Linear Predictive Synthesis Research Articles

Sound Analysis and Synthesis for Audio Simulation System of Flight Simulator

Through simulating the feelings of vision, hearing, motion and force during flight, the flight simulator can build a realistic flight environment on the ground. Therefore, Flight simulators have an important role in pilot training field. Audio simulation system is a key component of the flight simulator, it has direct impact on the realism and immersion of flight simulation. In this paper, we introduced the development procedure of audio simulation system shortly. Software implementation and its key development technology were mainly focused for expressing the necessity of extracting sound sources from original cockpit recordings. Based on the development method and practical recording conditions, we dicussed sound short-time Fourier analysis and synthesis technology, and sound linear prediction analysis and synthesis technology in detail. The objective of using these technologies was to extract sound sources from original recordings and this was also the preparation step of sound simulation.

Microphone array signal enhancement

A system and method facilitating signal enhancement utilizing an adaptive filter is provided. The invention includes an adaptive filter that filters an input based upon a plurality of adaptive coefficients and modifies the adaptive coefficients based on a feedback output. A feedback component provides the feedback output based, at least in part, upon a non-linear function of the acoustic reverberation reduced output. Optionally, the system can further include a linear prediction (LP) analyzer and/or a LP synthesis filter. The system can enhance signal(s), for example, to improve the quality of speech that is acquired by a microphone by reducing reverberation. The system utilizes, at least in part, the principle that certain characteristics of reverberated speech are measurably different from corresponding characteristics of clean speech. The system can employ a filter technology (e.g., reverberation reducing) based on a non-linear function, for example, the kurtosis metric.

Method for coding speech and music signals

The present invention provides a transform coding method efficient for music signals that is suitable for use in a hybrid codec, whereby a common Linear Predictive (LP) synthesis filter is employed for both speech and music signals. The LP synthesis filter switches between a speech excitation generator and a transform excitation generator, in accordance with the coding of a speech or music signal, respectively. For coding speech signals, the conventional CELP technique may be used, while a novel asymmetrical overlap-add transform technique is applied for coding music signals. In performing the common LP synthesis filtering, interpolation of the LP coefficients is conducted for signals in overlap-add operation regions. The invention enables smooth transitions when the decoder switches between speech and music decoding modes.

Enhancing the naturalness of synthesized speech by using the random fractalness of vowel source signals

Random fractal character observed in vowel source signals obtained by linear predictive analysis was investigated using a multiresolution analysis method based on Schauder expansion. With the results of the Schauder analysis, a technique for enhancing the naturalness of synthesized vowels was newly developed. Synthesized vowels produced by a linear predictive synthesis method tend to exhibit a buzzerlike unnatural quality due to oversimplifying the source signals as an impulse train. This study shows that such degraded voice quality can be improved by (1) removing the discontinuities in the impulse train, and (2) restoring the reduced power in the high-frequency region by employing random fractal interpolation technique devised in this study. Psychoacoustic experiments evaluated the validity of the proposed technique. © 2000 Scripta Technica, Electron Comm Jpn Pt 3, 84(1): 11–20, 2001

Wavelet analysis used in text-to-speech synthesis

This brief describes the use of wavelet analysis in the development of a Japanese text-to-speech (TTS) system for personal computers. The quality of synthesized speech is one of the most important features of any TTS system. Synthesis methods which are based on manipulation of the speech signal spectrum (e.g,, linear predictive coding synthesis and formant synthesis) produce comprehensible but unnatural sounding output. The lack of naturalness commonly associated with these methods results from the use of oversimplified speech models, small synthesis unit inventories, and poor handling of text parsing for prosody control. We developed four new technologies to overcome these difficulties and improve the quality of output from TTS systems: accurate pitch mark determination by wavelet analysis, speech waveform generation using a modified time domain pitch synchronous overlap-add method, speech synthesis unit selection using a context dependent clustering method, and efficient prosody control using a 3-phrase parser. All four technologies will be described; however, those which rely on wavelet techniques will be emphasized.

A new algorithm for adaptive lattice filter and its application in the speech linear predictive synthesis

In this paper, an adaptive line spectral pair filter is derived from an adaptive lattice filter. A least-mean-square(LMS) type adaptive algorithm used to calculate directly the line spectral pair (LSP) coefficients on a stage-by-stage basis is presented. Experimental results show that the algorithm has higher convergence rate and lower misadjustment as compared with the other algorithms. The LSP coefficients calculated by the algorithm have been used to carry out speech linear predictive synthesis, resulting in better results than PARCOR coefficients.

Speech coding and decoding methods using adaptive and random code books

An excitation vector of the previous frame stored in an adaptive codebook is cut out with a selected pitch period. The excitation vector thus cut out is repeated until one frame is formed, by which a periodic component codevector is generated. An optimum pitch period is searched for so that distortion of a reconstructed speech obtained by exciting a linear predictive synthesis filter with the periodic component codevector is minimized. Thereafter, a random codevector selected from a random codebook is cut out with the optimum pitch period and is repeated until one frame is formed, by which a repetitious random codevector is generated. The random codebook is searched for a random codevector which minimizes the distortion of the reconstructed speech which is provided by exciting the synthesis filter with the repetitious random codevector.

Female alaryngeal speech enhancement for improved speaker identification using linear predictive synthesis

Linear predictive synthesis was applied to female alaryngeal speech as an enhancement technique to reduce gender ambiguity during speaker identification. Esophageal speech, a form of alaryngeal speech resulting from the removal of the larynx due to cancer, perturbs the theoretical acoustic models for laryngeal speech. This form of pathological speech also changes the expected spectral characteristics and affects perception. The female esophageal speaker’s low fundamental frequency often causes her to be identified as a male speaker over telephone lines. The errors made in speaker identification can be reduced by obtaining vocal tract parameters of the aperiodic, esophageal speech input from a linear predictive coding (LPC) model and resynthesizing the LPC prediction coefficients using an appropriate pitch period that is better suited for female voice. This enhancement technique was implemented using the TMS320C30 floating point digital signal processor from Texas Instruments. This implementation supports the potential of a portable processor that enhances the perceptual quality of the female alaryngeal voice and improves the input speech signal used by both human listeners and automatic speech and speaker identification systems.

Apparatus for linear predictive coding and decoding of speech using residual wave form time‐acess compression

A speech coding and decoding apparatus for use in linear predictive coding of a speech signal includes coding apparatus having a pitch analyzing element for separating an analysis frame into one or more blocks and calculating the strength of correlativity between pitch periods of a residual waveform in each block. A residual partially compresing element compresses the time axis of the residual waveform in a block having a high correlativity strength. A residual quantizing element quantizes the compressed residual waveform while preferentially allotting quantization allotment bits to the compressed portion of the residual waveform. Decoding apparatus includes a residual inverse quantizing element for inversely quantizing the residual waveform by the same bit allotment. A residual partially expanding element expands the compressed portion of the residual waveform to its original length and supplies it to a linear predictive synthesis filter.

Voice coding process and device for implementing said process

The voice signal is analyzed to derive therefrom a low frequency base band signal, linear prediction coefficients and high frequency (HF) descriptors. Said HF descriptors include HF energy indications as well as indications relative to the phase shift between the low frequency and the high frequency band. Said HF descriptors are used during the voice synthesis operation to provide an inphase HF bandwidth component to be added to the base band prior to be used for driving a linear prediction synthesis filter tuned using said linear prediction parameters.

Linear autoregressive modeling of unvoiced speech

Recent interest in nonlinear prediction of speech has brought up the need to assess the performance limitations of linear speech models. While nonlinearity is essential in the sound production mechanisms this need not be reflected in a speech signal model. A linear model may produce the same perceptual effect for a human listener as the nonlinear one. The validity of a linear autoregressive model has been tested for the unvoiced segments of speech. In this experiment, a linear prediction residual is first computed from the speech signal. For unvoiced segments, the residual is replaced by white Gaussian noise with its energy matched every 5 ms. The residual is left unmodified for voiced segments. The linear prediction synthesis filter is excited with the composite residual signal to reconstruct speech. The resulting speech quality has been evaluated in a formal MOS test where 32 listeners rated speech material from 12 speakers uttering 3 sentences each. The present approach yielded an MOS of 4.0. In the same test, 16-bit linear PCM obtained a score of 4.1. In conclusion, a linear model with random excitation is sufficient to reproduce unvoiced speech with very high perceptual quality. A demonstration tape will be played during the presentation.

Speech synthesis using partial syllables

This article describes a method for synthesizing Arabic speech by using segments (synthesis units) which are parts of syllables. The method has several advantages. It is inherently capable of dealing with most phonetic variations of the sounds of modern standard Arabic; it permits speech to be synthesized using simple concatenation rules and synthesis techniques such as direct waveform concatenation of the synthesis units; and it allows the realization of prosodic features on segments because the segments are derived from Arabic syllables. Most intersyllabic coarticulations cannot be treated directly by this method. This has prompted a phonetic study of Arabic speech following the present method; we have investigated contextual variations of Arabic sounds and related these variations to the synthesis units. This enabled us to define a small number of allophones of the synthesis units which when included in the text-to-speech transcription process improved the quality of the synthesized speech. The segmentation of the Arabic syllables into synthesis units, on both phonetic and parametric levels, is similar to standard demisyllabic segmentation used by other speech synthesis researchers. However, it differs in one important respect which we have investigated. Synthesized speech is produced by two synthesis techniques: time-domain waveform concatenation and linear predictive (LPC) techniques. Time-domain waveform concatenation is used because it is simple and easy to carry out and is attractive since the juncture points between any contiguous synthesis units are stable. It involves direct concatenation of the digital data that represent the synthesis units. LPC synthesis is used because it eases the treatment of certain distortions which are introduced by the synthesis process. It is a parametric method that allows certain suprasegmental aspects of speech to be modified. This results in memory saving because it allows us to compare speech produced by the two synthesis techniques. The types of distortion introduced by concatenating the synthesis units are studied. It is shown how the stability of the juncture points between contiguous units and the phonological properties of Arabic syllables and speech make it possible to avoid these problems. This can be achieved by using simple synthesis techniques such as direct waveform concatenation and shows how problems can be avoided by using a parametric synthesis technique such as LPC. To demonstrate the suitability of the present method for the synthesis of Arabic speech, the method has been implemented as a non-real-time text-to-speech system on a personal computer (PC) and the intelligibility of the synthesized speech has been established by conducting perception tests.

A technique for using multipulse linear predictive speech synthesis in text-to-speech type systems

A scheme for alteration of pitch and timing in multipulse linear predictive synthesized speech is presented. The method is a hybrid of waveform coding and linear prediction. It attempts to combine the high-quality sound units available in waveform techniques with the ability to alter prosody available from linear prediction.

Multipulse linear predictive synthesis in text‐to‐speech systems

Multipulse linear prediction is a coding technique that can provide extremely high quality speech synthesis [B. S. Atal and J. R. Remde, IEEE Proc. ICASSP82, 614–617 (1982)]. It is therefore of interest to examine whether the technique can be used to provide correspondingly good quality in a text‐to‐speech system. In such a system a prosodic contour is imposed on a set of concentrated speech units. Various speech units (e.g., words and diphones) have been tried. First results suggest that the word is the appropriate unit to be used in such a system. The techniques for pitch and timing alteration, speech unit concatention, and the effects on the resulting synthetic speech will be discussed and demonstrated.

Perception of static and dynamic acoustic cues to place of articulation in initial stop consonants.

Two recent accounts of the acoustic cues which specify place of articulation in syllable-initial stop consonants claim that they are located in the initial portions of the CV waveform and are context-free. Stevens and Blumstein [J. Acoust. Soc. Am. 64, 1358-1368 (1978)] have described the perceptually relevant spectral properties of these cues as static, while Kewley-Port [J. Acoust. Soc. Am. 73, 322-335 (1983)] describes these cues as dynamic. Three perceptual experiments were conducted to test predictions derived from these accounts. Experiment 1 confirmed that acoustic cues for place of articulation are located in the initial 20-40 ms of natural stop-vowel syllables. Next, short synthetic CV's modeled after natural syllables were generated using either a digital, parallel-resonance synthesizer in experiment 2 or linear prediction synthesis in experiment 3. One set of synthetic stimuli preserved the static spectral properties proposed by Stevens and Blumstein. Another set of synthetic stimuli preserved the dynamic properties suggested by Kewley-Port. Listeners in both experiments identified place of articulation significantly better from stimuli which preserved dynamic acoustic properties than from those based on static onset spectra. Evidently, the dynamic structure of the initial stop-vowel articulatory gesture can be preserved in context-free acoustic cues which listeners use to identify place of articulation.

A Compact, Flexible LPC Vocoder Based on a Commercial Signal Processing Microcomputer

A very small, flexible, high-quality, full-duplex 2.4-kbit/s linear predictive vocoder has been implemented with commercially available integrated circuits. This fully digital realization is based on a distributed signal processing architecture employing three Nippon Electric Company (NEC) µPD7720 signal processing interface (SPI) single-chip microcomputers. One SPI implements the LPC analyzer, a second implements the Gold pitch and voicing decision algorithm, white the third µPD7720 implements the excitation generator and synthesizer. An Intel 8085-based 8-bit microcomputer is used for data transfer, control and multiplexing functions, and communications with the host terminal. The LPC chip set achieves high flexibility by accepting run time initialization options from the Intel 8085. These parameters include choice of linear predictive model (<= 15), analysis and synthesis frame size, and speech sampling frequency, as well as choice of speech input and output coding formats (linear or µ-255 law) and choice of analog or digjtal pre- and deemphasis. A total of 16 integrated circuits is used in the LPC vocoder with a power disipation of 5.5 W and occupying 18 in/sup 2/ of circuit area.

Electronic speech synthesis

Discusses the developments of electronic speech synthesizers due to the advances in microelectronic devices and due to novel signal processing techniques. The technique of vocal tract modeling is first described and then the following two methods of speech synthesis are dealt with in some detail: a method based on linear predictive coding and speech synthesis by rule

A note on complexity reduction for linear predictive speech synthesis

Most linear predictive speech synthesis (LPSS) algorithms employ a deemphasis filter with a fixed transfer function subsequent to the time varying LPSS filter. The excitation signal for the LPSS filter is usually a stored waveform sequence, repeated at the pitch period, for voiced sounds or a noise sequence for Unvoiced sounds. In either case the excitation signal is approximately white. A computational savings may be achieved by using an excitation signal whose spectrum reflects the frequency response of the deemphasis filter. A procedure is described for generating such an excitation signal and formal subjective intelligibility test results are presented to demonstrate that no performance degradation occurs when this method is employed.