of 1, 2, and 4 kHz. In the control condition (no masker), the ripple-density resolution was almost independent of sound level within a range of 40-90 dB SPL. When an on-frequency masker coincided with the probe band (that resulted in reduced ripple depth), resolution decreased slightly relative to the control condition but remained little dependent on level. With an off-frequency low-side masker, the ripple-density resolution was a little less than in the control but almost independent of level within a range of 40-60 dB SPL and progressively decreased with level increase from 70 to 90 dB SPL. The dependence on level was qualitatively similar at all probe frequencies and at various widths and positions of the low-side off-frequency masker band.

Mechanisms determining the salience of coloration in echoed sound: influence of interaural time and level differences

This study investigates whether the salience of the pitch associated with a single reflection of a broadband sound, such as noise, is determined by the monaural information mediated by the stimuli at the two ears, or by the relative locations of the primary sound and the reflection. Pitch strength was measured as a function of the reflection delay and the lateral displacement between the primary sound and the reflection. Thereby, lateral displacement was produced by means of interaural time differences (ITDs) in experiment 1 and interaural level differences (ILDs) in experiment 3. The results from both experiments are in accordance with the assumption that the strength of the pitch associated with a reflection is based on a central average of the internal representations of the stimuli at the two ears. This notion was corroborated by experiment 2, which showed that the results from experiment 1 could be mimicked by simply adding the stimuli from the two ears and presenting the merged stimulus identically to both ears.

Asymmetry of masking between complex tones and noise: partial loudness

This experiment examined the partial masking of periodic complex tones by a background of noise, and vice versa. The tones had a fundamental frequency (F0) of 62.5 or 250 Hz, and components were added in either cosine phase (CPH) or random phase (RPH). The tones and the noise were bandpass filtered into the same frequency region, from the tenth harmonic up to 5 kHz. The target alone was alternated with the target and the background; for the mixture, the background and target were either gated together, or the background was turned on 400 ms before, and off 200 ms after, the target. Subjects had to adjust the level of either the target alone or the target in the background so as to match the loudness of the target in the two intervals. The overall level of the background was 50 dB SPL, and loudness matches were obtained for several fixed levels of the target alone or in the background. The resulting loudness-matching functions showed clear asymmetry of partial masking. For a given target-to-background ratio, the partial loudness of a complex tone in a noise background was lower than the partial loudness of a noise in a complex tone background. Expressed as the target-to-background ratio required to achieve a given loudness, the asymmetry typically amounted to 12-16 dB. When the F0 of the complex tone was 62.5 Hz, the asymmetry of partial masking was greater for CPH than for RPH. When the F0 was 250 Hz, the asymmetry was greater for RPH than for CPH. Masked thresholds showed the same pattern as for partial masking for both F0's. Onset asynchrony had some effect on the loudness matching data when the target was just above its masked threshold, but did not significantly affect the level at which the target in the background reached its unmasked loudness. The results are interpreted in terms of the temporal structure of the stimuli.

Microsecond temporal resolution in monaural hearing without spectral cues?

The auditory system encodes the timing of peaks in basilar-membrane motion with exquisite precision, and perceptual models of binaural processing indicate that the limit of temporal resolution in humans is as little as 10-20 microseconds. In these binaural studies, pairs of continuous sounds with microsecond differences are presented simultaneously, one sound to each ear. In this paper, a monaural masking experiment is described in which pairs of continuous sounds with microsecond time differences were combined and presented to both ears. The stimuli were matched in terms of the excitation patterns they produced, and a perceptual model of monaural processing indicates that the limit of temporal resolution in this case is similar to that in the binaural system.

Asymmetry of masking between complex tones and noise: the role of temporal structure and peripheral compression

Thresholds for the detection of harmonic complex tones in noise were measured as a function of masker level. The rms level of the masker ranged from 40 to 70 dB SPL in 10-dB steps. The tones had a fundamental frequency (F0) of 62.5 or 250 Hz, and components were added in either cosine or random phase. The complex tones and the noise were bandpass filtered into the same frequency region, from the tenth harmonic up to 5 kHz. In a different condition, the roles of masker and signal were reversed, keeping all other parameters the same; subjects had to detect the noise in the presence of a harmonic tone masker. In both conditions, the masker was either gated synchronously with the 700-ms signal, or it started 400 ms before and stopped 200 ms after the signal. The results showed a large asymmetry in the effectiveness of masking between the tones and noise. Even though signal and masker had the same bandwidth, the noise was a more effective masker than the complex tone. The degree of asymmetry depended on F0, component phase, and the level of the masker. The maximum difference between masked thresholds for tone and noise was about 28 dB; this occurred when the F0 was 62.5 Hz, the components were in cosine phase, and the masker level was 70 dB SPL. In most conditions, the growth-of-masking functions had slopes close to 1 (on a dB versus dB scale). However, for the cosine-phase tone masker with an F0 of 62.5 Hz, a 10-dB increase in masker level led to an increase in masked threshold of the noise of only 3.7 dB, on average. We suggest that the results for this condition are strongly affected by the active mechanism in the cochlea.

Sustained magnetic fields reveal separate sites for sound level and temporal regularity in human auditory cortex

Magnetoencephalography was used to investigate the relationship between the sustained magnetic field in auditory cortex and the perception of periodic sounds. The response to regular and irregular click trains was measured at three sound intensities. Two separate sources were isolated adjacent to primary auditory cortex: One, located in lateral Heschl's gyrus, was particularly sensitive to regularity and largely insensitive to sound level. The second, located just posterior to the first in planum temporale, was particularly sensitive to sound level and largely insensitive to regularity. This double dissociation to the same stimuli indicates that the two sources represent separate mechanisms; the first would appear to be involved with pitch perception and the second with loudness. The delay of the offset of the sustained field was found to increase with interclick interval up to 200 ms at least, which suggests that the sustained field offset represents a sophisticated offset-monitoring mechanism rather than simply the cessation of stimulation.