By Topic

Acoustics, Speech and Signal Processing, IEEE Transactions on

Issue 2 • Date April 1975

Filter Results

Displaying Results 1 - 18 of 18
  • [Front cover and table of contents]

    Page(s): 0
    Save to Project icon | Request Permissions | PDF file iconPDF (154 KB)  
    Freely Available from IEEE
  • [Back cover]

    Page(s): c4
    Save to Project icon | Request Permissions | PDF file iconPDF (740 KB)  
    Freely Available from IEEE
  • Digital pulse compression via fast convolution

    Page(s): 189 - 201
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1472 KB)  

    The mathematical structure of the digital ambiguity function for a matched filtered linear FM (LFM) waveform is derived as a function of time-bandwidth product, sampling rate, and arbitrary delay and frequency shifts. It is found to be well behaved for sampling rates equal to or greater than the swept signal bandwidth, provided that time sidelobes are controlled using standard frequency domain weighting techniques. A digital convolution processor comprised of cascaded pipeline fast Fourier transforms (FFT's) is presented as a viable architecture for real-time filtering of moderately high bandwidth LFM signals, and tradeoffs among radix, pipeline clock rate, and convolutional efficiency are discussed. It is found that a modified floating-point computational scheme performs well in such a context and is especially useful if a large signal dynamic range must be accommodated. A radix-4 4096-point design example is considered and the effects of quantization and finite register length arithmetic upon the digital ambiguity function are demonstrated via simulation. It is found that input data, FFT coefficients, reference filter coefficients, and intermediate results can be represented with mantissas of modest bit length. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • On the approximation problem for recursive digital filters with arbitary attenuation curve in the pass-band and the stop-band

    Page(s): 202 - 207
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (448 KB)  

    The approximation of attenuation curve for a recursive digital filter is solved for arbitrary specifications in the stop-band and in the pass-band. The problem is split into two parts; the first one uses the classical theory of the transformed plane and the second one works with linear programming. The method is illustrated by some specific examples. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • The design of a spark discharge acoustic impulse generator

    Page(s): 157 - 162
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (656 KB)  

    In digital-processing techniques designed to remove room coloration from acoustic tests a short intense sound source is desirable. A spark discharge proves very suitable and this paper describes parameters influencing the design of such a generator. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • The existence of cepstra for two-dimensional rational polynomials

    Page(s): 242 - 243
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (296 KB)  

    The use of cepstral analysis is helpful for some problems where two one-dimensional signals are combined by convolution [1]. In such problems it is important to ensure that the phase function associated with the resultant signal may be defined so that it is a continuous, odd, and periodic function of frequency [2], [3]. One class of one-dimensional signals which have this property is the class whose z-transforms are rational polynomials [2]. In this correspondence, we shall show that these results are extendible to two dimensions, and that 2-D cepstra can be defined for 2-D rational polynomials. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Digital filter realization using successive multiplier-extraction approach

    Page(s): 235 - 239
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (464 KB)  

    A new approach to the digital filter realization problem is proposed in this paper. In this approach multipliers are extracted one at a time successively in the development of all realizations. The approach is illustrated by finding all possible realizations of a second-order transfer function containing two delays, four multipliers, and four two-input adders with the added restriction that no products of multipliers appear in the expression for the transfer function. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • The parsing program for automatic text-to-speech synthesis developed at the electrotechnical laboratory in 1968

    Page(s): 183 - 188
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (704 KB)  

    The paper describes a parsing program developed at the Electrotechnical Laboratory, Tokyo, Japan, in 1968 for automatic speech synthesis from ordinary English spelling. The parser handles unique problems for a speech production system, especially of phrase-structure analysis in regard to stress and pause assignments. The parsing program consists of a dictionary of about 1500 most frequently used words, a simple syntactic analyzer and a breath-group delimiter. The syntactic analyzer, with the assistance of information stored in the dictionary, divides the sentence into phrases, and assigns pause markers at major syntactic boundaries; the breath-group delimiter decides actual pauses and sentence stress. The output of the parsing program consists of a sequence of phonemes with stress marks and of phrase termination marks. These letters and marks are transformed into vocal tract shapes, duration, and pitch signals in the subsequent part of the synthesis system. The parsing program, written in the PL/I language, consists of about 1900 statements. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Fast convolution with finite field fast transforms

    Page(s): 240
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (144 KB)  

    The fast convolution procedure for processing discrete data requires that a transform of the data and the filter pulse response be formed, followed by the inverse transform of their (complex) product. The finite field fast transform eliminates any roundoff error due to internal multiplication, eliminates truncation of irrational coefficients, and requires only real arithmetic (addition and multiplication). This note develops a realization scheme for such a transform using the Chinese remainder theorem. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • New techniques for automatic speaker verification

    Page(s): 169 - 176
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1064 KB)  

    An interactive automatic speaker verification system has been augmented to include linear prediction parameters in addition to the already existing pitch and intensity analysis of sentence-long utterances. This improved system has been evaluated on a new and enlarged speaker population. A method for selecting optimum speaker-dependent features has been incorporated in this system which significantly improves its performance. The evaluation indicates that verification error rate is approximately 1 percent with respect to casual impostors and 4 percent with respect to well-trained mimics. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Convergence of an algorithm for extra-ripple filters

    Page(s): 244 - 245
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (304 KB)  

    It is shown that the convergence of an algorithm for designing extra-ripple digital filters can be inferred from an earlier paper in the mathematical literature. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Rapid measurement of digital instantaneous frequency

    Page(s): 207 - 222
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1568 KB)  

    A new method is proposed for estimating the frequency domain structure of digital signals which may be characterized as having a narrow-band, rapidly time-varying spectrum. The estimated parameter is termed the digital instantaneous frequency of the input and is defined in a manner similar to that used previously to describe frequency-modulated, continuous-time signals. Instantaneous frequency estimates are derived from a spectral computation based on the use of an adaptive linear prediction filter. The proposed method differs from previous techniques in that the coefficients of this filter are continuously updated as each new input data sample is received using a simple time-domain algorithm, A derivation of the algorithm and its properties are presented. Numerical examples are included which illustrate the properties of the procedure. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Selection of acoustic features for speaker identification

    Page(s): 176 - 182
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (880 KB)  

    The aim of this study was to determine a set of acoustic features in the speech signal that are effective for the identification of a speaker. The investigation examined a large number of theoretically attractive features. The analysis technique of linear prediction was incorporated to examine features that were previously ignored because their measurement was either too time consuming or not easily amenable to automatic measurement. A novel probability of error criterion was used to determine the the relative merits of the features. The experimental data base was collected over a3frac{1}{2}year period and afforded the oportunity to investigate the variation over time of the measurements. The measurements that were found to be the most important were the value of the second resonance (around 1000 Hz) in /n/, the value of the third or fourth resonance (1700-2000 Hz) in /m/ the values of the second, third and fourth formant frequencies in vowels, and the average fundamental frequency of the speaker. A speaker identification experiment using only the best five features was performed. The test data consisted of the multisession data of 11 speakers, and the test data was kept independent of the design data. One error was made in the identification of these speakers for 320 separate identification experiments. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Remarks on the zero-input behavior of second-order digital filters designed with one magnitude truncation quatizer

    Page(s): 240 - 242
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (400 KB)  

    A detailed description is given of the areas in the parameter plane where limit-cycles are possible in a second-order digital filter with one magnitude-truncation quantizer. It is shown that in those points for which limit-cycles can exist the possibility of occurrence is small. A method is described for breaking down these remaining limit-cycles. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • The fast Fourier transform

    Page(s): 245 - 246
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (328 KB)  

    First Page of the Article
    View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • On the power spectrum of the staircase function in linear delta modulation

    Page(s): 162 - 168
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (720 KB)  

    We consider the power spectrum Y(f) of the staircase sequence{y_{r}}in the linear delta modulation (LDM) of a band-limited signal sequence,{x_{r}}. Within the signal band Y(f) approximates the input spectrum X(f) with an accuracy that depends on two parameters: the LDM step-size Δ and the over-sampling F (ratio of sampling rate in LDM to the Nyquist rate for the bandlimited input), or equivalently, the correlation C between adjacent input samples. We demonstrate Y(f) dependencies on Δ and F (or C) using Gauss-Markov and speech inputs in a computer simulation. For speech, we consider the specific problem of preserving, in Y(J), the formant frequencies of a short-term input spectrum X(f). We observe, for example, when F = 9, that input resonances are shifted by amounts not exceeding perceptual limens, in an average sense, if Δ is within an estimated ± 6 dB of a step-size ΔOPT(which minimizes the mean-square-quantization error in the delta modulation of{x_{r}}). The fact that yris a summation of r binary quantities makes the computation of Y(f) much simpler than that of X(f), in general; specifically Y(f) can be computed without any multiply operations. Therefore, in problems where the power spectrum is the desired end result (for example, in the visual monitoring of formant frequencies in speech) Y(f) can provide a useful and simply computed approximation to the input spectrum. With such special applications in mind, we consider the problem of implementing a Y(f) analyzer, and note two specific analyzer configurations. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Applications of the partitioned and modified chirp z-transform to oceanographic measurements

    Page(s): 243 - 244
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (296 KB)  

    The partitioned modified chirp Z-transform (PAM-CZT) is a signal processing technique which aids in the analysis of acoustically propagated ocean noises. The technique utilizes the fast Fourier transform (FFT) to allow real time computation in which long time data are processed in short-time-ordered sequences, thus providing spectral analysis to any frequency resolution. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Digital matched filtering with pipelined floating point fast Fourier transforms (FFT's)

    Page(s): 222 - 234
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1352 KB)  

    A special floating point arithmetic technique for fast Fourier transform (FFT) processors has been developed. The implementation of a high-speed pipeline FFT matched filter using the method employs significantly fewer components than a fixed-point processor with an equivalent performance level. The special floating point process avoids the nonlinear effects of fixed-point processors while achieving performance levels of traditional floating point arithmetic. Computer simulations were used to examine the performance of the system for linear FM pulse compression under a variety of conditions. In a specific case, with 8 bits plus sign quantization of the complex I- and Q-mantissa words in the processor and a 12-stage (4096-coefficient) FFT convolver, the mean square error (MSE) of the sidelobes relative to the peak from a single point target was less than -70 dB. Changes in waveform duration and sampling rate had negligible effect on the error level. This error characteristic can be treated as a computational or self noise, added to the input thermal noise of the radar receiver. Quantization artifacts or noise peaks which occur at levels consistent with the mantissa quantization are below levels (-50 dB for 8 bits plus sign) which would normally cause difficulty in an operational system. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.

Aims & Scope

This Transactions ceased production in 1990. The current retitled publication is IEEE Transactions on Signal Processing.

Full Aims & Scope