This paper describes an 8 kbit/s ACELP speech coder with high performance for both speech and non-speech signals such as background noise. While the traditional waveform matching LPAS structure employed in many existing speech coders provides high quality for speech signals, it has significant performance limitations for, for example, background noise. The coder presented here employs a novel adaptive gain coding technique using energy matching in combination with a traditional waveform matching criterion providing high quality for both speech and background noise. The coder has a basic structure similar to that of the 7.4 kbit/s D-AMPS EFR coder, with a 10 th order LPC, high resolution adaptive codebook and a 4 pulse algebraic codebook. The performance for speech signals is equivalent to or better than that of state-of-the-art 8 kbit/s coders, while for background noise conditions the performance is significantly improved.

Phonetic classification of speech frames allows distinctive quantization and bit allocation schemes suited to the particular class. Separate quantization of the linear predictive coding (LPC) parameters for voiced and unvoiced speech frames is shown to offer useful gains for representing the synthesis filter commonly used in code-excited linear prediction (CELP) and other coders. Subjective test results are reported that determine the required bit rate and accuracy in the two classes of voiced and unvoiced LPC spectra for CELP coding with phonetic classification. It was found, in this context, that unvoiced spectra need 9 b/frame or more whereas voiced spectra need 25 b/frame or more with the quantization schemes used. New spectral distortion criteria needed to assure transparent LPC spectral quantization for each voicing class in CELP coders are presented. Similar subjective test results for speech synthesized from the true residual signal are also presented, leading to some interesting observations on the role of the analysis-by-synthesis structure of CELP. Objective performance assessments based on the spectral distortion measure are also presented. The theoretical distortion-rate function for the spectral distortion measure is estimated for voiced and unvoiced LPC parameters and compared with experimental results obtained with unstructured vector quantization (VQ). These results show a saving of at least 2 b/frame for unvoiced spectra compared to voiced spectra to achieve the same spectral distortion performance.

In this paper we propose a novel technique for vector quantizer design where the reconstruction vectors are given by a linear mapping of a binary block code (LMBC). The LMBC framework provides a relation between the index bits and the reconstruction vectors through mapping properties. We define a framework, show its flexibility, and give optimality conditions. We consider source optimized vector quantization (VQ), where the objective is to directly obtain a VQ with inherent good channel robustness properties. Several instructive theoretical results and properties of the distortion experienced due to channel noise are demonstrated. These results are used to guide the design process. Both optimization algorithms and a block code selection procedure are devised. Experimental results for Gauss-Markov sources show that quantization performance close to an unconstrained VQ is obtained with a short block code which implies a constrained VQ. The resulting VQs have better channel noise robustness than conventional VQs designed with the generalized Lloyd algorithm (GLA) and splitting initialization, even when a post-processing index assignment algorithm is applied to the GLA-based VQ. We have, thus, demonstrated a unique method for direct design resulting in an inherent good index assignment combined with small losses in quantization performance.

In CELP, the use of codebooks with entries with only a few non-zero samples provides high speech quality and facilitates fast computation. With decreasing bit-rate, the intervals between the pulses increase, and the quality of the reconstructed signal begins to suffer from a particular type of artifact, which is strongest for noise-like segments. In this paper we describe experiments which show that the perceived artifacts are mainly concentrated at frequencies above 3 kHz, and this is consistent with our understanding of auditory theory. Our analysis leads to simple strategies to eliminate the artifacts, even at lower bit rates. We describe both a non-adaptive and an adaptive post-processing method to remove the artifacts. The methods are demonstrated to be efficient when used in the ACELP algorithm. A closed-loop method for ACELP is also described.

In speech coding it is important to focus the coding effort on the perceptually important features of the speech signal. This paper describes new quantization techniques which take advantage of current knowledge of human perception in speech coders. The new procedures exploit the frequency-dependent frequency resolution of the human auditory system. The methods are applied to the waveform interpolation (WI) coder, and their effectiveness is confirmed with experimental results. The principles described in the paper are not restricted to the WI coder, but are also applicable to many other speech coding algorithms.

A recent independent survey found a 2.4 kbit/s waveform-interpolation (WI) algorithm to perform better than other state-of-the-art speech coders. However, this coder had a very high level of computational complexity. The introduction of various techniques, including a time-varying waveform sampling rate and a cubic B-spline waveform representation, has reduced the computational complexity by an order of magnitude. The new implementation allows full-duplex real-time operation on a single DSP device and on an average workstation (the latter using non-optimized compiled C source code). The new coder also contains a number of new features which improve the quality of the reconstructed speech signal.

In memoryless quantization, neither the encoder nor the decoder has memory, and quantization noise shaping is not used. We show that, by constraining the parameter dynamics during quantization at the encoder, the performance of speech coders can be enhanced significantly without adding to the delay. The proposed method retains the advantages of memoryless quantization, including channel-error robustness.

In this paper, we address the use of a vector quantizer (VQ) defined as a linear mapping of a block code (LMBC-VQ) in robust LPC spectrum quantization. Split VQ and multistage VQ of LSF's are considered. We demonstrate several properties for the LMBC-VQ including design methods and channel robustness theory. The LMBC-VQ codebooks are, guided by the robustness theory, directly designed with inherent robustness against channel errors. We demonstrate results where the resulting LMBC-VQ's outperform conventional codebooks designed with the generalized Lloyd algorithm in the presence of channel errors even when postprocessing index assignment is applied to the latter. Listening tests confirmed that the increased robustness is significant for the perceptual quality. We show that short block codes can be used to build the codebooks with no or small losses in quantization performance. This is advantageous for channel error robustness, lower memory requirement, and lower complexity. We also describe the structure of split VQ and multistage VQ using the LMBC framework. This allows us to assess the structural constraints and to generalize these VQ schemes, resulting in improved performance.

Variable rate quantization of the linear predictive coding (LPC) parameters based on phonetic classification of the speech frame results in substantial performance gain. Speech frames are classified as unvoiced or voiced and are separately quantized with VQ codebooks designed for each class. Performance results, including listening tests, show that for transparent quality roughly 9 bits is sufficient for unvoiced frames and 24 bits for voiced frames. Test results of LPC quantization are described for a variable rate phonetically segmented CELP coder and for the synthesis of speech from the prediction residual.

Studies the cepstral coefficients as a suitable representation of the linear prediction filter for spectral coding purposes. Spectral coding methods in predictive speech coders are usually evaluated using the spectral distance measure. The average spectral distance combined with a measure of the percentage of spectra with high distortion are used to predict the perceptual quality when quantizing the prediction filter. The authors show that the spectral distance is equivalent to a squared error in the cepstral domain. Methods for spectral quantization using vector quantization of cepstral coefficients are analyzed. Better results than for quantization of line spectrum frequencies are reported for both single-stage VQ at 11-14 bits as well as 2-stage VQ at 18-22 bits. It is concluded that the cepstral coefficients are the right representation for LPC spectral coding purposes.>

The use of line spectral pair (LSP)-frequencies for approaching the performance of a full vector quantizer for speech coding is discussed. An algorithm is presented that attains close to optimal performance if an accumulative constraint is introduced in the parameter encoding. Numerical results are presented from the training of code-books as well as from the evaluation of the quantizer. Some listening tests are discussed.>