Linear Prediction Vocoder Research Articles

Towards a minimally redundant linear predictive vocoder

We consider pitch-excited linear predictive vocoders, and present approaches to achieve minimal transmission rates while maintaining good-quality speech. Our contributions deal with quantization and encoding of predictor parameters before transmission, and their interpolation before synthesis. For purposes of quantization and transmission we have found the so-called log area ratios to be optimal in that they preserve filter stability under quantization, and they possess uniform spectral sensitivity characteristics (Makhoul and Viswanathan, BBN Report No. 2800). The quantized parameters are encoded using (i) a variable-rate scheme that transmits parameters only when speech characteristics have sufficiently changed, and (ii) a variable-wordlength scheme (Huffman code) that uses parameter statistics to code parameters with variable numbers of bits. At the receiver, the synthesizer parameters are reset at the corresponding times at which they were extracted during the analysis. In addition, between every two settings, the parameters are further updated pitch asynchronously by periodic interpolation between the settings. With these features, transmission of good-quality 10-kHz sampled speech is achieved at average bit rates as low as 1650 bits/sec.

Residual excited linear predictive coder

We describe a linear predictive vocoder excited by the speech residual with the transmission bit rate below 8k bits/sec. The features of the residual excited linear predictive (RELP) vocoder system are as follows: (1) no pitch extraction is needed; (2) the bit rate is relatively low; and (3) the system is simple for hardware implementation. The RELP vocoder combines all the attractive concepts of linear predictive coding (LPC), voice excited vocoder (VEV), and adaptive delta modulation (ADM); that is, the system employs LPC for a better spectral representation of the vocal tract, the concept of VEV for bandwidth compression of excitation signal, and ADM for its simplicity of implementation and ability of accurate coding of a low-frequency signal. The LPC residual is generated by a feed-forward LPC analyzer, is low-pass filtered at cutoff frequency 800 Hz, and then is coded by ADM. At the synthesizer the decoded residual is passed through a spectral flattener and its output is finally fed as an excitation signal to the LPC synthesizer to obtain synthesized speech. The quality of synthesized speech is, in our opinion, reasonably good compared with those of other vocoders with the same bit rate. [The study was supported by the Defense Communications Agency.]

A linear prediction vocoder simulation based upon the autocorrelation method

A detailed discussion of the computer simulation of a linear prediction vocoder system is presented. The basic technique used for analysis is the autocorrelation method of linear prediction. New results include modifications to the simplified inverse filter tracking (SIFT) algorithm for more efficient pitch extraction, coding algorithms for low-bit rate transmission, a simplified synthesizer gain calculation, and a bias correction for the synthesizer driving function. Experimental results are presented which illustrate both the capabilities and limitations of linear prediction vocoders.

Adaptive Preprocessing for Linear Predictive Speech Compression Systems

In any vocoder, the process of parameter quantization is crucial for reducing the transmission rate while maintaining the quality of the synthesized speech. We have observed that, for linear predictive vocoders, the short-time spectral dynamic range of the speech signal is the single most important factor that affects the quantization properties of transmission parameters, irrespective of the particular choice of parameters. We describe two preprocessing methods by which the spectral dynamic range is reduced, thereby improving quantization properties. In the first method, adaptive optimal (linear predictive) preemphasis is applied either directly to the speech signal or more easily to the autocorrelation coefficients through convolution. Experiments indicate that an appropriate set of preemphasis filters can be pre-selected; for any speech segment being analyzed, the filter from this set that is closest to the optimal choice can be used for preemphasis. The advantages of such a scheme as well as the effect of first- and second-order preemphasis on quantization and speech quality are discussed. The second method of preprocessing (the SIGMA method) involves multiplying the impulse response of the inverse linear prediction filter by an exponential window. Some preliminary results obtained using this method are reported.

Results of a Study of the Linear Prediction Vocoder

The Linear Prediction Vocoder (LIPREDER) System is an analysis-synthesis speech processing system which efficiently converts speech into digital form at voiceband data rates. The objective of our study was to optimize the LIPREDER system in terms of performance and complexity for a wide selection of talkers and sentences at data rates of 3600 and 7200 b/s. The optimized LIPREDER system produces synthetic speech quality which represents a substantial improvement over previous voice encoders operating at voiceband data rates. Evaluation results indicate that for most talkers and text there is little or no perceptible quality difference between the synthetic speech produced by LIPREDER systems at 3600 and 7200 b/s. Additionally, implementation of the LIPREDER system appears to be technically feasible.