Categorical perception--phenomenon or epiphenomenon: evidence from experiments in the perception of melodic musical intervals

Categorical perception was investigated in a series of experiments on the perception of melodic musical intervals (sequential frequency ratios). When procedures equivalent to those typically used in speech-perception experiments were employed, i.e., determination of identification and discrimination functions for stimuli separated by equal physical increments), musical intervals were perceived categorically by trained musicians. When a variable-step-size (adaptive) discrimination procedure was used, evidence of categorical perception (in the form of smaller interval-width DL's for ratios at identification category boundaries than for ratios within categories), although present initially, largely disappeared after subjects had reached asymptotic performance. However, equal-step-size discrimination functions obtained after observers had reached asymptotic performance in the adaptive paradigm were not substantially different from those initially obtained. The results of other experiments imply that this dependence of categorical perception on procedure may be related to differences in stimulus uncertainty between the procedures. An experiment on the perception of melodic intervals by musically untrained observers showed no evidence for the existence of "natural" categories for musical intervals.     

Subjective acceptability of various regular twelve-tone tuning systems in two-part musical fragments

Musically trained subjects rated the overall acceptability of the performance of two-part musical fragments. With the help of a computer these fragments were produced with intonations according to various regular 12-tone tuning systems: Pythagorean tuning (tempering T of the fifths equal to 0.0 cent), equal temperament (T = -2.0 cents), Silbermann (T = -3.9 cents), mean-tone (T = -5.4 cents), and Salinas tuning (T = -7.2 cents). In experiment 1, two systems in which T = 2.0 or T = -10.0 cents were also included. In both the lower and the higher parts of the fragments we used complex tones with a spectral-envelope slope of -6 dB/oct. Mean acceptability ratings were about the same for -5.4 less than or equal to T less than or equal to 0.0 cent, whereas for T greater than 0.0 and T less than -5.4 cents the ratings strongly decreased. This effect of tuning system was also found when acceptability was determined by means of the method of paired comparisons. The specific effect of tuning system was not affected by the tempo of the fragments. The perception of beats in the harmonic intervals was manipulated in experiment 2 by varying the spectral content of the tones. The condition in which the interference of the nearly coinciding harmonics was canceled resulted in higher acceptability. The effect of tuning system was the same as in experiment 1. In both experiments overall acceptability could be accurately predicted from a linear combination of the purity ratings of the isolated harmonic fifths and major thirds. It is not precluded, however, that for T greater than 0.0 and T less than -5.4 cents, the subjects based their acceptability ratings partly on the subjective purity of the melodic intervals.     

Pitch-interval discrimination and musical expertise: Is the semitone a perceptual boundary?

The ability to discriminate pitch changes (or intervals) is foundational for speech and music. In an auditory psychophysical experiment, musicians and non-musicians were tested with fixed- and roving-pitch discrimination tasks to investigate the effects of musical expertise on interval discrimination. The tasks were administered parametrically to assess performance across varying pitch distances between intervals. Both groups showed improvements in fixed-pitch interval discrimination as a function of increasing interval difference. Only musicians showed better roving-pitch interval discrimination as interval differences increased, suggesting that this task was too demanding for non-musicians. Musicians had better interval discrimination than non-musicians across most interval differences in both tasks. Interestingly, musicians exhibited improved interval discrimination starting at interval differences of 100 cents (a semitone in Western music), whereas non-musicians showed enhanced discrimination at interval differences exceeding 125 cents. Although exposure to Western music and speech may help establish a basic interval-discrimination threshold between 100 and 200 cents (intervals that occur often in Western languages and music), musical training presumably enhances auditory processing and reduces this threshold to a semitone. As musical expertise does not decrease this threshold beyond 100 cents, the semitone may represent a musical training-induced intervallic limit to acoustic processing.     

Vibrotactile intensity discrimination measured by three methods

The difference threshold for the detection of changes in vibration amplitude was measured as a function of the intensity and frequency of stimuli delivered through a 2.9-cm2 contactor to the thenar eminence. Stimuli were either 25- or 250-Hz sinusoids, narrow-band noise centered at 250 Hz, or wideband noise. Thresholds were measured by two-interval, forced-choice tracking under three methods of stimulus presentation. In the gated-pedestal method, subjects had to judge which of two 700-ms bursts of vibration separated by 100 ms was more intense. In the continuous-pedestal method, subjects had to detect a 700-ms increment in the amplitude of an ongoing pedestal of vibration. In the two-burst-continuous-pedestal method with 1500-ms pedestals, the subject had to detect which of two successively presented pedestals contained a 500-ms amplitude increment. Thresholds were consistently lower for detecting increments in the amplitude of a continuous pedestal of vibration than for detecting amplitude differences between briefly presented successive pedestals or amplitude increments in successive pedestals. A "near miss" to Weber's law was found both for sinusoidal and for noise stimuli. The difference threshold was not affected by stimulus frequency condition.     

Age effects in discrimination of repeating sequence intervals

The study measured listener sensitivity to increments in the inter-onset intervals (IOIs) of successive 20-ms 4000-Hz tone bursts in isochronous sequences. The stimulus sequences contained two-six tone bursts, separated equally by silent intervals, with tonal IOIs ranging from 25 to 100 ms. Difference limens (DLs) for increments of the tonal IOIs were measured to assess listener sensitivity to changes of sequence rate. Comparative DLs were also measured for increments of a single interval located within six-tone isochronous sequences with different tone rates. Listeners included younger normal-hearing adults and two groups of older adults with and without high-frequency sensorineural hearing loss. The results, expressed as Weber fractions (DL/IOI), revealed that discrimination improved as the sequence tone rate decreased and the number of tonal components increased. Discrimination of a single sequence interval also improved as the number of sequence components increased from two to six but only for brief intervals and fast sequence rates. Discrimination performance of the older listeners with and without hearing loss was equivalent and significantly poorer than that of the younger listeners. The discrimination results are examined and discussed within the context of multiple-look mechanisms and possible age-related differences in the sensory coding of signal onsets.     

Aging and temporal discrimination in auditory sequences.

This study examined age-related changes in temporal sensitivity to increments in the inter-onset intervals (IOI) of successive components in tonal sequences. Temporal discrimination was examined using reference stimulus patterns consisting of five 50-ms, 4000-Hz components with equal tonal IOIs selected from the range 100-600 ms. Discrimination was examined in separate conditions by measuring the relative difference limen (DL) for increments of tonal IOI in comparison sequences. In some conditions, comparison sequences featured equal increments of all tonal lOIs to examined listener sensitivity to uniform changes of sequence rate, or tempo. Other conditions measured the DL for increments of a single target IOI within otherwise uniform-rate comparison sequences. For these measurements, the single target IOI was either fixed in sequence location, or randomized in location across listening trials. Listeners in the study included four groups of young and elderly adults with and without high-frequency hearing loss. The results for all listeners showed the relative DL for rate discrimination to decrease from a maximum at the 100-ms IOI to a smaller stable value across the range of longer sequence IOI. All listeners also exhibited larger relative DLs for discrimination of single target intervals compared to rate discrimination for equivalent reference IOI values. Older listeners showed poorer performance than younger listeners in all conditions, with the largest age differences observed for discrimination of brief single intervals that were varied randomly in sequence location. None of the results revealed significant effects of hearing loss on performance of younger and older listeners.     

Frequency and intensity discrimination in humans and monkeys

Frequency and intensity DLs were compared in humans and monkeys using a repeating standard "yes-no" procedure in which subjects reported frequency increments, frequency decrements, intensity increments, or intensity decrements in an ongoing train of 1.0-kHz tone bursts. There was only one experimental condition (intensity increments) in which monkey DLs (1.5-2.0 dB) overlapped those of humans (1.0-1.8 dB). For discrimination of both increments and decrements in frequency, monkey DLs (16-33 Hz) were approximately seven times larger than those of humans (2.4-4.8 Hz), and for discrimination of intensity decrements, monkey DLs (4.4-7.0 dB) were very unstable and larger than those of humans (1.0-1.8 dB). For intensity increment discrimination, humans and monkeys also exhibited similar DLs as SL was varied. However, for frequency increment discrimination, best DLs for humans occurred at a high (50 dB) SL, whereas best DLs for monkeys occurred at a moderate (30 dB) SL. Results are discussed in terms of various neural mechanisms that might be differentially engaged by humans and monkeys in performing these tasks; for example, different amounts of temporal versus rate coding in frequency discrimination, and different mechanisms for monitoring rate decreases in intensity discrimination. The implications of these data for using monkeys as models of human speech sound discrimination are also discussed.     

Dichotic fusion of two tones one octave apart: evidence for internal octave templates

Stimuli consisting of two simultaneous and sinusoidally frequency-modulated pure tones were dichotically presented to four listeners. Two component tones of each stimulus were approximately an octave apart. They were both modulated at 2 Hz, and the frequency swing resulting from each modulation corresponded to one tenth of the carrier frequency. The listeners' task was to detect phase differences between the modulation waveforms of the two simultaneous tones: With an adaptive 2IFC procedure, just-noticeable values of phi, the phase angle of the modulation waveforms, were measured as a function of the interval formed by the carrier frequencies (one octave, i.e., 1200 cents, +/- 0, 25, 50, or 100 cents). When the carrier frequencies were not too high, just-noticeable values of phi often varied nonmonotonically with the interval, showing a minimum at or near 1200 cents. An additional experiment indicated that most, if not all, of these octave effects were not due to some form of beat detection. As a whole, the results reported here provide evidence for the existence of internal octave templates. Such templates might play an important role in the perceptual segregation of simultaneous harmonic signals, as well as in pitch perception.     

Intensity discrimination, increment detection, and magnitude estimation for 1-kHz tones

Intensity difference limens (DLs) were measured over a wide intensity range for 200-ms, 1-kHz gated tones and for 200-ms increments in continuous 1-kHz tones. Magnitude estimates also were obtained for the gated tones over a comparable intensity range. The discrimination data are in general agreement with those from earlier studies but they extend them by showing: (1) good discrimination for gated tones over at least a 115-dB dynamic range; (2) a slight increase in the relative DL (delta I/I) as intensity increases above 95 dB SPL; (3) smaller DLs for increments than for gated tones, with the difference approximately independent of intensity; (4) negligible "negative masking" when thresholds are expressed as intensity differences (delta I). For two of the three subjects, magnitude estimates do not conform to a single-exponent power law for suprathreshold intensities. Over the middle range of intensities where a single exponent is appropriate, the value of the exponent is less than 0.1 for all subjects.     

Detection and discrimination of simulated motion of auditory targets in the horizontal plane

Three experiments investigated subjects' ability to detect and discriminate the simulated horizontal motion of auditory targets in an anechoic environment. "Moving" stimuli were produced by dynamic application of stereophonic balancing algorithms to a two-loudspeaker system with a 30 degree separation. All stimuli were 500-Hz tones. In experiment 1, subjects had to discriminate a left-to-right moving stimulus from a stationary stimulus pulsed for the same duration (300 or 600 ms). For both durations, minimum audible "movement" angles ("MAMA's") were on the order of 5 degrees for stimuli presented at 0 degrees azimuth (straight ahead), and increased to greater than 30 degrees for stimuli presented at +/- 90 degrees azimuth. Experiment 2 further investigated MAMA's at 0 degrees azimuth, employing two different procedures to track threshold: holding stimulus duration constant (at 100-600 ms) while varying velocity; or holding the velocity constant (at 22 degrees-360 degrees/s) while varying duration. Results from the two procedures agreed with each other and with the MAMA's determined by Perrott and Musicant for actually moving sound sources [J. Acoust. Soc. Am. 62, 1463-1466 (1977b)]: As stimulus duration decreased below 100-150 ms, the MAMA's increased sharply from 5 degrees-20 degrees or more, indicating that there is some minimum integration time required for subjects to perform optimally in an auditory spatial resolution task. Experiment 3 determined differential "velocity" thresholds employing simulated reference velocities of 0 degrees-150 degrees/s and stimulus durations of 150-600 ms. As with experiments 1 and 2, the data are more easily summarized by considering angular distance than velocity: For a given "extent of movement" of a reference target, about 4 degrees-10 degrees additional extent is required for threshold discrimination between two "moving" targets, more or less independently of stimulus duration or reference velocity. These data suggest that for the range of simulated velocities employed in these experiments, subjects respond to spatial changes--not velocity per se--when presented with a "motion" detection or discrimination task.     

Multidimensional tactile displays: identification of vibratory intensity, frequency, and contactor area

Experiments were conducted to determine the ability of subjects to identify vibrotactile stimuli presented to the distal pad of the middle finger. The stimulus sets varied along one or more of the following dimensions: intensity of vibration, frequency of vibration, and contactor area. Identification performance was measured by information transfer. One-dimensional stimulus sets produced values in the range 1-2 bits and, for most subjects, three-dimensional sets produced values in the range 4-5 bits. Of the three dimensions considered, performance on the intensity variable was most affected, and performance on contactor area least affected, by simultaneous variations in the other dimensions.     

Intensity difference limens at high frequencies

Thresholds for amplitude modulation detection were obtained from four subjects at frequencies of 2, 4, 6, 8, and 10 kHz for sensation levels of 15, 30, 45, and 60 dB and modulation rates of 2, 4, and 8 Hz. High-frequency difference limens calculated from amplitude modulation thresholds were found to change nonmonotonically as a function of sensation level, independent of modulation rate. This nonmonotonic relation stemmed mainly from a gradual reduction of the difference limen at the lowest sensation level with increasing frequency. Difference limens for pulsed tone discrimination were also measured in two of the subjects at 2, 6, and 10 kHz and sensation levels of 15, 30, 45, and 60 dB. The relation between intensity discrimination and sensation level was similar to that found for amplitude modulation detection. These findings are interpreted as indicating that the nonmonotonic relation between sensation level and intensity resolution is a general characteristic of stimulus processing at higher frequencies.     

Memory for pitch versus memory for loudness

The decays of pitch traces and loudness traces in short-term auditory memory were compared in forced-choice discrimination experiments. The two stimuli presented on each trial were separated by a variable delay (D); they consisted of pure tones, series of resolved harmonics, or series of unresolved harmonics mixed with lowpass noise. A roving procedure was employed in order to minimize the influence of context coding. During an initial phase of each experiment, frequency and intensity discrimination thresholds [P(C) = 0.80] were measured with an adaptive staircase method while D was fixed at 0.5 s. The corresponding physical differences (in cents or dB) were then constantly presented at four values of D: 0.5, 2, 5, and 10 s. In the case of intensity discrimination, performance (d') markedly decreased when D increased from 0.5 to 2 s, but was not further reduced when D was longer. In the case of frequency discrimination, the decline of performance as a function of D was significantly less abrupt. This divergence suggests that pitch and loudness are processed in separate modules of auditory memory.     

Age-related differences in discrimination of an interval separating onsets of successive tone bursts as a function of interval duration

The study measured listener sensitivity to increments in the inter-onset interval (IOI) separating pairs of successive 20-ms 4000-Hz tone pulses. A silent interval between the tone pulses was adjusted across conditions to create reference tonal IOI values of 25-600 ms. For each condition, a duration DL for increments of the tonal IOI was measured in listeners comprised of young normal-hearing adults and two groups of older adults with and without high-frequency hearing loss. Discrimination performance of all listeners was poorest for the shorter reference IOIs, and improved to stable levels for longer reference intervals exceeding about 200 ms. Temporal sensitivity of the young listeners was significantly better than that of the elderly listeners in each condition, with the largest age-related differences observed for the shortest reference interval. Age-related differences were also observed for duration DLs measured using single 4000-Hz tone bursts set to three reference durations in the range 50-200 ms. The tone DLs of all listeners were smaller than the corresponding tone-pair IOI DLs, particularly for the shorter reference stimulus durations. There were no significant performance differences observed between the older listeners with and without hearing loss for either discrimination task.     

Formant frequency discrimination by Japanese macaques (Macaca fuscata).

These studies investigated formant frequency discrimination by Japanese macaques (Macaca fuscata) using an AX discrimination procedure and techniques of operant conditioning. Nonhuman subjects were significantly more sensitive to increments in the center frequency of either the first (F1) or second (F2) formant of single-formant complexes than to corresponding pure-tone frequency shifts. Furthermore, difference limens (DLs) for multiformant signals were not significantly different than those for single-formant stimuli. These results suggest that Japanese monkeys process formant and pure-tone frequency increments differentially and that the same mechanisms mediate formant frequency discrimination in single-formant and vowel-like complexes. The importance of two of the cues available to mediate formant frequency discrimination, changes in the phase and the amplitude spectra of the signals, was investigated by independently manipulating these two parameters. Results of the studies indicated that phase cues were not a significant feature of formant frequency discrimination by Japanese macaques. Rather, subjects attended to relative level changes in harmonics within a narrow frequency range near F1 and F2 to detect formant frequency increments. These findings are compared to human formant discrimination data and suggest that both species rely on detecting alterations in spectral shape to discriminate formant frequency shifts. Implications of the results for animal models of speech perception are discussed.     

频域分辨率和时域包络周期性对音乐音高分辨的影响

系统地分析与探讨频域分辨率及时域包络周期性对不同音色及频率覆盖范围的音乐音高分辨的影响。选择钢琴、小提琴、小号及单簧管四种乐器的乐音和特定的复合音作为测试音源。利用噪声调制的声码器模型调控音乐信号的频域分辨率和时域包络周期性。十位正常听力者参与了该项音高分辨测试。实验结果表明,随着频域分辨率的提高,受试者对音高分辨的准确率呈上升趋势,16个频带已可获得较好的音高分辨效果;当时域包络周期性信息增加时,未见其对音高分辨产生一致性积极影响。     

Sound intensity processing by the goldfish

Capacities of the goldfish for intensity discrimination were studied using classical respiratory conditioning and a staircase psychophysical procedure. Physiological studies on single saccular (auditory) nerve fibers under similar stimulus conditions helped characterize the dimensions of neural activity used in intensity discrimination. Incremental intensity difference limens (IDLs in dB) for 160-ms increments in continuous noise, 500-ms noise bursts, and 500-ms, 800-Hz tone bursts are 2 to 3 dB, are independent of overall level, and vary with signal duration according to a power function with a slope averaging - 0.33. Noise decrements are relatively poorly detected and the silent gap detection threshold is about 35 ms. The IDLs for increments and decrements in an 800-Hz continuous tone are about 0.13 dB, are independent of duration, and are level dependent. Unlike mammalian auditory nerve fibers, some goldfish saccular fibers show variation in recovery time to tonal increments and decrements, and adaptation to a zero rate. Unit responses to tone increments and decrements show rate effects generally in accord with previous observations on intracellular epsp's in goldfish saccular fibers. Neurophysiological correlates of psychophysical intensity discrimination data suggest the following: (1) noise gap detection may be based on spike rate increments which follow gap offset; (2) detection of increments and decrements in continuous tones may be determined by steep low-pass filtering in peripheral neural channels which enhance the effects of spectral "splatter" toward the lower frequencies; (3) IDLs for pulsed signals of different duration can be predicted from the slopes of rate-intensity functions and spike rate variability in individual auditory nerve fibers; and (4) at different sound pressure levels, different populations of peripheral fibers provide the information used in intensity discrimination.     

Effects of musical training and absolute pitch ability on event-related activity in response to sine tones.

The neural correlates of music perception have received relatively little scientific attention. The neural activity of listeners without musical training (N = 11), highly trained musicians (N = 14), and musicians possessing "absolute pitch" (AP) ability (N = 10) have been measured. Major differences were observed in the P3, an endogenous event-related potential (ERP), which is thought to be a neurophysiological manifestation of working memory processing. The P3 was elicited using the classical "oddball" paradigm with a sine-tone series. Subjects' musical backgrounds were evaluated with a survey questionnaire. AP ability was verified with an objective pitch identification test. The P3 amplitude, latency and wave shape were evaluated along with each subjects' performance score and musical background. The AP subjects showed a significantly smaller P3 amplitude than either the musicians or nonmusicians, which were nearly identical. The P3 latency was shortest for the AP subjects, and was longer for the nonmusicians. Performance scores were uniformly high in all three groups. It is concluded that AP subjects do indeed exhibit P3 ERPs, albeit with smaller amplitudes and shorter latencies. The differences in neural activity between the musicians and AP subjects were not due to musical training, as the AP subjects had similar musical backgrounds to the musician group. It is also concluded that persons with the AP ability may have superior auditory sensitivity at cortical levels and/or use unique neuropsychological strategies when processing tones.     

Effects of spectral resolution and temporal envelope periodicity to musical pitch discrimination

The main goal of this study was to systematically investigate the relative contribution of spectral resolution and temporal envelope periodicity to musical pitch discrimination. Stimuli from four instruments(clarinet,trumpet,piano and violin) and synthetic complex tone were utilized.A noise-excited vocoder was used to control the spectral resolution and temporal envelope periodicity of the musical stimuli.Ten normal-hearing subjects were recruited for the study.Psychoacoustic experiments on pitch discrimination were carried out.Results showed that the spectral cue was important for musical pitch discrimination.Relative good performance could be achieved when there were 16 frequency channels.No clear effect was found for the temporal envelope periodicity to pitch discrimination.     

Audibility and recognition of stop consonants in normal and hearing-impaired subjects

The purpose of this study was to examine the effect of spectral-cue audibility on the recognition of stop consonants in normal-hearing and hearing-impaired adults. Subjects identified six synthetic CV speech tokens in a closed-set response task. Each syllable differed only in the initial 40-ms consonant portion of the stimulus. In order to relate performance to spectral-cue audibility, the initial 40 ms of each CV were analyzed via FFT and the resulting spectral array was passed through a sliding-filter model of the human auditory system to account for logarithmic representation of frequency and the summation of stimulus energy within critical bands. This allowed the spectral data to be displayed in comparison to a subject's sensitivity thresholds. For normal-hearing subjects, an orderly function relating the percentage of audible stimulus to recognition performance was found, with perfect discrimination performance occurring when the bulk of the stimulus spectrum was presented at suprathreshold levels. For the hearing-impaired subjects, however, it was found in many instances that suprathreshold presentation of stop-consonant spectral cues did not yield recognition equivalent to that found for the normal-hearing subjects. These results demonstrate that while the audibility of individual stop consonants is an important factor influencing recognition performance in hearing-impaired subjects, it is not always sufficient to explain the effects of sensorineural hearing loss.