Comodulation masking release as a function of bandwidth and time delay between on-frequency and flanking-band maskers

The threshold for a signal masked by a narrow band of noise centered at the signal frequency (the on-frequency band) may be reduced by adding to the masker a second band of noise (the flanking band) whose envelope is correlated with that of the first band, an effect called comodulation masking release (CMR). This paper examines CMR as a function of masker bandwidth and time delay between the envelopes of the on-frequency and flanking bands. The 1.0-kHz sinusoidal signal had a duration of 400 ms. The on-frequency band was presented alone (reference condition) or with the flanking band. The flanking-band envelope was either correlated or uncorrelated with that of the on-frequency band. Flanking-band center frequencies ranged from 0.25-2.0 kHz. The flanking band was presented either in the same ear as the on-frequency band (monaural condition) or in the opposite ear (dichotic condition). The noise bands had bandwidths of 6.25, 25, or 100 Hz. In the correlated conditions, the flanking-band envelope was delayed with respect to that of the on-frequency band by 0, 5, 10, or 20 ms. For the 100-Hz bandwidth, CMRs were small (typically less than 1 dB) in both monaural and dichotic conditions at all delay times. For the 25-Hz bandwidth, CMRs were about 3.5 dB for the 0-ms delay, and decreased to about 1.5 dB for the 20-ms delay. For the 6.25-Hz bandwidth, CMRs averaged about 5 dB and were almost independent of delay time. The results suggest that the absolute delay time is not the critical variable determining CMR. The magnitude of CMR appears to depend on the correlation between the envelopes of the on-frequency and flanking bands. However, the results do not support a model of CMR that assumes that signal threshold corresponds to a constant change in across-band envelope correlation when the correlation is transformed to Fisher's z.     

Comodulation masking release (CMR): effects of signal frequency, flanking-band frequency, masker bandwidth, flanking-band level, and monotic versus dichotic presentation of the flanking band

In experiment I, thresholds for 400-ms sinusoidal signals were measured in the presence of a continuous 25-Hz-wide noise centered at signal frequencies (fs) ranging from 250 to 8000 Hz in 1-oct steps. The masker was presented either alone or together with a second continuous 25-Hz-wide band of noise (the flanking band) whose envelope was either correlated with that of the on-frequency band or was uncorrelated; its center frequency ranged from 0.5 fs to 1.5 fs. The flanking band was presented either in the same ear (monotic condition) as the signal plus masker or in the opposite ear (dichotic condition). The on-frequency band and the flanking band each had an overall level of 67 dB SPL. The comodulation masking release, CMR (U-C), is defined as the difference between the thresholds for the uncorrelated and correlated conditions. The CMR (U-C) showed two components: a broadly tuned component, occurring at all signal frequencies and all flanking-band frequencies, and occurring for both monotic and dichotic conditions; and a component restricted to the monotic condition and to flanking-band frequencies close to fs. This sharply tuned component was small for low signal frequencies, increased markedly at 2000 and 4000 Hz, and decreased at 8000 Hz. Experiment II showed that the sharply tuned component of the CMR (U-C) was slightly reduced in magnitude when the level of the flanking band was 10 dB above that of the on-frequency band and was markedly reduced when the level was 10 dB below, whereas the broadly tuned component and the dichotic CMR (U-C) were only slightly affected. Experiment III showed that the sharply tuned component of the CMR (U-C) was markedly reduced when the bandwidths of the on-frequency and flanking bands were increased to 100 Hz, while the broadly tuned component and the dichotic CMR (U-C) decreased only slightly. The argument here is that the sharply tuned component of the monotic CMR (U-C) results from beating between the "carrier" frequencies of the two masker bands. This introduces periodic zeros in the masker envelope, which facilitate signal detection. The broadly tuned component, which is probably a "true" CMR, was only about 3 dB.     

Comodulation detection differences using noise-band signals

In a variant of the standard paradigm employed to study comodulation masking release (CMR), a narrow noise band was used as a signal in the presence of "cue" bands which had either the same or different temporal envelopes. The number of cue bands present ranged from zero to four; when there were two or four cue bands, they were either all presented at the same overall level or the spectral profile was "scrambled" in a haphazard manner. Different noise samples were presented within and across trials. The result was in the opposite direction from the standard CMR outcome; that is, better performance was obtained when the envelopes of the cue band(s) were uncorrelated with those of the signal band than when they were correlated. These comodulation detection differences (CDDs) ranged from a decibel or two up to 10-12 dB in different conditions, and were generally larger the more cue bands present. Standard CMR conditions, which were run as controls, revealed that the detectability of a tonal signal does not increase as the number of cue bands is increased from one to four-an outcome which differs from those obtained in profile analysis experiments. The data taken with the equal-level and the scrambled-level cues differed little in both the CDD and the CMR conditions. All noise bands were 100 Hz wide, and approximately 250 ms in duration. The signal band in CDD and the masker band in CMR were centered at 2500 Hz. The psychophysical procedure was two-interval forced choice.     

Comodulation masking release for various monaural and binaural combinations of the signal, on-frequency, and flanking bands

The threshold for a signal masked by a narrow band of noise centered at the signal frequency (the on-frequency band) may be reduced by adding to the masker a second band of noise (the flanking band) whose envelope is correlated with that of the first band. This effect is called comodulation masking release (CMR). These experiments examine two questions. (1) How does the CMR vary with the number and ear of presentation of the flanking band(s)? (2) Is it possible to obtain a CMR when a binaural masking level difference (BMLD) is already present, and vice versa? Thresholds were measured for a 400-ms signal in a continuous 25-Hz-wide noise centered at signal frequencies (fs) of 250, 1000, and 4000 Hz. This masker was presented either alone or with one or more continuous flanking bands whose envelopes were either correlated or uncorrelated with that of the on-frequency band; their frequencies ranged from 0.5fs to 1.5fs. CMRs were measured for six conditions in which the signal, the on-frequency band, and the flanking band(s) were presented in various monaural and binaural combinations. When a single flanking band was used, the CMR was typically around 2-3 dB. The CMR increased to 5-6 dB if an additional flanking band was added. The effect of the additional band was similar whether it was in the same ear as the original band or in the opposite ear. At the lowest signal frequency, a large CMR was observed in addition to a BMLD and vice versa. At the highest signal frequency, the extra release from masking was small. The results are interpreted in terms of the cues producing the CMR and the BMLD.     

Comodulation masking release: effects of varying the level, duration, and time delay of the cue band

The phenomenon of comodulation masking release (CMR) was studied in a series of experiments. When the relative level of the correlated cue band was more than about 10 dB less than that of the masker band, the CMR was abolished. When the duration of the tonal signal was varied with continuous maskers and cues, the course of the standard temporal-integration function (about -10 dB/decade) was followed by both the correlated-cue and the uncorrelated-cue conditions. In a burst masker paradigm employing several burst durations, the data for the correlated-cue condition closely followed the previously determined temporal-integration function. Finally, when the cue band was time delayed more than about 1.6 ms, the CMR began to decline, and it was abolished somewhere between 3 and 15 ms of delay, depending upon the subject. This latter outcome was essentially the same for masker and cue bands of both 75 and 100 Hz in width; in neither instance was there evidence of a cyclic, autocorrelation-like pattern following the period of the envelope. Supplementary experiments revealed two facts: The detectability of a masked narrow-band signal is not improved by the simultaneous presence of a correlated (or uncorrelated) noise band, and a small CMR can be obtained under conditions of forward masking.     

Comodulation masking release in a forward-masking paradigm

Waveforms that yield comodulation masking release (CMR) when they are presented simultaneously with a signal were used in a standard forward-masking procedure. The signal was a 25-ms sample of a 2500-Hz tone. The masker was a band of noise centered at 2500 Hz, 100 Hz in width, and 200 ms in duration. Presented with the masker were two or four cue bands, each 100 Hz wide and centered at various distances from the masker band. These cue bands either all had the same temporal envelope as the masker band (correlated condition) or their common envelope was different from that of the masker band (uncorrelated condition). In the initial experiments, (1) detectability of the tonal signal was 7-18 dB better when the masker band was accompanied by cue bands than when it was not--an effect that would be expected from past research on lateral suppression--but further, (2) the signal was about 3 dB more detectable in the correlated conditions than in the uncorrelated conditions. In follow-up experiments, these CMR-like differences between the correlated and uncorrelated conditions were substantially reduced (although not eliminated) by presenting a contralateral, wideband noise that was gated synchronously with the masker and/or cue bands. The implications are that the initial results were attributable in part to the "confusion effects" known to exist in certain temporal-masking situations, and that listeners are able to obtain greater information about the temporal extent of a masker band from correlated cue bands than from uncorrelated bands.(ABSTRACT TRUNCATED AT 250 WORDS)     

Comodulation masking release (CMR) as a function of masker bandwidth, modulator bandwidth, and signal duration

These experiments examine how comodulation masking release (CMR) varies with masker bandwidth, modulator bandwidth, and signal duration. In experiment 1, thresholds were measured for a 400-ms, 2000-Hz signal masked by continuous noise varying in bandwidth from 50-3200 Hz in 1-oct steps. In one condition, using random noise maskers, thresholds increased with increasing bandwidth up to 400 Hz and then remained approximately constant. In another set of conditions, the masker was multiplied (amplitude modulated) by a low-pass noise (bandwidth varied from 12.5-400 Hz in 1-oct steps). This produced correlated envelope fluctuations across frequency. Thresholds were generally lower than for random noise maskers with the same bandwidth. For maskers less than one critical band wide, the release from masking was largest (about 5 dB) for maskers with low rates of modulation (12.5-Hz-wide low-pass modulator). It is argued that this release from masking is not a "true" CMR but results from a within-channel cue. For broadband maskers (greater than 400 Hz), the release from masking increased with increasing masker bandwidth and decreasing modulator bandwidth, reaching an asymptote of 12 dB for a masker bandwidth of 800 Hz and a modulator bandwidth of 50 Hz. Most of this release from masking can be attributed to a CMR. In experiment 2, the modulator bandwidth was fixed at 12.5 Hz and the signal duration was varied. For masker bandwidths greater than 400 Hz, the CMR decreased from 12 to 5 dB as the signal duration was decreased from 400 to 25 ms.(ABSTRACT TRUNCATED AT 250 WORDS)     

Comodulation masking release and auditory grouping

The detectability of a pure-tone signal masked by a band of noise centered on the signal can be improved by the addition of flanking noise bands, provided that the temporal envelopes of the flanking bands are correlated with that of the on-signal band. This phenomenon is referred to as comodulation masking release (CMR). The present study examined CMR in conditions in which some flanking noise bands were comodulated with the on-signal band, but other flanking bands (termed "deviant" bands) were not. Past research has indicated that CMR is often substantially reduced when deviant bands are present at spectral locations close to the signal frequency. An investigation was undertaken to determine whether the disruptive effects of such bands could be reduced by factors related to auditory grouping. The signal frequency was 100 Hz. In one condition, only 20-Hz-wide comodulated bands, centered on 400, 600, 800, 1000, 1200, 1400, and 1600 Hz, were present. The CMR for this condition, referenced to threshold for the on-signal band only, was approximately 15 dB. In a second condition, two deviant bands were added at 900 and 1100 Hz; their presence reduced the CMR to only 3-4 dB. The number of deviant bands was then increased progressively, from two to eight bands. Deviant bands either shared a common envelope (codeviant), or had unique envelopes (multideviant). The number of bands that were comodulated with the on-signal band was held constant at six.(ABSTRACT TRUNCATED AT 250 WORDS)     

An analysis of quasi-frequency-modulated noise and random-sideband noise as comparisons for amplitude-modulated noise

Experiments were performed to determine under what conditions quasi-frequency-modulated (QFM) noise and random-sideband noise are suitable comparisons for AM noise in measuring a temporal modulation transfer function (TMTF). Thresholds were measured for discrimination of QFM from random-sideband noise and AM from QFM noise as a function of sideband separation. In the first experiment, the upper spectral edge of the noise stimuli was at 2400 Hz and the bandwidth was 1600 Hz. For sideband separations up to 256 Hz, at threshold sideband levels for discriminating AM from QFM noise, QFM was indiscriminable from random-sideband noise. For the largest sideband separation used (512 Hz), listeners may have used within-stimulus envelope correlation in the QFM noise to discriminate it from the random-sideband noise. Results when stimulus bandwidth was varied suggest that listeners were able to use this cue when the carrier was wider than a critical band, and the sideband separation approached the carrier bandwidth. Within-stimulus envelope correlation was also present in AM noise, and thus QFM noise was a suitable comparison because it made this cue unusable and forced listeners to use across-stimulus envelope differences. When the carrier bandwidth was less than a critical band or was wideband, QFM noise and random-sideband noise were equally suitable comparisons for AM noise. When discrimination thresholds for QFM and random-sideband noise were converted to modulation depth and modulation frequency, they were nearly identical to those for discrimination of AM from QFM noise, suggesting that listeners were using amplitude modulation cues in both cases.     

The detectability of a tone added to narrow bands of equal-energy noise.

The ability to detect a 2000-Hz tone added to bands of noise centered at 2000 Hz was measured using a two-interval, forced-choice, pulsed-masker paradigm. The stimuli ranged in duration from 50-200 ms, and the maskers ranged in bandwidth from 5-320 Hz. In one condition, the bands of noise had equal energy across the two intervals of each trial and in a second condition the levels of the stimuli were independently and randomly chosen from a 30-dB range on a presentation-by-presentation basis. The energy model failed to predict the data obtained either in the presence or in the absence of level variation. Control experiments showed that exposure to level variation yielded an overall reduction in sensitivity, suggesting that the presence of level variation leads to changes in the listeners' detection strategies. Computer simulations indicated that changes in either the fine structure or envelopes of the waveforms were sufficient to account for detection when changes in stimulus energy were not reliable.     

Signal detection in temporally modulated and spectrally shaped maskers.

The first part of this paper presents several experiments on signal detection in temporally modulated noise, yielding a general approach toward the concept of comodulation masking release (CMR). Measurements were made on masked thresholds of both long- and short-duration, narrow-band signals presented in a 100% sinusoidally amplitude-modulated (SAM) noise masker (modulation frequency 32 Hz), as a function of masker bandwidth from 1/3 oct up to 13/3 octs, while the masker band was geometrically centered at signal frequency. With the short-duration signals placed in the valley of the masker, a substantial CMR (i.e., a decrease of masked threshold with increasing masker bandwidth) was found, whereas for the long-duration signals CMR was smaller. Furthermore, investigations were carried out to determine whether CMR changes when the bandwidth of the signals, consisting of bandpass impulse responses, is increased. The data indicate that substantial CMR remains even when all masker bands contain a signal component, thus minimizing across-channel differences. This finding is not in line with current models accounting for the CMR phenomenon. The second part of this paper concerns signal detection in spectrally shaped noise. Also investigated was whether release from masking occurs for the detection of a pure-tone signal at a valley or a peak of a simultaneously presented masking noise with a sinusoidally rippled power spectrum, when this masker was preceded and followed by a second noise (temporal flanking burst) with an identical spectral shape as the on-signal noise. Similar to CMR effects for temporal modulations, the data indicate that coshaping masking release (CSMR) occurs when the signal is placed in a valley of the spectral envelope of the masker, whereas no release from masking is found when the signal is placed at a peak of the spectral envelope of the masker. The implications of these experiments for measures of spectral and temporal resolution are discussed.     

Influence of place synchrony on detection of a sinusoid

A series of experiments was performed to study the ability of the ear to code the temporal envelope of a waveform as demonstrated by comodulation masking release (CMR). The stimulus for all experiments was composed of a tone-burst signal, a 100-Hz-wide masker band centered at the signal frequency, and a second 100-Hz-wide noise band of variable frequency, the cue band. The cue band had a temporal envelope which was either correlated with or independent of that of the masker. The signal was a 100-Hz tone burst for most experiments. For the monotic stimulus, the correlated cue band results in lowered signal detection thresholds over a range extending from around 2/3 oct below the signal frequency to 1/3 oct above that frequency. When measured dichotically, with the signal and masker band in one ear and the cue band in the opposite ear, that effective range is expanded but the detection threshold shifts are a bit smaller. The greatest CMR is observed when the stimulus is presented diotically. With regard to effects of level and frequency, our data show CMR increasing with increasing stimulus level for a cue band lower in frequency than the signal, but show little effect of level for a cue band higher in frequency. Similarly, CMR increases with increasing stimulus frequency when the cue band is lower in frequency, but shows little effect of frequency for a cue band higher in frequency.     

The ability of listeners to use recovered envelope cues from speech fine structure

Recent work has demonstrated that auditory filters recover temporal-envelope cues from speech fine structure when the former were removed by filtering or distortion. This study extended this work by assessing the contribution of recovered envelope cues to consonant perception as a function of the analysis bandwidth, when vowel-consonant-vowel (VCV) stimuli were processed in order to keep their fine structure only. The envelopes of these stimuli were extracted at the output of a bank of auditory filters and applied to pure tones whose frequency corresponded to the original filters' center frequencies. The resulting stimuli were found to be intelligible when the envelope was extracted from a single, wide analysis band. However, intelligibility decreases from one to eight bands with no further decrease beyond this value, indicating that the recovered envelope cues did not play a major role in consonant perception when the analysis bandwidth was narrower than four times the bandwidth of a normal auditory filter (i.e., number of analysis bands > or =8 for frequencies spanning 80 to 8020 Hz).     

Temporal decline of masking and comodulation masking release.

Masking sounds can be continuously present, gated simultaneously with the signal, or gated somewhat prior to the signal. This continuum of relative onset times was explored using waveforms of the sort commonly employed in studies of comodulation masking release (CMR). There was a 50-Hz masker band centered on the 1250-Hz tonal signal, and four 50-Hz flanker bands centered at 850, 1050, 1450, and 1650 Hz. In some conditions, all four flanker bands had the same temporal envelope, and the masker band either had that same envelope (correlated presentations) or a different envelope (uncorrelated presentations). In other conditions, all five bands had different temporal envelopes (all-uncorrelated presentations). The masker band and/or the four flanker bands were either gated nearly simultaneously with the signal (burst conditions) or were gated prior to the signal by a duration that was systematically varied (fringed conditions). The eight listeners could be partitioned into three groups on the basis of their response to these fringing manipulations. Two listeners (the large fringers) showed a gradual improvement in detectability with increasing fringe duration (called a temporal decline of masking), while three others (the small fringers) showed little improvement in detectability. For the remaining three subjects, there was evidence of a "learning" effect that changed them from large fringers to small fringers over a 10-week period of listening. When present, the temporal decline of masking was greater for the correlated than for the uncorrelated comodulation condition; as a consequence, the difference in detectability between them (the comodulation masking release or CMR) increased with fringe duration. By fringing the masker and flanker bands separately and in combination, it was revealed that the temporal declines of masking were primarily attributable to the fringing of the flanker bands. In contrast, large CMRs required long fringes on both the masker and flanker bands. The above results were obtained with 50-ms signals, but generally similar data were obtained with a signal duration of 240 ms. The difficulties raised for experimentalists and theorists by such long-term practice effects are discussed.     

Uncertainty about the correlation among temporal envelopes in two comodulation tasks

The threshold of a 1250-Hz tonal signal was measured in the presence of five noise bands (each 50 Hz wide, centered at 850, 1050, 1250, 1450, and 1650 Hz) under five conditions of uncertainty about the waveform type ("correlated" or "uncorrelated"), and/or the specific waveform sample to be presented. The waveform type was correlated when the temporal envelopes of all of the noise bands were the same, and was uncorrelated when the temporal envelope of the band centered on the signal differed from the common envelope of the other bands. At the low-uncertainty end of the continuum of conditions, the same waveform type was presented throughout an entire block of trials, and, in addition, the same waveform sample was presented on the two observation intervals of a single trial (but changed across trials). At the high-uncertainty end of the continuum, both the waveform type and the waveform sample were chosen at random for every observation interval. Threshold estimates obtained from trials in which both observation intervals contained the same waveform type were not affected by uncertainty about the waveform sample within a trial, nor by uncertainty about the waveform type introduced across trials. Thus the comodulation masking release, or CMR (the difference in the thresholds obtained with the uncorrelated and correlated waveforms), calculated from these types of trials was robust across all of the uncertainty conditions. However, on those trials in which one correlated interval and one uncorrelated interval were paired, threshold estimates were influenced by a bias for listeners to choose the uncorrelated interval as the signal interval, whether or not it actually contained the signal. This bias reveals the importance of recognizing the contribution of the nonsignal interval in experiments involving masker uncertainty. Parallel results were obtained using the comodulation detection difference (CDD) task. In some conditions, marked individual differences were observed.     

Comodulation masking release: evidence for multiple cues

Signal detection was determined in conditions where the masker was a 10-Hz-wide noise band centered on the signal, and in conditions where either a comodulated or noncomodulated noise band (centered at 0.8 times the signal frequency) was also present. Signal frequencies of 500 or 2000 Hz were investigated. In one condition of the first experiment, the signal was exactly the same 10-Hz-wide noise band as the masker, added to the masker in phase. This condition was designed to limit the availability of cues based upon dip listening, suppression, beating, or across-frequency differences in noise envelope correlation, but to afford a cue based upon across-frequency envelope amplitude difference. The narrow-band noise signal resulted in approximately the same magnitude of comodulated masking release (CMR) as was found for a pure-tone signal. This result suggested that one important cue for CMR is an across-frequency difference in envelope amplitude. Stimulus conditions in the second experiment were intended to disrupt cues of across-frequency envelope amplitude difference, but to afford cues based upon across-frequency differences in noise envelope correlation. In this experiment, cues based upon envelope amplitude were impoverished by randomly varying the level of the flanking band from interval to interval, and by adjusting the level in the on-signal band to be the same in the nonsignal intervals as the level of noise plus signal in the signal interval. Again, substantial CMRs occurred, suggesting that another cue for CMR may be envelope pattern or correlation. The results of these experiments indicated that CMR is probably based upon more than one stimulus variable.     

The effect of across-frequency differences in masking level on spectro-temporal pattern analysis

Masking noise well separated in frequency from the signal may improve the detectability of the signal if the masking noise is modulated. This effect is referred to as co-modulation masking release (CMR). The present experiments examine the effect of across-frequency differences in masking noise level on CMR. Three experiments were performed, each using a different method to create modulated noise stimuli having across-frequency differences in the spectrum level. All stimulation was monaural. Experiment I used a notched noise method (selectively reducing the level for the critical band centered on the signal). Experiment II used a method in which the level of a 100-Hz-wide masker centered on the signal was varied, and flanking noise bands were of constant level. Experiment III used a method in which flanking noise bands were varied in level, and the 100-Hz-wide masker centered on the signal was of constant level. The signal was a 1000-Hz, 300-ms pure tone. The CMR effect was negated by small spectral notches centered on the signal (experiment I). However, CMR proved to be relatively robust to across-frequency level differences in experiments II and III (a CMR effect occurred for across-frequency differences in spectrum level as great as 20 dB). Low CMR's obtained in experiment I were probably due to relatively poor correlation of across-frequency modulation pattern which occurred with notched noise. The results of experiments II and III suggest that the fluctuation pattern is of primary importance in providing release from masking, and that information on absolute levels, coded across frequency, is of less importance.     

Lateralization of low-frequency tones and narrow bands of noise

It is well known and universally accepted that people's ability to use ongoing interaural temporal disparities conveyed via pure tones is limited to frequencies below 1600 Hz. We wish to determine if this limitation is the result of the constant amplitude and periodic axis-crossings which characterize pure tones. To this end, an acoustic pointing task was employed in which listeners varied the interaural intensitive difference of a 500-Hz narrow-band noise (the pointer) so that the position of its intracranial image matched that of a second, experimenter-controlled stimulus (the target). Targets were either pure tones or narrow bands of noise (50 or 100 Hz wide). The narrow bands of noise were delayed interaurally in two distinct manners: Either the entire waveform or only the carrier was delayed. In the latter case, the envelopes and phase-functions of the bands of noise were identical interaurally. This resulted in noises which resemble the pure tone case in that the interaural delay is manifested as a constant phase-shift and resemble ordinary noises in that the envelope and phase are random functions of time. Surprisingly, it appears that all three targets were lateralized virtually identically regardless of frequency or bandwidth. Apparently, the dynamically changing envelopes and phases did not affect the listeners' use of interaural temporal disparities in any discernible fashion.     

Differential contribution of envelope fluctuations across frequency to consonant identification in quiet

Two experiments investigated the effects of critical bandwidth and frequency region on the use of temporal envelope cues for speech. In both experiments, spectral details were reduced using vocoder processing. In experiment 1, consonant identification scores were measured in a condition for which the cutoff frequency of the envelope extractor was half the critical bandwidth (HCB) of the auditory filters centered on each analysis band. Results showed that performance is similar to those obtained in conditions for which the envelope cutoff was set to 160 Hz or above. Experiment 2 evaluated the impact of setting the cutoff frequency of the envelope extractor to values of 4, 8, and 16 Hz or to HCB in one or two contiguous bands for an eight-band vocoder. The cutoff was set to 16 Hz for all the other bands. Overall, consonant identification was not affected by removing envelope fluctuations above 4 Hz in the low- and high-frequency bands. In contrast, speech intelligibility decreased as the cutoff frequency was decreased in the midfrequency region from 16 to 4 Hz. The behavioral results were fairly consistent with a physical analysis of the stimuli, suggesting that clearly measurable envelope fluctuations cannot be attenuated without affecting speech intelligibility.     

The role of the envelope in processing iterated rippled noise.

Iterated rippled noise (IRN) is generated by a cascade of delay and add (the gain after the delay is 1.0) or delay and subtract (the gain is -1.0) operations. The delay and add/subtract operations impart a spectral ripple and a temporal regularity to the noise. The waveform fine structure is different in these two conditions, but the envelope can be extremely similar. Four experiments were used to determine conditions in which the processing of IRN stimuli might be mediated by the waveform fine structure or by the envelope. In experiments 1 and 3 listeners discriminated among three stimuli in a single-interval task: IRN stimuli generated with the delay and add operations (g = 1.0), IRN stimuli generated using the delay and subtract operations (g = -1.0), and a flat-spectrum noise stimulus. In experiment 2 the listeners were presented two IRN stimuli that differed in delay (4 vs 6 ms) and a flat-spectrum noise stimulus that was not an IRN stimulus. In experiments 1 and 2 both the envelope and waveform fine structure contained the spectral ripple and temporal regularity. In experiment 3 only the envelope had this spectral and temporal structure. In all experiments discrimination was determined as a function of high-pass filtering the stimuli, and listeners could discriminate between the two IRN stimuli up to frequency regions as high as 4000-6000 Hz. Listeners could discriminate the IRN stimuli from the flat-spectrum noise stimulus at even higher frequencies (as high as 8000 Hz), but these discriminations did not appear to depend on the pitch of the IRN stimuli. A control experiment (fourth experiment) suggests that IRN discriminations in high-frequency regions are probably not due entirely to low-frequency nonlinear distortion products. The results of the paper imply that pitch processing of IRN stimuli is based on the waveform fine structure.