Search

Research in physics education has indicated that the traditional lecture-style class is not the most efficient way to teach introductory physical science courses at the university level. Current best teaching practices focus on creating an active-learning environment and emphasize the students' role in the learning process. Several of the recommended techniques have recently been applied to Brigham Young University's introductory acoustics course, which has been taught for more than 40 years. Adjustments have been built on a foundation of establishing student-based learning outcomes and attempting to align these objectives with assessments and course activities. Improvements have been made to nearly every aspect of the course including use of class time, assessment materials, and time the students spend out of the classroom. A description of the progress made in improving the course offers suggestions for those seeking to modernize or create a similar course at their institution. In addition, many of the principles can be similarly applied to acoustics education at other academic levels.

Piano tones have partials whose frequencies are sharp relative to harmonic values. A listening test was conducted to determine the effect of inharmonicity on pitch for piano tones in the lowest three octaves of a piano. Nine real tones from the lowest three octaves of a piano were analyzed to obtain frequencies, relative amplitudes, and decay rates of their partials. Synthetic inharmonic tones were produced from these results. Synthetic harmonic tones, each with a twelfth of a semitone increase in the fundamental, were also produced. A jury of 21 listeners matched the pitch of each synthetic inharmonic tone to one of the synthetic harmonic tones. The effect of the inharmonicity on pitch was determined from an average of the listeners' results. For the nine synthetic piano tones studied, pitch increase ranged from approximately two and a half semitones at low fundamental frequencies to an eighth of a semitone at higher fundamental frequencies.

This talk was motivated by time spent in Australia at the University of New England in Armidale, NSW, during the latter part of 1980. My wife and I and our three youngest sons spent five months there while I carried out research on the flute with Neville Fletcher and Ron Silk. As you might suspect from the title of the talk, Tom was also at the University of New England and our times there overlapped. Tom was solicitous of our sons which enriched their Australian experience. Tom's research was concerned with acoustical aspects of percussion instruments. Though discussions of our respective research projects was limited, our shared experience in Australia led to our continuing interaction during the ensuing years. A major part of the talk will consider interactions with Tom at other times and in other places.

Recent experimental developments have enabled directivity measurements of played musical instruments with high angular resolution. The measurement system assesses directional radiation characteristics while including diffraction and absorption effects of seated musicians. The results are advantageous to better understand and visualize sound produced by the instruments, provide benchmarks for physical modeling, predict and auralize sound in rehearsal and performance venues, and improve microphone placement techniques. This paper explores steady-state directivities, contrasting key differences between brass, tone-hole, and string instruments. As expected, brass instruments generate relatively predictable results because of their single radiation elements. Tone-hole instruments produce notable interference patterns due to radiation from multiple instrument openings. String instruments produce even more complex directivities because of radiation from distributed vibrating structures and instrument openings. T...

This research investigates the modeling of harmonic generation occurring in trombone sound production at loud dynamic levels. The profile of a tenor trombone was approximated as a set of contiguous cylindrical tubes. A linear model of the trombone was obtained by requiring continuity of pressure and flow between adjacent tubes. The mouthpiece and radiated pressure spectra were measured in an anechoic chamber for several pitches played in first position at several dynamic levels. The measured mouthpiece spectra were processed by the model, taking into account harmonic generation due to wave‐front steepening. Results comparing the calculated radiated spectra with the measured radiated spectra will be reported, along with an overview of the algorithm.

A simplified clarinet was modeled as a closed–open tube of uniform cross section with an open tonehole lattice near its open end. The tube had a length of 50 cm and a diameter of 1.5 cm. The tonehole lattice consisted of five identical toneholes spaced at 2.8‐cm intervals. Each tonehole was 0.8 cm in diameter and had an effective length of 1.3 cm. The open end of the tube and each of the toneholes was treated as a simple source, and a corresponding source strength (volume velocity) was calculated at frequencies of interest. The far‐field radiated pressure was calculated at a constant radius in a plane that included the tonehole lattice. The pressure was calculated at angular positions from 0 deg (open end) to 180 deg (closed end) for normal mode frequencies. (Circular symmetry about the tube axis was assumed.) Significant directional patterns were seen for frequencies above 2000 Hz. Experiments were performed on a similar structure to measure the directivity and for comparison with the calculations.

The purpose of the present study was to obtain detailed photographic sequences and lip motion data on which lip models for brass instruments may be more accurately based. The study expands upon an earlier study by Martin [J. Acoust. Soc. Am. 13, 305–307 (1942)] by using advanced fiber‐optic stroboscopy, a real instrument mouthpiece, and by studying two dynamic levels. The trombone was selected as representative of the brass family because its relatively large mouthpiece permitted the use of an optic probe. Lip motion was observed from the front and side for six notes (Bflat2, F3, Bflat3, D4, F4, Aflat4) played at loud and soft dynamic levels. The video sequences were used to obtain information on lip opening area, lip motion perpendicular to airflow, and lip motion parallel to airflow. The data are tabulated and represented in graphic form.

This presentation will review research conducted to determine the effect of inharmonic partials on the pitch of piano tones [JASA, 117, 3268–3272 (2005)]. Synthetic piano tones were created based on recordings of an upright piano. One set of tones were made to have inharmonic partials that matched the partial frequencies of the piano tone recording, whereas another set of tones consisted of harmonic partials. Listeners compared a synthetic, inharmonic piano tone to a set of synthetic, harmonic piano tone with varying fundamental frequencies. They were asked to select the harmonic tone whose pitch best matched that of the inharmonic tone. The pitch of the inharmonic tones was perceived to be sharp relative to the harmonic tones (sharp in terms of the comparison of their respective fundamental frequencies).

This paper investigates some effects of harmonicity in the resonant frequencies of a brass wind instrument. Computer methods for calculating input impedance and impulse response are first applied to several cylinder-bell combinations, and show that the bell produces a strong positive overshoot in the first reflected pulse. A trumpet with optimum harmonicity maximizes this effect, producing an impulse response somewhat similar to that of a closed-closed tube. By contrast, several inharmonic ‘trumpets’ with catenoidal and conical bells are lacking in this characteristic. The trumpet's impulse response is then convolved with a hypothetical periodic input volume velocity having a rich harmonic content, and the RMS value of the corresponding mouthpiece pressure is plotted as a function of input frequency. The resulting curve exhibits a set of distinctive, harmonic ‘playing frequencies’. A similar curve is computed for an inharmonic ‘conical trumpet’ and found to be defective.

Graduate studies in acoustics at Brigham Young University prepare students for jobs in industry, research, and academia by complementing in-depth coursework with publishable research. In the classroom, a series of five graduate‐level core courses provides students with a solid foundation in core acoustics principles and practices. The associated lab work is substantial and provides hands-on experience in diverse areas of acoustics: calibration, directivity, scattering, absorption, Doppler vibrometry, lumped-element mechanical systems, equivalent circuit modeling, arrays, filters, room acoustics measurements, active noise control, and near-field acoustical holography. In addition to coursework, graduate students complete independent research projects with faculty members. Recent thesis and dissertation topics have included active noise control, directivity of acoustic sources, room acoustics, radiation and directivity of musical instruments, energy-based acoustics, aeroacoustics, propagation modeling, nonlinear propagation, and high-amplitude noise analysis. In addition to their individual projects, graduate students often serve as peer mentors to undergraduate students on related projects and often participate in field experiments to gain additional experience. Students are expected to develop their communication skills, present their research at multiple professional meetings, and publish it in peer-reviewed acoustics journals. In the past five years, nearly all graduate students have published at least one refereed paper.

Acoustics research at Brigham Young University has a rich history extending back over half a century. The tradition of acoustics continues to this day as BYU is home to one of the largest student chapters of the ASA. The academic roots of current and former faculty will be presented. The locations of the branches (students) will also be presented to illustrate the growth and development of the program and the impact on the broader acoustics community.

Directivity measurements for live performance of musical instruments present several experimental challenges, including the need for musicians to play consistently and reproducibly. Some researchers have chosen to implement fixed, limited-element microphone arrays surrounding instruments for rough directivity assessments. Unfortunately, with practical numbers of microphones and data acquisition channels, this approach limits spatial resolution and field decomposition bandwidth. Higher-resolution data may be obtained with a given microphone and channel count by rotating a musician in sequential azimuthal angle increments under a fixed semicircular microphone array. The musician plays a selected note sequence with each increment, but corrections must be made for playing variability. This paper explores the development of this method, which also uses rotating reference frame microphones and frequency response function measurements. The initial developments involve a loudspeaker, with known directivity, to simulate a live musician. It radiates both idealized signals and anechoic recordings of musical instruments with random variations in amplitude. The presentation will discuss how one can reconstruct correct source directivities from such signals and the importance of reference microphone placement when using frequency response functions. It will also introduce the concept of coherence maps as tools to establish directivity confidence.

The presentation will review the computational simulation of two clarinet-like models. The Stewart and Strong model (J. Acoust. Soc. Am. 68, July 1980) consisted of a uniform cylindrical tube with an attached clarinet mouthpiece and reed. The tube was represented by 0.25-cm sections of lumped element transmission line. The tapered mouthpiece was similarly represented with 0.1-cm sections. The reed was represented as a non-uniform bar clamped at one end. The simulation was carried out on a DEC PDP-15 computer and had a running time of roughly 250,000 times real time. The Sommerfeldt and Strong model (J. Acoust. Soc. Am. 83, May 1988) was similar to the Stewart model, but included seven toneholes. The impedance of the tube with mouthpiece was calculated and used to calculate the impulse response. This was used to determine the interaction between pressure and flow at the reed opening. A player's airway was also included in the model because, at the time, there was some interest in its effect on tone product...

Neville Fletcher's wide ranging interests in excitation mechanisms and sound production in musical instruments and other sound sources have resulted in many related reports and publications. The presenter had the good fortune to work with Fletcher while holding a Senior Fulbright Fellowship at the University of New England. The joint research on acoustical aspects of the flute resulted in two papers: “Acoustical characterization of flute head joints” (Fletcher, Strong, and Silk, JASA 71, 1255-1260) and “Numerical calculation of flute impedances and standing waves” (Strong, Fletcher, and Silk, JASA 77, 2166-2172). These two papers will be reviewed in the presentation.

Pure sound piano wire is a stainless‐steel wire that has only recently entered the U.S. market. Because of its different composition, which makes the wire more malleable, it should have a considerably lower amount of inharmonicity compared to regular steel wire. Measurements were conducted on several pianos with regular piano wire (Rosslau and Mapes) and pure sound in order to assess their differences in inharmonicity and tonal qualities. The pure sound wire produced measurably less inharmonicity than regular wire on the same kind of piano. However, this difference is much smaller compared to the difference between small and large pianos. This presentation will explain the methods of measurement and analysis, as well as simulations that were implemented in order to analyze the effect that inharmonicity has on the temperament, octave stretching, and interval patterns. [The Brigham Young University Department of Physics is acknowledged for the funding it provided for this research.]

Wind instrument research at Brigham Young University during the 1970s, 1980s, and 1990s will be reviewed. In more or less chronological order, the research was as follows: (1) Calculation of input impedances of an oboe; (2) predictor coefficient analysis/synthesis of French horn tones; (3) functional model of a simplified clarinet; (4) calculated and measured modal frequencies of brass instruments; (5) calculation and measurement of flute impedances and standing waves; (6) calculation and measurement of horn impedances; (7) simulation of a player‐clarinet system; (8) optimization of clarinet toneholes; (9) simulation of a player‐trumpet system; (10) stroboscopic measurement of the lip motions of a trombone player; (11) calculation and measurement of nonlinear effects in loud trombone tones; (12) finite element modeling of the lips of a trombone player. [Work supported by the Institute for Research in Musical Acoustics.]

Following his long and illustrious career at Bell Laboratories, Harvey Fletcher retired to Provo, UT and continued his career at Brigham Young University. Building on his research at Bell Labs, Fletcher designed and constructed a large anechoic chamber at BYU and also pursued numerous research projects with students at the university. These projects often required the development of research equipment that was very innovative in its time, much of which is still in existence today. This paper will summarize some of the research carried out during those years, and the measurement equipment and apparatus developed by Fletcher and his students to obtain the desired measurements.

A brief background will be given of ‘‘Descriptive Acoustics of Music and Speech,’’ an introductory course taught at Brigham Young University for the past 40 years. Several conceptual and physical models used in the course will be described. Some phenomena observed in music and speech will be explained in terms of the models. Diagrams, sound samples, and video clips will be used to illustrate the phenomena and their explanation.

A geometrical display of speech spectra intended as an adjunct to lipreading was developed. Spectra were calculated at 5‐ms intervals from speech sound pairs ambiguous to lipreaders. The spectra were displayed as sequences of irregular decagons. Human subjects were asked to discriminate between pairs of spectral decagon sequences derived from pairs of ambiguous speech sounds. Subjects were able to discriminate between most of the visual spectral patterns derived from ambiguous sounds. However, spectral patterns associated with the voiced/unvoiced contrast in some stop pairs were not discriminated consistently.

Many syllables are ambiguous to lip readers because of their similar appearance on the lips. An aid is necessary to make distinction of these syllables possible. In a previous study [E. J. Hunter and W. J. Strong, J. Acoust. Soc. Am. 102, 3166–3167 (1997)], speech spectra were processed at 5‐ms intervals and displayed as sequences of irregular decagons. Subjects were asked to discriminate between pairs of decagons representing ambiguous syllables. Subjects successfully discriminated most pairs, but were inconsistent in discrimination of voiced‐/unvoiced‐stop pairs. This study is concerned with display formats aimed at improving discrimination of ambiguous pairs to develop a more effective aid for lip readers. In particular, ‘‘pie slice’’ and ray formats are evaluated against polygons. Filled shapes are evaluated against open shapes. Five‐band speech spectra are evaluated against ten‐band spectra. Display color is evaluated against line thickness for representing overall sound level.

Some numerical calculations of impedances of an oboe showed relatively good qualitative agreement with experimental data [G. R. Plitnik and W. J. Strong, J. Acoust. Soc. Am. 65, 816–825 (1979)]. With the apparent success of the numerical method, input impedances were calculated for a number of ‘‘woodwind‐like’’ structures. On the basis of these calculations a paper was prepared and submitted for publication. Some six months after its submission it was returned with numerous comments regarding errors in its content. A recurring comment from the reviewer was ‘‘the digital computer where used alone has contributed directly to most of the difficulties of this paper.’’ Previous to receiving the reviewer’s comments the author was using the same numerical method to calculate resonance frequencies of a flute. Difficulty was being experienced in getting reasonable agreement with experimental results. However, after incorporating empirical data into the representation of toneholes rather good results were obtained ...

The creation of a physical model is discussed in the context of the lips of a player/trombone system. The lip model is designed to incorporate the salient features of the physical structure of the lips, especially the geometry and material properties. A finite‐element model is developed to approximate this structure, while the aerodynamic system is represented by flow equations solved with a finite difference scheme. To drive this model three pressures are applied. The mouth pressure is generated as a constant lung pressure delivered across an airway resistance; the pressure between the lips is dependent on the stream velocity, and the pressure in the mouthpiece is given by the convolution of the volume velocity with the impulse response of the trombone. Together these drive the oscillation of the finite element lip structure, which may be compared with that of previous models and experimental studies.

The strings on a piano have stiffness which gives rise to inharmonic partials that are stretched in frequency relative to harmonic values. In addition, most notes on a piano are strung with more than one string. Even though the multiple strings of a note are tuned to the same nominal frequency, they are never precisely in tune. When the multiple strings are sounded together the resulting tone has partials that increase and decrease in amplitude which results from their in‐plane and out‐of‐phase action on the bridge. In order to illustrate these phenomena, pianolike tones for the note C4 were synthesized in which the parameters of 15 partials were specified for each of three strings. Each of the 45 partials was represented with a decaying sinusoid whose frequency, initial amplitude, and decay rate were specified. The amount of inharmonicity and out‐of‐tuneness of the strings were varied over a considerable range in order to make their effects apparent. Audio samples were recorded for various combinations of partial parameters.

A series of studies of eight acted emotions compared acoustical measurements, respondents’ identifications, and vocal ratings. Predictions were made as to which of these three approaches would be most accurate in differentiating between emotions within a pair. Emotion pairs were classified as logically similar and/or perceptibly similar. Logically similar emotion pairs (such as sadness/depression or anger/hate) were better distinguished by vocal/acoustic analysis. Perceptibly similar but logically dissimilar emotions (such as anger, fear, and joy) were better distinguished by respondents’ identifications. Multivariate statistics were used to compare the respondents’ identification space, and vocal rating space, and the acoustical measurement space.

An extensive set of carefully recorded utterances provided a speech database for investigating acoustic correlates among eight emotional states. Four actors and four actresses simulated the emotional states of anger, joy, neutral, nervousness, hate, fear, sadness, and depression. Many acoustic parameters were extracted from analyses of the 64 (8×8) emotion portrayals. A reduced set of acoustic parameters was obtained by eliminating some highly correlated parameters. The talkers realized ‘‘unique’’ emotions, such as anger, with consistent values of the parameters. Different talkers realized ‘‘ambiguous’’ emotions, such as neutral and nervousness, with different values of the parameters. Emotion ‘‘pairs,’’ such as sadness and depression, had similar parameter values. The acoustic parameters tended to correlate strongly with ‘‘unique’’ emotions but less well with ‘‘ambiguous’’ emotions.

The method of images was used to calculate impulse responses for a large (30 m×23 m×18 m) three‐dimensional space, a large (30 m×23 m) two‐dimensional space, and a large (30 m) one‐dimensional space. The reverberation time for these spaces was set to approximately 1.5 s and all sound absorption took place at the walls. Similar impulse responses were calculated for small rooms whose dimensions were one‐tenth those of the large rooms and whose reverberation times were smaller. A short (10‐s) sample of singing was recorded in an anechoic chamber and then convolved with the various impulse responses. The resulting six reverberated samples of singing were recorded binaurally on audio tape. Graphs of the impulse responses and the taped examples will be presented. (Although the taped examples will be presented via loudspeakers, they are best heard via earphones.)

Although the bulk of his work was concerned with various aspects of speech and hearing in communication, Harvey Fletcher had a longtime interest in musical acoustics. This interest is evidenced in his papers on perceptual aspects of musical tones, which stressed the complex interdependences of the percepts of loudness, pitch, and timbre on the physical parameters of intensity, frequency, and spectrum. However, his active involvement in musical acoustics research came to fruition only after he was released from other research, administrative, and teaching responsibilities. In conjunction with others, he explored perceptual aspects of piano, organ, violin, and percussion tones. His approach was that of analysis of natural tones, followed by synthesis of these tones based on the analysis results. Real and synthetic tones were presented to listeners who were asked to judge the tones as being real or synthetic. When listeners were unable to distinguish between the two, the synthetic tones were considered to co...

More than 50 years ago, Martin investigated lip vibrations in a coronet mouthpiece using stroboscopic photography [D. Martin, J. Acoust. Soc. Am. 13, 305–307 (1942)]. Since then, several researchers have based lip models on Martin’s data. Unfortunately, due to the quality of the photographs, it is difficult to obtain anything more than a limited quantitative description of the lip motion. The purpose of this study is to obtain more detailed photographic sequences and lip motion data on which new models may be based. The trombone was selected as representative of the lip reed family. A computer‐controlled fiber optic stroboscope was used to capture the motion of a player’s lips on video. By inserting the optic bundle through small holes drilled in the mouthpiece, lip motion was observed from the front and side for six notes (Bflat2, F3, Bflat3, D4, F4, G4) played at loud and soft dynamic levels. Video sequences and resulting lip motion data will be presented and discussed.

Transmission of sound in a large room was simulated by means of an image method incorporating artificial diffusion. The simulation model was validated by comparing the standard deviations of its transmission spectra to theoretical and experimental results obtained by others. Several combinations of source and receiver positions were simulated in order to assess their effects on spectral variability. A single source and single receiver arrangement resulted in the greatest variability as expected. Other combinations, including single‐source/two‐receiver (to represent two ears), two‐source (to represent spatial extent of source)/two‐receiver, and five‐source/five‐receiver, were used to assess their effects. The effects of frequency modulation and critical bands of the ear were also simulated. In an attempt to relate simulation results to stability in perception, a limit of 6‐dB difference in relative harmonic levels between received tones and source tones was chosen as the criterion. On the basis of this criterion, a single‐source and two‐receiver arrangement incorporation critical band effects and frequency modulation was found adequate to provide stable perception.

Speech having a high signal‐to‐noise ratio and digitized at 16 kHz was used for the comparison. An ‘‘ideal’’ synchronous analysis was obtained by handmarking the speech events—voicing periods, plosive bursts, and so on. An asynchronous analysis was obtained from 25‐ms Hamming windowed segments of speech stepped at 7.5‐ms intervals. In both cases the speech sample—event or windowed segment—was Fourier transformed and displayed in a ‘‘trispectral’’ space. The synchronous analysis produced erratic event‐to‐event spectra for short voicing periods. The asynchronous analysis produced erratic segment‐to‐segment spectra for long voicing periods. It also tended to smear short‐duration, transitional features. Various examples of trispectral plots obtained from the two methods will be shown for comparison.

A resonator is something that responds to and enhances some frequencies with respect to other frequencies. In the simplest case of voiced speech production, an air‐filled tract acts as a resonator that selectively enhances certain frequencies of the periodic glottal airflow as it passes from one end to the other end of the tract. As a simple example for demonstrating the effect of a resonator, the tract is represented by a cylindrical tube closed at the ‘‘glottal’’ end and open at the ‘‘mouth’’ end. The resonance frequencies of the tube−those frequencies of the airflow that will be enhanced as the air passes through the tube−can be determined easily from the tube’s length and end conditions. The resonance frequencies−formant frequencies in the terminology of speech science−are verified by clapping a cylindrical tube to produce a tube spectrum. The tube is then driven with an artificial larynx as a source to produce a ‘‘voiced’’ sound. The spectrum of the sound is seen to be the sum of the source spectrum and the tube spectrum (when expressed in terms of sound levels). The production of ‘‘vowel’’ sounds by this method will be demonstrated.

In a previous meeting, the authors presented an algorithm for extracting microevents (e.g., pitch periods, explosive bursts, packets of energy during frication) from continuous speech [D. A. Berry and W. J. Strong, J. Acoust. Soc. Am. 89, 1893 (A) (1991)]. After initially characterizing microevents by their one‐third octave spectra, two methods were studied for plotting microevents in reduced, two‐dimensional spectral spaces. The first method involved principal components in which the first two components captured approximately 60% of the spectral variance. However, microevents clustered into relatively small regions of the space and the resulting plots were often counterintuitive and difficult to interpret. The second method was a ‘‘trispectral’’ approach adapted from previous tristimulus studies (H. F. Pollard and E. V. Jansson, Acustica 51, 162–171). Low‐, mid‐, and high‐frequency one‐third octave filters were grouped so as to obtain a fairly uniform distribution of microevents throughout the trispectral space. The resulting plots were intuitively easy to interpret, with the distribution of vowel microevents being analogous to typical F1‐F2 plots. Moreover, because the trispectral approach was based on filter outputs rather than formants, it permitted voiced and unvoiced events to be plotted on the same graph. Microevent distributions and trajectories from continuous speech will be presented in both spectral spaces.

An event‐synchronous technique has been designed in an attempt to optimize time and frequency resolution in speech analysis. The technique isolates “microevents” in the speech waveform and then analyzes them, thus differing from commonly used asynchronous methods that employ a fixed frame length stepped forward in constant time increments. A microevent (ME) is associated with a “packet of energy” in the waveform and is initiated by some underlying input or fluctuation of energy. There are four basic types of MEs: (1) a voiced ME is initiated by a pitch pulse; (2) a plosive ME is initiated by a plosive burst; (3) a noise ME is initiated by a positive fluctuation in energy; and (4) a mixture ME. An ME is terminated at the initiation of the next ME or when the energy of the speech signal falls below the background level. ME durations are constrained to lie within a range of 2–20 ms. The current algorithm, developed and tested with portions of the 1988 DARPA TIMIT acoustic‐phonetic continuous speech database,...

Intonation and stability of clarinet tones are influenced by the resonance frequencies of the instrument that are influenced in turn by the placement and size of its tone holes. Dimensions of an instrument of moderate quality were measured and served as the starting point for the optimization procedure. Input impedances were calculated for 47 different fingerings of the instrument. Minimization of the frequency differences between four resonance frequencies—modes 1 and 2 in the chalumeau register, mode 2 in the clarion register, and mode 3 in the altissimo register—and four reference frequencies served as the optimization criteria. The reference frequencies were arbitrarily chosen from an equal tempered chromatic scale tuned to A4 = 440 Hz. Tone‐hole position, diameter, and height were the optimization parameters. In one method, these parameters were modified one at a time, while in a second method all three were modified together. Both methods produced similar results with average reductions in “frequency error” of 9 cents for chalumeau mode 1, 28 cents for chalumeau mode 2, and 16 cents for modes 2 and 3 of the combined clarion and altissimo registers. Optimization procedures and results will be discussed.

Some time ago, John Backus developed an experiment to investigate harmonic generation in trumpet tones [J. Backus and T. C. Hundley, J. Acoust. Soc. Am. 49, 509–519 (1971)]. Incorporating Martin's conclusion that a player's lip opening varies almost sinusoidally [D. W. Martin, J. Acoust. Soc. Am. 13, 305–308 (1942)], he constructed a rectangular slit whose opening could be controlled by a sinusoidal driver. Attaching the slit to the mouthpiece of a trumpet and applying a constant blowing pressure, he then measured and analyzed the mouthpiece pressure as a function of time. The purpose of this paper is to report a computer simulation of that experiment, in which different parameters of the system are varied to assess their relative importance in the generation of tones. The effects of varying the slit dimensions, blowing pressure, and amount of closure occurring in each cycle will be examined, and curves will be shown to illustrate the slit opening, mouthpiece pressure, and input volume velocity as functio...

In an effort to generate realistic trumpet mouthpiece pressure waveforms, a player‐trumpet model was created that consisted of the player's lungs, the player's airway, the player's upper lip, and the trumpet. The player's lungs were modeled as a constant pressure source. The player's airway (from lungs to lips) was modeled as a resistance of 20 acoustic ω (cgs). The player's lip was modeled as a swinging‐sliding door having 1 deg of rotational freedom and one degree of translational freedom. The trumpet was modeled in terms of its pressure response to a volume‐flow impulse. Equations were solved iteratively to give (1) lip opening as a dynamic function of the forces and torques acting on the lip, (2) volume flow as a function of lip opening and the pressure difference across the lips, and (3) mouthpiece pressure as a convolution of volume flow with trumpet impulse response. Waveforms were calculated and displayed for the lip‐opening area, the volume flow through the lips, and the pressure in the mouthpiece. The lip‐opening area and mouthpiece pressure waveforms were realistic when compared to experimental waveforms in the literature. However, there are still too few experimental data to constrain the model in terms of lip closure time and total air flow on sustained notes. Simulation data will be presented and compared with experimental data. [The work to be reported was carried out while the author was on leave at IRCAM in Paris. Rene Causse supplied player and trumpet data and Duanne Dudley supplied trumpet impulse data.]

Two preliminary studies comparing the intelligibilities of predictor-coefficient versus formant-frequency-coded speech and the intelligibilities of predictor-coefficient-coded speech using different numbers of coefficients are reported.

The resonant frequencies of a sinusoidally driven system are usually defined as those for which some physical response is a relative maximum. There is also a tendency to define them as frequencies for which the reactance of the system vanishes, since zero‐reactance frequencies are often approximately equal to maximum‐response frequencies and are sometimes easier to calculate. However, there are many systems for which the two types of frequencies are significantly different. There are also systems for which the reactance does not vanish in certain frequency ranges, though maximum responses still occur. It is concluded that vanishing reactance is not valid as a general criterion for resonance, and students should be warned against its use.

The purpose of this study was to investigate a numerical method for calculating impedances and standing wave patterns of flute structures. To this end, the physical dimensions of flute joints and tone holes were used to compute impedance and standing waves as a function of frequency for several different fingerings. Numerically computed resonance frequencies for head joint, middle and foot joints, and complete flute are compared to experimentally measured values. Computed pressure standing wave patterns for two fingerings of A6 are compared to experimental values. Reasons for the observed discrepancies between the predicted and experimental frequencies are discussed.

The human vocal cord is treated as an elastic material capable of propagating compressional, shear, and surface waves. A mathematical formulation for commonly observed vibrational modes is developed. This includes horizontal as well as vertical tissue displacements. Current vocal cord models are discussed and evaluated in terms of their ability to describe adequately the vibrational degrees of freedom of the cords.Subject Classification: 70.20.

A model has been developed which is designed to preserve some of the naturalness that is usually lost in speech synthesis. A parametrized function is used to produce an approximation to the cross‐sectional area through the glottis. A circuit model of the subglottal and glottal system is used with the supraglottal pressure to generate the glottal volume velocity. The tract used to obtain the supraglottal pressure is represented by its input‐impedance inpulse response, which can be calculated from the area function of the tract. A convolution of the input‐impedance impulse response with the volume velocity determines the supraglottal pressure. The two coupled equations for the volume velocity are solved simultaneously. The output of the model is generated by convolving the resulting glottal volume velocity with the transfer‐function impulse response of the tract. This technique preserves the interaction between the glottal flow and the vocal tract, which is usually lost. Comparisons are made between ‘‘complete tract loading’’ and ‘‘inductive tract loading.’’ Magnitude spectra of the various pressures and the glottal volume velocity are examined in detail. Effects of varying the glottal parameters are examined for one vowel. Listening tests showed that vowels systhesized with the interaction were preferred as more natural sounding than those without the interactions.

Using the voices of six subjects, representing various social and educational backgrounds, fifty-four synthetic voices were generated by computer. Each normal voice was both increased and decreased in rate by 121/2, 25, 371/2, and 50 per cent. Judges evaluated the fifty-four voices using a series of adjectives representing two major personality factors of " competence " and " benevolence ". Several statistical analyses were performed, and it was found that the competence factor was much more sensitive to rate manipulations than was the benevolence factor. Ratings of competence were found to increase as rate increases and decrease as rate decreases, in a linear fashion. Benevolence had an inverted U-relationship with speech rate; the highest benevolence ratings occurred with normal speech rate.

The intelligibility of unprocessed and of low‐pass filtered speech was compared to that of speech processed using three versions of an all‐harmonic code consisting of many harmonic sinusoids, a largest harmonic code consisting of four harmonic sinusoids closest to the formants, and a formant code consisting of three sinusoids scaled from the formant frequencies. Fundamental frequency and formant frequencies were scaled down in frequency by different amounts in the various codes. Normal‐hearing subjects were tested on three different types of tests. The Diagnostic Rhyme Test (DRT) was used on the two speech varieties and on codes that were not frequency lowered, a Diagnostic Discrimination Test (DDT) was used on frequency‐lowered speech codes, and a prosodic test was run on all versions of the speech and speech codes. Results of each test are presented and compared for the various talker, speech, and speech‐code combination; they show that the low‐pass‐filtered speech was always more intelligible than any low‐frequency speech code tested.