Index of content:
Volume 71, Issue S1, April 1982
- PROGRAM OF THE 103RD MEETING OF THE ACOUSTICAL SOCIETY OF AMERICA
- Session A. Physiological Acoustics I: Hearing Development
- Invited Papers
71(1982); http://dx.doi.org/10.1121/1.2019266View Description Hide Description
The developmental changes in interaural distance, pinna size, and ear‐canal dimensions will be noted in a variety of species. The contribution of these changes to sounddiffraction about the head interaural distance and ear canal resonance will then be explored. We will suggest that the acoustic properties of the neonatal outer ear exert an influence on sound input to the TM at frequencies substantially higher than that found in the adult. Consequently, the outer‐ear system in the very young neonate may contribute only minimally to hearing ability. The late stages of development in the middle‐ear conductive apparatus show wide variations among species, but within a given species there is remarkable synchrony in the maturation of the ossicles, TM, middle‐ear muscles, and the tympanic cavity. Evidence of functional middle‐ear development will be summarized from human, hamster, and chick. We will consider the idea that middle‐ear maturation plays a significant role in determining many of the patterns of development seen in more central regions of the auditory system. [Work supported by the Deafness Research Foundation.]
71(1982); http://dx.doi.org/10.1121/1.2019267View Description Hide Description
The organ of Corti of the fetal mouse in culture provides a unique opportunity to study the development of the afferent innervation of the hair cells. Cultures were explanted between 13 to 18 days of gestation and maintained up to 14 days. The leading finding is that all the major fiber systems that are present in the intact animal develop also in culture. Formation of radial patterns in the spiral ganglion follows (1) the disappearance of the subepithelial plexus and a pronounced reduction in the total number of peripheral fibers, and (2) rearrangement of spiral ganglion cells into distinct groups that give rise to individual radial bundles. Development of the innervation of inner hair cells begins with an arrest of the growth of radial fibers around the inner hair cells,formation of neurofibrillar caps, and formation of the inner spiral plexus and bundle. Outer spiral fibers penetrate the outer hair cell region as the “deep radial fibers” of Tello. This is followed in succession by (a) a stationary period during which growth is suspended, (b) positioning of the growing tips toward the base, (c) a period of resumed growth in a spiral direction, and (d) rearrangement of fibers into three outer spiral bundles and formation of neurofibrillar caps upon the outer hair cells. Our findings show that orderly innervation patterns, indistinguishable from those observed in the intact animal, are produced in culture. Thus the development of innervation in the organ of Corti of the mouse appears to be determined by local factors. [Work supported by NSH grants NS15061, NS12732, and NS2P3‐03352.]
71(1982); http://dx.doi.org/10.1121/1.2019319View Description Hide Description
The development of synapses clearly differentiates the two kinds of cochlear receptors: Inner (IHC) and outer (OHC) hair cells. An adultlike pattern of innervation can be seen very early below the IHCs: Both afferent and efferent (mainly axodendritic) synapses are formed before the onset of function. At the OHC level, drastic changes occur: Numerous afferent dendrites retract as the large efferent endings arrive and synapse with the hair cell. This synaptic OHC remodeling coincides with maturation of compound action potential (AP) parameters, such as the “S” shape of the input‐output curves and the sharpening of AP tuning curves. Two major points can be emphasized from these findings. First, ontogeny brings other evidence for two separate efferent systems. The one maturing first is mainly connected with afferent dendrites issuing from the IHCs and is probably responsible for the inhibitory effect on the AP, which can be observed as soon as the first AP can be recorded. The other efferent system matures slowly and synapses mainly with the OHCs. This leads to the second point about the timing coincidence which tends to link the late development of the OHCs and their efferent synapses with the maturation of discriminative physiological properties of the cochlea.
Morphogenesis of the cochlea and cochlear nucleus: Studies of normal and experimentally altered avion embryos71(1982); http://dx.doi.org/10.1121/1.2019320View Description Hide Description
Cochlear ganglion cells form foot‐shaped synaptic endings on hair cells and large axonal end bulbs on adendritic neurons in nuc. magnocellularis (cochlear nuc.). These two distinctive types of endings, despite their separate locations, develop synchronously and their morphogenesis correlates with structural changes in their respective target cells. For example, immature synapses appear simultaneously in the cochlea and nuc. magnocellularis, contemporaneous with the differentiation of hair cells. The differentiation of end bulbs parallels the transformative loss of dendrites from their post‐synaptic neurons [Jhaveri and Morest, Neurosci. (in press, 1982)]. Hair cells, ganglion cells, and magnocellular neurons may influence one another trophically to greater or lesser degrees. The survival of cochlear ganglion cells,in vitro, is depressed by excluding the receptor epithelium [Ard and Morest, Neurosci. Abstr. 7, 168 (1981)]. In nuc. magnocellularis fewer and smaller neurons survive when deprived of cochlear axons. Remarkably, however, the remaining magnocellular neurons, although contacted by abnormal boutonal endings, exhibit several normal features. Similarly, hair cells differentiate after ganglion cells are removed; in other experiments, normal synapses form between hair cells and ganglion cells even when the latter cells lack a central connection. These studies suggest that for auditory cells, some developmental events result from trophic influences while certain morphological processes may occur autonomously. [Supported by NIH grants NS 14354 and NS 16993.]
71(1982); http://dx.doi.org/10.1121/1.2019321View Description Hide Description
The auditory system continues to develop well into postnatal life. In the cat, the discharge properties of neurons in the cochlear nuclei, the first synaptic stations in the auditory pathway, undergo marked changes during the first few postnatal weeks when these structures, as well as those peripheral to them, are completing their morphologic development. The structural changes in these parts of the auditory pathway are accompanied by changes in neuronal threshold, firing rate, frequency selectivity, dynamic range, and in the timing of the unit discharge. Many of these changes reflect the development of structures peripheral to the cochlear nuclei, including the middle ear, cochlea, and auditory nerve. Others are likely the result of maturation of neuronal circuits within the cochlear nuclei. Not all functional properties recorded in a given neuron reach maturity at the same time. Those functional properties that emerge prior to the onset of normal auditory sensitivity would appear to be largely under genetic control. Other functional properties appear later in life and, therefore, their development may come under the influence of the acoustic environment. [NIH Grant HDO3353.]
71(1982); http://dx.doi.org/10.1121/1.2019322View Description Hide Description
There exist certain features in the electrophysiological development of auditory responses that behave consistently across mammalian species. Some of these are the decline of response latency with age, the paradoxical maturation of frequency sensitivity, the rapid maturation of response threshold, and the growth of response amplitude during the neonatal and early postnatal period. Our work has focused on the development of the auditory brainstem response (ABR) in unanesthetized kittens. Using free‐field click stimuli, we have discerned two developmental stages. The first or early stage begins at about 10 days postpartum, coincident with the appearance of ABRs at high intensities. It is characterized by ABR wave latencies that decline rapidly and ABR wave amplitudes that increase rapidly with age. These behave more‐or‐less linearly. The second or later stage begins at about 20 days of age. ABR wave latencies mature less rapidly during this stage, reaching asymptotic values at about 60 days of age for wave V. These behave exponentially. ABR waves originating from higher auditory brainstem nuclei mature more slowly than waves originating from more peripheral nuclei. ABR responses to tone pips also develop in two stages, with development of low‐frequency sensitivity lagging behind development of middle‐ and high‐frequency sensitivity. In kittens, cortical‐evoked responses develop in parallel with ABR responses, but their measurement is considerably less reliable. When the relative rates of maturation are taken into account, it is seen that the same sequence of events occurs in human infants, except that the early stage of ABR development happens prenatally. [Work supported by grant NS‐14880 from NINCDS and by a biomedical research supplement grant from NIH.]
71(1982); http://dx.doi.org/10.1121/1.2019374View Description Hide Description
The purpose of this paper is to review the evoked potential correlates of human auditory development with special emphasis on the brainstem auditory evoked response (BAER). A number of developmental dependencies have been identified in the BAER, including threshold changes, latency changes, absolute and relative amplitude changes, sensitivity to varying rate of stimulation, and age dependent topography. These age dependent phenomena will be reviewed and new data on developmental aspects of adaptation, broadband and narrow‐band masking, and frequency specificity will be discussed. Primary emphasis will be placed on the definition of the source of the age dependent changes in the absolute latency of wave V, including the relative contribution of mechanical and neural factors to the observed changes. The picture emerging from these observations is that the determinants of auditory development, as reflected in the auditory evoked response, are multifactorial. Each of these factors has its own time constant and rarely acts independently of the other factors.
71(1982); http://dx.doi.org/10.1121/1.2019375View Description Hide Description
Two structure‐function relationships which influence the development of hearing will be discussed. First the evidence of a paradoxical relationship between the ontogeny of responsiveness as a function of frequency and developmental gradients seen in the inner ear and central nervous system will be considered. To account for this paradox, I will propose that the mechanical tuning properties of the basilar membrane are systematically changing during development. Two lines of indirect evidence, one plotting developmental changes in the position of damage produced by high‐intensity pure tones and the other showing ontogenetic changes in the tonotopic map in the central nervous system, support the proposed interpretation. Second, the influence of afferent interactions on the development of hearing and of neuronal structure in the arian brainstem auditory pathways will be examined. Data will be presented to show that afferent integrity and activity profoundly influence the normal development of structure and function. [Work supported by NIH RCDA ♯00305‐04 and grants NS15478 and NS15395.]
- Session B. Architectural Acoustics I and Noise I: Computer Modeling Simulation of Room Acoustics
71(1982); http://dx.doi.org/10.1121/1.2019422View Description Hide Description
The computer program, giving a simultaneous picture of the space‐, time‐, and directional distribution of energy on any room surface, has been used as a design tool both in room acoustical planning and in designing sound reinforcement systems. The newest extension of the program and design cases are presented.
71(1982); http://dx.doi.org/10.1121/1.2019423View Description Hide Description
The practical calculation of the first reflections from elements comprising the boundaries of an auditorium (such as rigid plane panels, rigid curved panels, and panels having reflection coefficients other than unity) will be reviewed. Measuredreflection coefficients of porous layers and auditorium seats at grazing incidence will be shown, concentrating on the time domain. Next, multiple reflections between two plane panels, two rigid convex panels, two plane panels with reflection coefficients different from unity, and those inside a rigid concave panel are calculated, based on the calculation of the first reflections mentioned above. These results are in good agreement with those measured. Finally, the calculated pulse response of the early reflections in a scale model auditorium, which is the convolution of the calculated impulse response and a direct soundgenerated by a point source, will be compared with the measured one. Comparison will also be made of their frequency characteristics.
71(1982); http://dx.doi.org/10.1121/1.2019424View Description Hide Description
Speech quality, a subjective measure of overall excellence determined by preference testing, was originally developed for evaluating speech system designs. The application of speech quality measurements to the evaluation of auditorium acoustics design required two models: a real‐time multichannel electroacoustic model to simulate physical sound fields, and a linear/scalar products mathematical model to analyze subjective preferences (together with an analysis of variance method for determining statistical significance and scalability levels). In the present experiments, speech quality is shown to be either multidimensional or unidimensional in configuration, and valid at either interval or ordinal levels of measurement. Four consensus dimensions (reflections ratio; excessive reverberation; discrete echo disturbance; naturalness) and three individual differences dimensions (reverberance; spatial coloration; evenness of echogram) were identified. The objective index, signal‐to‐noise ratio, gave a high correlation with speech quality for simulated sound fields without excessive background noise. [Work supported by the SERC, United Kingdom.]
71(1982); http://dx.doi.org/10.1121/1.2019426View Description Hide Description
This paper presents a method of calculating the subjective preference of sound fields in concert halls before their construction. Subjective preference judgments (paired comparison tests) were systematically performed in relation to the fully independent objective parameters of the acoustic information which is included in the signals to both ears. The sound fields with various combinations of listening level, early discrete reflections, subsequent reverberation time, and magnitude of the interaural cross correlation were simulated with the aid of a digital computer. The optimal conditions maximizing the subjective preference could be found for each objective parameter, because the parameter affected almost independently the subjective preference judgments. Based on the linear scale value which is obtained by applying the law of comparative judgment, we can calculate a total preference value according to the “principle of superposition.” Examples of the calculation of preference values using the plan and the cross section of a concert hall are demonstrated. [Work supported by Scientific Foundation of Ministry of Education, Japan.]
71(1982); http://dx.doi.org/10.1121/1.2019478View Description Hide Description
In this talk I shall present some results on quantifying the effects of small room reverberation on diotic subjective preference of speech under conditions of no interfering noise. All reverberation was generated digitally using an image model of a rectangular room [Allen and Berkley, J. Acoust. Soc. Am. 65, 943 (1979)]. Our main result is given by the equation , where P is the subjective preference in some arbitrary units, P max is the maximum possible preference, σ is the room spectral variance [Jetzt, J. Acoust. Soc. Am. 65, 1204–1211 (1979)], and T 60 is the reverberation time. The spectral variance is a known function of Δ/Δ c , where Δ is the source‐receiver distance and Δ c is the critical distance (Jetzt). In our experiment, σ went from 0.5 to 5.5, and T 60 ranged between 75 and 500 ms. The above result will be explained in some detail.
71(1982); http://dx.doi.org/10.1121/1.2019479View Description Hide Description
A unified approach to simulating auditorium acoustics electronically is described. The technique begins with a modeling phase in which the geometry and acoustical characteristics of an auditorium are programmed into a computer. The computer calculates the directional impulse response characterizing the transmission of sound from the source to the listener. The impulse response is convolved with an audio signal in real time by using a high‐speed transversal filter. The result of the convolution can be presented to a listener with a few speakers and still achieve a result which is natural‐sounding and realistic.
71(1982); http://dx.doi.org/10.1121/1.2019480View Description Hide Description
The design of Godot, a computer system for computer‐aided room acousticsmodeling and simulation, was described in an earlier paper [J. Acoust. Soc. Am. Suppl. 1 69, S36 (1981)]. In this paper, issues affecting its use in the production environment are discussed and current research issues are examined briefly. The system has been moved from a mainframe computer to a minicomputer having graphics facilities suitable for interactive design; we will review the way room acousticsmodeling and simulation are treated in the larger context of computer‐aided architectural design. Taped examples of simple simulations will be presented, and compared with examples recorded in the actual rooms. Godot currently traces sound beams incrementally; in this way it avoids exhaustive intersection testing with all faces in the room representation. Current work aims at generalizing the orientation‐dependent “bounding boxes” used in coarse “hit” detection to yield an intrinsic bounding volume for each wall element. As a side benefit this method will allow us to model moving portions of a room. [Work supported by the Science Council of British Columbia, grant ♯41 (RC‐4).]
- Session C. Speech Communication I: Speech Processing: Analysis and Synthesis
- Contributed Papers
71(1982); http://dx.doi.org/10.1121/1.2019520View Description Hide Description
Speech recognition systems based on spectral template‐matching have proven successful in restricted speech understanding tasks: especially, speaker‐dependent recognition of isolated words with acoustically distinct vocabularies. Such systems, however, have only been of limited utility where vocabulary items are acoustically similar, as in the set B, C, D, E, G, P, T, V, and Z. We will describe a hybrid system which supplements the spectral template with a set of feature measurements extracted from the signal. A learning mechanism automatically extracts a weighting vector, which emphasizes informative features as well as informative frames in the template. Recognition performance of the hybrid system is superior to the performance of the system using either the spectral information alone or the featural information alone. Hybrid recognition architectures are therefore useful in overcoming present limitations of speech recognition systems. [Work supported by NSF.]
71(1982); http://dx.doi.org/10.1121/1.2019521View Description Hide Description
The vector quantization method described by Linde et al. [IEEE Trans. Commun. 28, 84–95 (1980)] was used to optimally quantize speech spectral principal components. The spectral principal components, obtained from linear combinations of spectral band energies, are an optimal uncorrelated parameter set containing a very large amount of spectral variance with a relatively small number of dimensions. For a given percentage of quantization distortion, the vector quantizer required fewer bits per frame than individual quantization of principal components, thus indicating a clustering of the speech data in the uncorrelated parameter space. Using the principal components parameters to measure both gain and spectral shape and using eight bits per frame, the mean square vector quantization error was about 2.5% in a four‐dimensional space, and about 5% in a seven‐dimensional space. Speech synthesized from the vector quantized principal components using nine bits per frame was somewhat inferior in quality to speech synthesized with full precision principal components, but quite intelligible. [Work supported by NSF.]
71(1982); http://dx.doi.org/10.1121/1.2019522View Description Hide Description
An unconstrained optimization technique is used to find the values of parameters of a combination of an articulatory and a vocal tract model that minimize the difference between model spectra and natural speech spectra. The articulatory model is anatomically realistic and the vocal tract model is a “lossy” Webster equation for which a method of solution is given. For English vowels in the steady state, anatomically reasonable articulatory configurations whose corresponding spectra match those of human speech to within 2 dB have been computed in fewer than ten iterations. Results are also given which demonstrate a limited ability of the system to track the articulatory dynamics of voiced speech. This is a continuation of work reported earlier [S. E. Levinson and C. E. Schmidt, J. Acoust. Soc. Am. Suppl. 1 68, S19 (1980)].
71(1982); http://dx.doi.org/10.1121/1.2019523View Description Hide Description
This paper will present a discussion of approaches used in the area of language identification. Past efforts have attempted to develop an automatic language identification system based upon the employment of linguistic units, i.e., phones or phonemes, with moderate results. An approach currently under investigation at SCRL is the application of pattern analysis techniques to features extracted from the speech signal, without regard to specific linguistic units. Upon completion of the first phase of the research, it is likely that a very unusual set of features will prove to be the most useful one in identifying languages. It will be of interest to examine the relationships that may exist between those features which are determined to be the greatest contributors to the classification process and basic linguistic units. [Work supported by RADC.]
71(1982); http://dx.doi.org/10.1121/1.2019524View Description Hide Description
We developed an approach to abridging the short‐term LPC residual and using the abridged information to generate speech at 16.0 kilobits per second (kb/s). The abridgement logic requires an average of less than one multiply per speech input sample. It offers a simple way to add a high‐rate, improved quality capability to existing 2.4 kb/s LPC voice communication systems. Our simulated 16.0 kb/s residual‐excited linear prediction (RELP) transmission system using abridgement generates very good quality speech output. Our abridgement approach is independent of voiced/unvoiced decisions, restricts the influence of errors to a single frame, and strongly resists brief input anomalies. We designed this medium‐rate system as an alternative to the baseband residual‐excited coder currently used in the NRL Multirate Processor (MRP). Our (computationally) simpler approach scores lower overall on the Diagnostic Rhyme Test than the MRP baseband system at 16.0 kb/s: 87.4 vs 93.9. Listening suggests that the speech quality from our approach is closer to the MRP quality for continuous speech than for the DRT words. For some applications, the computational efficiency and stability of our approach could out‐weigh the quality advantage of the MRP.