Pub Date : 1972-09-01DOI: 10.1002/J.1538-7305.1973.TB01965.X
P. Bonyhard, T. Nelson
Bubble technology offers several operations that have no equivalents in technologies based on magnetic recording. Examples of such operations are: transfer, reversal of the direction of propagation, and opening and closing of gaps in the data stream. This paper∗ shows how such operations can be used to dynamically reallocate data in the bubble memory, causing it to become an integrated memory hierarchy. A considerable improvement in performance results. A model is presented which relates the bubble memory with dynamic reallocation to stack processing, a technique used in the evaluation of memory hierarchies. With the aid of this model it becomes possible to calculate the performance of the bubble memory using published data derived from the traces of selected typical programs. Memory design is optimized for the execution of such programs. Design parameters are proposed for a 2-Mb bubble memory with 128 detectors which, in the execution of the type of program for which data were available, requires an average of only 8.8 shifts for access and an average of 12.1 shifts per memory cycle. If bubbles are propagated at a rate of 1 MHz, the average access and cycle times for this memory become 8.8 µs and 12.1 µs, respectively. Such performance, in conjunction with the low cost per bit offered by bubble technology, is expected to have a major impact. The performance of this memory, when operated in conjunction with a faster buffer, is also calculated. The use of a 64-kb buffer is shown to reduce the average number of shifts for access to 1.05, and the average number of shifts per cycle to 1.9.
{"title":"Dynamic data reallocation in bubble memories","authors":"P. Bonyhard, T. Nelson","doi":"10.1002/J.1538-7305.1973.TB01965.X","DOIUrl":"https://doi.org/10.1002/J.1538-7305.1973.TB01965.X","url":null,"abstract":"Bubble technology offers several operations that have no equivalents in technologies based on magnetic recording. Examples of such operations are: transfer, reversal of the direction of propagation, and opening and closing of gaps in the data stream. This paper∗ shows how such operations can be used to dynamically reallocate data in the bubble memory, causing it to become an integrated memory hierarchy. A considerable improvement in performance results. A model is presented which relates the bubble memory with dynamic reallocation to stack processing, a technique used in the evaluation of memory hierarchies. With the aid of this model it becomes possible to calculate the performance of the bubble memory using published data derived from the traces of selected typical programs. Memory design is optimized for the execution of such programs. Design parameters are proposed for a 2-Mb bubble memory with 128 detectors which, in the execution of the type of program for which data were available, requires an average of only 8.8 shifts for access and an average of 12.1 shifts per memory cycle. If bubbles are propagated at a rate of 1 MHz, the average access and cycle times for this memory become 8.8 µs and 12.1 µs, respectively. Such performance, in conjunction with the low cost per bit offered by bubble technology, is expected to have a major impact. The performance of this memory, when operated in conjunction with a faster buffer, is also calculated. The use of a 64-kb buffer is shown to reduce the average number of shifts for access to 1.05, and the average number of shifts per cycle to 1.9.","PeriodicalId":55391,"journal":{"name":"Bell System Technical Journal","volume":"148 1","pages":"307-317"},"PeriodicalIF":0.0,"publicationDate":"1972-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88660550","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1972-07-08DOI: 10.1002/J.1538-7305.1972.TB02652.X
J. McKenna, J. Morrison
A special class of transmission lines is considered, in which the modes decompose into two noninteracting sets. Both a single transmission line with constant characteristic impedance and variable propagation factor, and two transmission lines with equal propagation factors and variable coupling, in which the forward modes do not interact with the backward modes, are investigated. Exact expressions are obtained for the reflection and transmission coefficients when a section of such a transmission system connects two semi-infinite transmission systems consisting of constant impedance and admittance lines. These results hold for arbitrarily varying propagation factors and coupling; and while they are of independent interest in the case of deterministic variations, we make an application of them here in the case of stochastic variations. Exact results are obtained for the ensemble averages of the transmission coefficient and transmitted power, and their variances, for the inserted section of single line, when the variable propagation factor is a random function involving either a Gaussian process or the random telegraph process. Asymptotic results are also obtained in the general case of weak fluctuations and long inserted sections. Analogous results may be obtained for the inserted section of two lines when they are randomly coupled, and the results are given in the case of matched lines, for which no reflections occur. Finally, some of the time domain statistics for lossless lines are considered, and expressions are derived for the ensemble averages of the transmitted pulse, due to pulses incident on the inserted section.
{"title":"Exact solutions to some deterministic and random transmission line problems","authors":"J. McKenna, J. Morrison","doi":"10.1002/J.1538-7305.1972.TB02652.X","DOIUrl":"https://doi.org/10.1002/J.1538-7305.1972.TB02652.X","url":null,"abstract":"A special class of transmission lines is considered, in which the modes decompose into two noninteracting sets. Both a single transmission line with constant characteristic impedance and variable propagation factor, and two transmission lines with equal propagation factors and variable coupling, in which the forward modes do not interact with the backward modes, are investigated. Exact expressions are obtained for the reflection and transmission coefficients when a section of such a transmission system connects two semi-infinite transmission systems consisting of constant impedance and admittance lines. These results hold for arbitrarily varying propagation factors and coupling; and while they are of independent interest in the case of deterministic variations, we make an application of them here in the case of stochastic variations. Exact results are obtained for the ensemble averages of the transmission coefficient and transmitted power, and their variances, for the inserted section of single line, when the variable propagation factor is a random function involving either a Gaussian process or the random telegraph process. Asymptotic results are also obtained in the general case of weak fluctuations and long inserted sections. Analogous results may be obtained for the inserted section of two lines when they are randomly coupled, and the results are given in the case of matched lines, for which no reflections occur. Finally, some of the time domain statistics for lossless lines are considered, and expressions are derived for the ensemble averages of the transmitted pulse, due to pulses incident on the inserted section.","PeriodicalId":55391,"journal":{"name":"Bell System Technical Journal","volume":"32 1","pages":"1269-1292"},"PeriodicalIF":0.0,"publicationDate":"1972-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76115355","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1972-07-08DOI: 10.1002/J.1538-7305.1972.TB02653.X
N. Jayant, L. Rabiner
By adding a pseudo-random “dither” noise to a signal X that is to be quantized, and by subtracting an identical noise sequence from the quantizer output, it is possible to break up undesirable signal-dependent patterns in the quantization error sequence, without increasing the variance of the error E. The idea has been widely discussed in the context of picture coding, and it is the purpose of this paper to demonstrate application of the technique to the quantization of speech signals. Computer simulations have shown how the use of dither whitens the quantization error sequence in PCM encoding, and renders it more acceptable than signal-correlated errors of equal variance. We demonstrate, for conditions of dither and no dither, typical speech recordings, illustrative error waveforms, and data on signal-to-error correlation C, and indicate how the advantage of dithering increases monotonically with crudeness of signal quantization and becomes significant when the number of bits per sample is less than about six. While the parameter C is a simple criterion for demonstrating the effect of dither, it must be emphasized that the truly relevant criterion is the statistical independence of E and X, and not merely the decorrelation of these functions. Thus, for example, we show that for the case of a reciprocal PDF (probability density function) for X, a zero value of C can be achieved without dither. For purposes of implementation, it is desirable to employ dither noise values characterized by a discrete PDF, with a support that is equal to an integral multiple of the step-size Δ x in the quantizer. We show that for effective dithering, the step-size Δ N in the noise PDF need be no smaller, typically, than Δ x /4. Finally, we indicate an application of dither to the quantization of speech signals by delta modulation.
通过添加一个伪随机“优柔寡断”噪声信号X是量子化的,减去一个相同的噪声序列从量化器输出,可以打破不良相互依赖模式的量化误差序列,不增加误差的方差大肠的想法已经被广泛讨论的上下文中图片编码,它的目的是本文演示的应用技术,语音信号的量化。计算机模拟显示了抖动如何使PCM编码中的量化误差序列白化，并使其比等方差的信号相关误差更容易接受。对于抖动和无抖动的条件，我们展示了典型的语音记录、说白了的误差波形和信号误差相关C的数据，并指出抖动的优势如何随着信号量化的粗糙程度单调增加，并在每个样本的比特数小于约6时变得显著。虽然参数C是证明抖动效应的一个简单准则，但必须强调的是，真正相关的准则是E和X的统计独立性，而不仅仅是这些函数的去相关。因此，例如，我们证明了对于X的倒数PDF(概率密度函数)的情况，C的零值可以在没有抖动的情况下实现。为了实现目的，希望采用离散PDF特征的抖动噪声值，其支持等于量化器中步长Δ x的整数倍。我们表明，为了有效抖动，噪声PDF中的步长Δ N通常不需要小于Δ x /4。最后，我们指出了抖动在增量调制语音信号量化中的应用。
{"title":"The Application of dither to the quantization of speech signals","authors":"N. Jayant, L. Rabiner","doi":"10.1002/J.1538-7305.1972.TB02653.X","DOIUrl":"https://doi.org/10.1002/J.1538-7305.1972.TB02653.X","url":null,"abstract":"By adding a pseudo-random “dither” noise to a signal X that is to be quantized, and by subtracting an identical noise sequence from the quantizer output, it is possible to break up undesirable signal-dependent patterns in the quantization error sequence, without increasing the variance of the error E. The idea has been widely discussed in the context of picture coding, and it is the purpose of this paper to demonstrate application of the technique to the quantization of speech signals. Computer simulations have shown how the use of dither whitens the quantization error sequence in PCM encoding, and renders it more acceptable than signal-correlated errors of equal variance. We demonstrate, for conditions of dither and no dither, typical speech recordings, illustrative error waveforms, and data on signal-to-error correlation C, and indicate how the advantage of dithering increases monotonically with crudeness of signal quantization and becomes significant when the number of bits per sample is less than about six. While the parameter C is a simple criterion for demonstrating the effect of dither, it must be emphasized that the truly relevant criterion is the statistical independence of E and X, and not merely the decorrelation of these functions. Thus, for example, we show that for the case of a reciprocal PDF (probability density function) for X, a zero value of C can be achieved without dither. For purposes of implementation, it is desirable to employ dither noise values characterized by a discrete PDF, with a support that is equal to an integral multiple of the step-size Δ x in the quantizer. We show that for effective dithering, the step-size Δ N in the noise PDF need be no smaller, typically, than Δ x /4. Finally, we indicate an application of dither to the quantization of speech signals by delta modulation.","PeriodicalId":55391,"journal":{"name":"Bell System Technical Journal","volume":"12 1","pages":"1293-1304"},"PeriodicalIF":0.0,"publicationDate":"1972-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87902104","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1972-07-08DOI: 10.1002/J.1538-7305.1972.TB02660.X
A. Bobeck, S. Blank, H. Levinstein
A conventional magnetic bubble material consists of a magnetic garnet film deposited on a nonmagnetic substrate. Garnet films with stress- and/or growth-induced uniaxial anisotropy are deposited by chemical vapor deposition (CVD) or liquid phase epitaxy (LPE) usually on Gd 3 Ga 5 O 12 substrates. In this B.S.T.J. Brief we report on the properties of multilayer garnet films deposited by LPE.
{"title":"Multilayer epitaxial garnet films for magnetic bubble devices-hard bubble suppression","authors":"A. Bobeck, S. Blank, H. Levinstein","doi":"10.1002/J.1538-7305.1972.TB02660.X","DOIUrl":"https://doi.org/10.1002/J.1538-7305.1972.TB02660.X","url":null,"abstract":"A conventional magnetic bubble material consists of a magnetic garnet film deposited on a nonmagnetic substrate. Garnet films with stress- and/or growth-induced uniaxial anisotropy are deposited by chemical vapor deposition (CVD) or liquid phase epitaxy (LPE) usually on Gd 3 Ga 5 O 12 substrates. In this B.S.T.J. Brief we report on the properties of multilayer garnet films deposited by LPE.","PeriodicalId":55391,"journal":{"name":"Bell System Technical Journal","volume":"61 1","pages":"1431-1435"},"PeriodicalIF":0.0,"publicationDate":"1972-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88518898","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1972-07-08DOI: 10.1002/J.1538-7305.1972.TB02650.X
D. Marcuse
Using coupled power equations to describe the average performance of a multimode waveguide with random coupling, it is shown that a Gaussian input pulse remains approximately Gaussian with a pulse width that increases proportionally to the square root of the length of the waveguide. The proportionality factor is determined for the model of a slab waveguide. Since coupling between guided modes of necessity causes coupling of some of the guided modes to radiation modes, radiation losses are un-avoidable. A desired improvement in pulse distortion that is accomplished by coupling the guided modes intentionally to each other must be paid for by a certain loss penalty. This loss penalty is also evaluated for the special case of the slab waveguide model. Pulse dispersion improvement can be achieved by providing intentional roughness of the core-cladding interface of the dielectric waveguide. The “power spectrum” of the core-cladding interface function must be designed very carefully in order to minimize the radiation loss penalty that accompanies any attempt to reduce pulse dispersion. The dependence of the loss penalty on the shape of the “power spectrum” of the core-cladding interface function is studied in this paper. Design criteria for the improvement of multimode pulse dispersion are given based on the slab waveguide model. The connection between the slab waveguide model and the round optical fiber is pointed out.
{"title":"Pulse propagation in multimode dielectric waveguides","authors":"D. Marcuse","doi":"10.1002/J.1538-7305.1972.TB02650.X","DOIUrl":"https://doi.org/10.1002/J.1538-7305.1972.TB02650.X","url":null,"abstract":"Using coupled power equations to describe the average performance of a multimode waveguide with random coupling, it is shown that a Gaussian input pulse remains approximately Gaussian with a pulse width that increases proportionally to the square root of the length of the waveguide. The proportionality factor is determined for the model of a slab waveguide. Since coupling between guided modes of necessity causes coupling of some of the guided modes to radiation modes, radiation losses are un-avoidable. A desired improvement in pulse distortion that is accomplished by coupling the guided modes intentionally to each other must be paid for by a certain loss penalty. This loss penalty is also evaluated for the special case of the slab waveguide model. Pulse dispersion improvement can be achieved by providing intentional roughness of the core-cladding interface of the dielectric waveguide. The “power spectrum” of the core-cladding interface function must be designed very carefully in order to minimize the radiation loss penalty that accompanies any attempt to reduce pulse dispersion. The dependence of the loss penalty on the shape of the “power spectrum” of the core-cladding interface function is studied in this paper. Design criteria for the improvement of multimode pulse dispersion are given based on the slab waveguide model. The connection between the slab waveguide model and the round optical fiber is pointed out.","PeriodicalId":55391,"journal":{"name":"Bell System Technical Journal","volume":"18 1","pages":"1199-1232"},"PeriodicalIF":0.0,"publicationDate":"1972-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"91055172","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1972-07-08DOI: 10.1002/J.1538-7305.1972.TB02649.X
L. Rabiner
In this paper it is shown how standard linear programming techniques can be applied to designing finite impulse response digital filters. Attention is concentrated on designing filters having exactly linear phase, and arbitrary magnitude response. The design method is illustrated by examples of the design of frequency sampling filters with constraints on in-band ripple, optimal filters where the passband and stopband cutoff frequencies may be specified exactly, and filters with simultaneous constraints on the time response and frequency response.
{"title":"The design of finite impulse response digital filters using linear programming techniques","authors":"L. Rabiner","doi":"10.1002/J.1538-7305.1972.TB02649.X","DOIUrl":"https://doi.org/10.1002/J.1538-7305.1972.TB02649.X","url":null,"abstract":"In this paper it is shown how standard linear programming techniques can be applied to designing finite impulse response digital filters. Attention is concentrated on designing filters having exactly linear phase, and arbitrary magnitude response. The design method is illustrated by examples of the design of frequency sampling filters with constraints on in-band ripple, optimal filters where the passband and stopband cutoff frequencies may be specified exactly, and filters with simultaneous constraints on the time response and frequency response.","PeriodicalId":55391,"journal":{"name":"Bell System Technical Journal","volume":"40 1","pages":"1177-1198"},"PeriodicalIF":0.0,"publicationDate":"1972-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77699905","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1972-07-08DOI: 10.1002/J.1538-7305.1972.TB02662.X
A. Rosencwaig
A new class of magnetic bubbles designated as hard (and intermediate) bubbles has been found to be a common feature in many bubble garnet films.1 These hard bubbles have been very disruptive to the operation of bubble circuits since they not only have a much lower mobility than normal bubbles, but also tend to move at an angle rather than parallel to the direction of the driving field gradient. Fortunately, it has recently been found that the presence of a second magnetic layer apparently eliminates these hard bubbles.2,3 This second layer can either be a growth layer with a sufficiently small moment so that its magnetization under a bias field is always oppositely directed to the magnetization within the bubble,2 or it can be a layer with magnetization perpendicular to the bubble magnetization. This latter layer might be produced by ion implantation to the point where a stress-induced uniaxial anisotropy in the plane of the film overcomes the previously existing anisotropy.3 We propose in this B.S.T.J. Brief that the apparent elimination of these hard bubbles is due to the presence of the domain wall between the bubble and this second layer.
{"title":"The effect of a second magnetic layer on hard bubbles","authors":"A. Rosencwaig","doi":"10.1002/J.1538-7305.1972.TB02662.X","DOIUrl":"https://doi.org/10.1002/J.1538-7305.1972.TB02662.X","url":null,"abstract":"A new class of magnetic bubbles designated as hard (and intermediate) bubbles has been found to be a common feature in many bubble garnet films.1 These hard bubbles have been very disruptive to the operation of bubble circuits since they not only have a much lower mobility than normal bubbles, but also tend to move at an angle rather than parallel to the direction of the driving field gradient. Fortunately, it has recently been found that the presence of a second magnetic layer apparently eliminates these hard bubbles.2,3 This second layer can either be a growth layer with a sufficiently small moment so that its magnetization under a bias field is always oppositely directed to the magnetization within the bubble,2 or it can be a layer with magnetization perpendicular to the bubble magnetization. This latter layer might be produced by ion implantation to the point where a stress-induced uniaxial anisotropy in the plane of the film overcomes the previously existing anisotropy.3 We propose in this B.S.T.J. Brief that the apparent elimination of these hard bubbles is due to the presence of the domain wall between the bubble and this second layer.","PeriodicalId":55391,"journal":{"name":"Bell System Technical Journal","volume":"15 1","pages":"1440-1444"},"PeriodicalIF":0.0,"publicationDate":"1972-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88188895","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1972-07-08DOI: 10.1002/J.1538-7305.1972.TB02651.X
K. Ishizaka, J. Flanagan
A model of voiced-sound generation is derived in which the detailed acoustic behavior of the human vocal cords and the vocal tract is computed. The vocal cords are approximated by a self-oscillating source composed of two stiffness-coupled masses. The vocal tract is represented as a bilateral transmission line. One-dimensional Bernoulli flow through the vocal cords and plane-wave propagation in the tract are used to establish acoustic factors dominant in the generation of voiced speech. A difference-equation description of the continuous system is derived, and the cord-tract system is programmed for interactive study on a DDP-516 computer. Sampled waveforms are calculated for: acoustic volume velocity through the cord opening (glottis); glottal area; and mouth-output sound pressure. Functional relations between fundamental voice frequency, subglottal (lung) pressure, cord tension, glottal area, and duty ratio of cord vibration are also determined. Results show that the two-mass model duplicates principal features of cord behavior in the human. The variation of fundamental frequency with subglottal pressure is found to be 2 to 3 Hz/cm H 2 O, and is essentially independent of vowel configuration in the programmed tract. Acoustic interaction between tract eigenfrequencies and glottal volume flow is strong. Phase difference in motion of the cord edges is in the range of 0 to 60 degrees, and control of cord tension leads to behavior analogous to chest/falsetto conditions in the human. Phonation-neutral, or rest area of cord opening, is shown to be a critical factor in establishing self-oscillation. Finally, the complete synthesis system suggests an efficient, physiological description of the speech signal, namely, in terms of subglottal pressure, cord tension, rest area of cord opening, and vocal-tract shape.
{"title":"Synthesis of voiced sounds from a two-mass model of the vocal cords","authors":"K. Ishizaka, J. Flanagan","doi":"10.1002/J.1538-7305.1972.TB02651.X","DOIUrl":"https://doi.org/10.1002/J.1538-7305.1972.TB02651.X","url":null,"abstract":"A model of voiced-sound generation is derived in which the detailed acoustic behavior of the human vocal cords and the vocal tract is computed. The vocal cords are approximated by a self-oscillating source composed of two stiffness-coupled masses. The vocal tract is represented as a bilateral transmission line. One-dimensional Bernoulli flow through the vocal cords and plane-wave propagation in the tract are used to establish acoustic factors dominant in the generation of voiced speech. A difference-equation description of the continuous system is derived, and the cord-tract system is programmed for interactive study on a DDP-516 computer. Sampled waveforms are calculated for: acoustic volume velocity through the cord opening (glottis); glottal area; and mouth-output sound pressure. Functional relations between fundamental voice frequency, subglottal (lung) pressure, cord tension, glottal area, and duty ratio of cord vibration are also determined. Results show that the two-mass model duplicates principal features of cord behavior in the human. The variation of fundamental frequency with subglottal pressure is found to be 2 to 3 Hz/cm H 2 O, and is essentially independent of vowel configuration in the programmed tract. Acoustic interaction between tract eigenfrequencies and glottal volume flow is strong. Phase difference in motion of the cord edges is in the range of 0 to 60 degrees, and control of cord tension leads to behavior analogous to chest/falsetto conditions in the human. Phonation-neutral, or rest area of cord opening, is shown to be a critical factor in establishing self-oscillation. Finally, the complete synthesis system suggests an efficient, physiological description of the speech signal, namely, in terms of subglottal pressure, cord tension, rest area of cord opening, and vocal-tract shape.","PeriodicalId":55391,"journal":{"name":"Bell System Technical Journal","volume":"12 1 1","pages":"1233-1268"},"PeriodicalIF":0.0,"publicationDate":"1972-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87823008","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1972-07-08DOI: 10.1002/J.1538-7305.1972.TB02655.X
Yo-Sung Cho
Two methods are described for the optimal equalization of a channel with “Bump” Equalizers composed of several adjustable-gain Bode Networks. The first method is a general one and applies a steepest descent algorithm which minimizes the total mean-squared error (MSE) of the equalized channel. It requires continuous gradient information on the error-gain relationship in order to determine exactly the optimum equalizer adjustments and involves a relatively complicated procedure to calculate the gradient. However, the second method, which also applies a steepest descent algorithm, develops the necessary gradient information with knowledge of the error signal only at selected frequencies across the bandwidth occupied by the channel. Under idealized assumptions, it is shown that the gradients obtained by the second method are exact. When the assumptions do not apply exactly, it is shown by computer simulation that the difference between the gradients obtained by the two methods is very small. A significant potential advantage of the second method lies in the hardware realization which only requires the measurement of the channel error at 2M — 1 frequencies at the equalizing station (where M is the number of Bode Networks in the equalizer). From these frequency domain errors, the gradients can be generated as real-time signals and applied to the appropriate adjustable elements to obtain the optimum gain settings for minimum MSE.
{"title":"Optimal equalization of wideband coaxial cable channels using “bump” equalizers","authors":"Yo-Sung Cho","doi":"10.1002/J.1538-7305.1972.TB02655.X","DOIUrl":"https://doi.org/10.1002/J.1538-7305.1972.TB02655.X","url":null,"abstract":"Two methods are described for the optimal equalization of a channel with “Bump” Equalizers composed of several adjustable-gain Bode Networks. The first method is a general one and applies a steepest descent algorithm which minimizes the total mean-squared error (MSE) of the equalized channel. It requires continuous gradient information on the error-gain relationship in order to determine exactly the optimum equalizer adjustments and involves a relatively complicated procedure to calculate the gradient. However, the second method, which also applies a steepest descent algorithm, develops the necessary gradient information with knowledge of the error signal only at selected frequencies across the bandwidth occupied by the channel. Under idealized assumptions, it is shown that the gradients obtained by the second method are exact. When the assumptions do not apply exactly, it is shown by computer simulation that the difference between the gradients obtained by the two methods is very small. A significant potential advantage of the second method lies in the hardware realization which only requires the measurement of the channel error at 2M — 1 frequencies at the equalizing station (where M is the number of Bode Networks in the equalizer). From these frequency domain errors, the gradients can be generated as real-time signals and applied to the appropriate adjustable elements to obtain the optimum gain settings for minimum MSE.","PeriodicalId":55391,"journal":{"name":"Bell System Technical Journal","volume":"483 1","pages":"1327-1345"},"PeriodicalIF":0.0,"publicationDate":"1972-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76364506","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1972-07-08DOI: 10.1002/J.1538-7305.1972.TB02661.X
R. Wolfe, J. North
Hard bubbles are cylindrical magnetic domains which differ from normal bubbles in their static and dynamic properties.1–3 They form and collapse at fields considerably higher than normal bubble domains (hence the name “hard bubbles”) and, under the action of a gradient in the perpendicularly directed magnetic field, they have a component of motion either to the right or to the left rather than straight down the gradient (right- and left-handed hard bubbles). Under the influence of a small alternating field superimposed on the dc bias field, normal bubbles oscillate in size or strip out in a random manner, whereas hard bubbles tend to strip out in an S-shape and rotate in either a counterclockwise direction for the normal S-shape or clockwise for the reverse S-shape. It has been found that high-speed propagation using permalloy overlay circuits is impossible in materials which are prone to hard bubble formation.
{"title":"Suppression of hard bubbles in magnetic garnet films by Ion implantation","authors":"R. Wolfe, J. North","doi":"10.1002/J.1538-7305.1972.TB02661.X","DOIUrl":"https://doi.org/10.1002/J.1538-7305.1972.TB02661.X","url":null,"abstract":"Hard bubbles are cylindrical magnetic domains which differ from normal bubbles in their static and dynamic properties.1–3 They form and collapse at fields considerably higher than normal bubble domains (hence the name “hard bubbles”) and, under the action of a gradient in the perpendicularly directed magnetic field, they have a component of motion either to the right or to the left rather than straight down the gradient (right- and left-handed hard bubbles). Under the influence of a small alternating field superimposed on the dc bias field, normal bubbles oscillate in size or strip out in a random manner, whereas hard bubbles tend to strip out in an S-shape and rotate in either a counterclockwise direction for the normal S-shape or clockwise for the reverse S-shape. It has been found that high-speed propagation using permalloy overlay circuits is impossible in materials which are prone to hard bubble formation.","PeriodicalId":55391,"journal":{"name":"Bell System Technical Journal","volume":"53 1","pages":"1436-1440"},"PeriodicalIF":0.0,"publicationDate":"1972-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90281004","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：