As a result of using the first machine of speech trainer produced for trial, various educational effects have become evident. However, because the use of this per pupil is restricted in very short time, the durability of training effects is liable to cause a serious problem. Therefore, in order to solve such a problem, we deal with in this paper the structure and function of a new speech training system which consists of an instruction trainer and a numbers of self-study trainers. The instruction trainer described here carries out the teaching and training of pronunciation for many deaf pupils in a school room, and this trainer is mainly used for measuring the training effects cause from the self-study trainer and selecting suitable spectral patterns for each pupil. Accordingly, the instruction trainer stands out clearly because of the following functions : (1) It has internally a great number of spectral patterns used as examples in speech training, of vowels and any pupil can show these patterns on a Braun tube in simple operation. (2) It can indicate the deviation from the standard spectral pattern, which in shown by a pupil on the special pattern. (3) It can show the time transition of pitch patterns caused by both teacher and pupil, and draw these traces on a Braun tube at rest, by memorializing and reproducing at the some time of utterance. (4) It can reproduce on a Braun tube the three-dimensional correspondence of time, frequency and magnitude of each frequency component in the spectral patterns. On the other hand, the self-study trainer is a simple device which uses the suitable spectral patterns selected by the instruction trainer as examples of utterance, and this is very useful for speech training under the instruction of parents at home. After all, this is safely said to be a simple, but efficacious, pedagogical device, in view of its economical standard as well as its function.

View full abstract

In the past decade, the of development of digital processing technique of signals with computers has yielded a remarkable progress in the acoustic aspects of musical instrument tones and several papers of this analysis have been published. As a part systematic investigations of the analysis and synthesis of musical instrument tones, is reported here an application of the Discrete Fourier Transform to musical instrument tones with special regard to the temporal fluctuation of physical parameters of the tones. New mathematical procedure of error estimation is proposed, and an analysis of physical features with actual example is presented. Generally, the following three assumptions are postulated for the application of Discrete Fourier Transform to musical instrument tones : (1) The spectrum of a musical instrument tone is of harmonic constitution. (2) In the duration of a musical note, the fundamental frequency is held constant. (3) The results of the DFT analysis in a transient state represent averaged values for amplitude and frequency in the sampled interval. First, on these assumptions, a duration of a musical instrument tone was divided into sequential waves of short-period frames (Fig. 2). Next, a time series of spectrum of each frame was constructed, with the application of DFT to each frame. The number of samples in a frame set at 2^γ(γ : integer), since the FFT algorithm is based on 2, and the sampling frequency was set to make both the wave form and data window pitch synchronous. Since, however, the actual pitch of a musical instrument tone is regarded as fluctuating instantaneously, it is suspected that the pitch might become asynchronous with the window and errors might be brought into the results of the DFT analysis. Then, a new estimation procedure of errors is proposed, utilizing the properties of a convolution of the spectral window of the data window and the Fourier transform of singnals. The spectrum of a musical instrument sound wave from the DFT is represent by G'(mω)=Σ^^^m___&ltm=-m&gt G(mω_1)・W(nm-mω_1), where G(mω_1) is the Fourier transform of the signal and W(nm-mω_1) is an arbitrary spectrum window. Here G'(mω) is the theoretical spectrum including errors due to a slight variation of fundamental angular frequency ω_1 of the signal deviated from the angular frequency ω, which is pitchsynchronous with the data window. The errors were estimated by calculating the ratio of this spectral value to a modelized, true spectral value. If a Hamming window is used as the data window, the spectral error up to the 8th harmonic in the vicinity of the attack transient portions is less than 3 dB owing to the large fluctuation in pitch of actual musical instrument tone. However, in the steady state the error is of the order of 0. 1 to 0. 3 dB, showing that sufficient accuracy is achieved by this analytical method on the three assumption stated above. Although this analytical procedure is applied to various natural musical instrument tones, only the results of violin A_4 are presented here. In Fig. 8, it is shown that the attack transient form the violin tone is exponential with an attack time of 100 msec, and that even in the steady state, marked modulations in frequency (vibrato) and amplitude are observed.

View full abstract

There are Various problems to be solved concerning propagation of noise in our environment. One of the important problems is to investigate the effect of directivity character of sources on the noise field. From the practical point of view the following subjects are studied in this paper. In Chapter 2, we consider the noise due to a random distribution of directional point sources, simulating a free flow of vehicles on a straight lane. The cumulants of sound intensity are derived in Eq. (2. 3) and compared with those of usual omnidirectional point sources. Figs. 2 and 3 show directivity patterns of each component source and of composed ones respectively. In Chapter 3, we consider radiation characteristics of noise from an aperture of enclosure in which the sound field is perfectly diffuse. By using the Fresnel-Kirchhoff diffraction theory, it is derived that each point on the aperture is equivalent to a point source whose directivity factor is given in Eq. (3. 4). The results are applied to predict the sound field around the aperture of a cutting road and tunnel. Directivity pattern and attenuation characteristics of noise emitted from the cutting road are given in Figs. 5 and 6 respectively, which show the intensity of lateral direction is less than that of vertical direction by about 7 dB in the far field. Similar directivity pattern of noise emitted from the end of tunnel with semi-circular cross section is shown in Fig. 8. In Chapter 4, we also consider a random distribution of directional point sources on a plane as a simple mathematical model of urban noise. The cumulants of sound intensity along the vertial axis are expressed by Eq. (4. 4) provided that the directivity factor of each source is given by Eq. (4. 1). We see that the mean sound intensity is independent of height z. The variance, however, is inversely proportional to z^2 and hence decreases with increasing height. Qualitatively, these results are consistent with measured ones for noises in urban area.

View full abstract

The deaf may be exposed to a number of dangers in his daily life when he cannot notice a voice of warning or the approaching sound of a car from behind or out of his visible range. Nevertheless, there is scarcely any attempt to develop a prosthetic sound localizer via his remaining tactile or visual sense to assist him. The authors tried to find a practical method for a tactual localizer of sound for the deaf. It is reported that two channel vibratory stimuli presented to each of two points on the surface of the skin cause a phantom sensation between these two stimulated points. Based on this phantom sensation, we have tried to substitute auditory sound localization by corresponding the position of the tactile phantom sensation with the direction of a sound source. First of all, we compared the tactile sense with the auditory sense by means of psychophysical techniques (Fig. 1). As the result, the phantom sensation by intensity defference between two channels was similar to the auditory lateralization, however in regard to the time difference, it was much inferior in comparison to that of the auditory system (Figs. 3, 4, 5, and 6). It is also reported that the auditory sense localizes a sound source by the intensity difference between the sounds received by right and left ears for the higher tones with the frequencies more than 1000〜1500 Hz, while for lower tones, by time difference, as the intensity deifference is very little for lower frequency tones. On the other hand, a car noise or a voice subject to the localization in our daily life has relatively higher energy in the range of lower frequencies than 500〜1000 Hz. Hence, it seemed necessary to transform the time defference of the lower sound into intensity defference to obtain better tactile recognition of sound localization. Considering these characteristics of the tactile sense and other conditions, we designed a time intensity converter (Fig. 7, 8, 9, and 10). Using our tactual sound localizer with this converter, a subject who was equipped with a pair of vibrators 4〜20 cm apart on the skin sueface of hand, abdomen, breast, or back could detect the approaching direction of a car with a velocity of 20 km/h and its movement from rihgt to left. Problems of localization of multiple sound sources and approximate distinction of sound have not been solved. We are trying to find new techniques for these problems.

View full abstract

The finite element approch for acoustic transmission systems was discussed by Kagawa, and Young and Crocker, and the usability of the technique has been verified experimentally for acoustic filters with arbitrary circular cross section. The treatment is based on the electrical four-terminal network analogy with regard to the input and output ports of the filter, which is considered tobe ideally reactive or free from dissipation. The present work is the extension of our previous work to a more general system with a wall of arbitrary acoustic imdedance. The load impedance at the output port can simply be regarded as a part of the wall. so that the transmission characterstic can be computed directly. The finite element formulation is developed on the basis of the true-adjoint system approach. For the discretization, the second order polynomial is used for the traial function of the trangular ring element. A computer program is developed, with which the transmission loss of the expansion-type acoustic filters is calculated. The wall of the chamber is partly treated with sound-absorbing felt. The calculation results coincide well with the experiment results (Figs. 8〜11). The program is also applied to the simulation of the acoustic horn problem, another example of an acoustic transmission sytem with variable cross section. A model is considered in which the radiation half sphere is assumed to be surrounded by a hypothetical wall with an air impedance ρc and the finete element calculation is applied inside. The driving-point impedance of the horn throat is calculated, the result of which again agrees reasonably with the measured result (Fig. 13). It is to be concluded that the simulation technique thus developed can provide a means of predicting and designing the behaviour and the characteristics of acousitc transmission systems of this kind.

View full abstract