Manfred R. Schroeder

Code-excited linear prediction(CELP): High-quality speech at very low bit rates

We describe in this paper a code-excited linear predictive coder in which the optimum innovation sequence is selected from a code book of stored sequences to optimize a given fidelity criterion. Each sample of the innovation sequence is filtered sequentially through two time-varying linear recursive filters, one with a long-delay (related to pitch period) predictor in the feedback loop and the other with a short-delay predictor (related to spectral envelope) in the feedback loop. We code speech, sampled at 8 kHz, in blocks of 5-msec duration. Each block consisting of 40 samples is produced from one of 1024 possible innovation sequences. The bit rate for the innovation sequence is thus 1/4 bit per sample. We compare in this paper several different random and deterministic code books for their effectiveness in providing the optimum innovation sequence in each block. Our results indicate that a random code book has a slight speech quality advantage at low bit rates. Examples of speech produced by the above method will be played at the conference.

New Method of Measuring Reverberation Time

A new method of measuring reverberation time is described. The method uses tone bursts (or filtered pistol shots) to excite the enclosure. A simple integral over the tone‐burst response of the enclosure yields, in a single measurement, the ensemble average of the decay curves that would be obtained with bandpass‐filtered noise as an excitation signal. The smooth decay curves resulting from the new method improve the accuracy of reverberation‐time measurements and facilitate the detection of nonexponential decays.

Optimizing digital speech coders by exploiting masking properties of the human ear

In any speech coding system that adds noise to the speech signal, the primary goal should not be to reduce the noise power as much as possible, but to make the noise inaudible or to minimize its subjective loudness. ’’Hiding’’ the noise under the signal spectrum is feasible because of human auditory masking: sounds whose spectrum falls near the masking threshold of another sound are either completely masked by the other sound or reduced in loudness. In speech coding applications, the ’’other sound’’ is, of course, the speech signal itself. In this paper we report new results of masking and loudness reduction of noise and describe the design principles of speech coding systems exploiting auditory masking.

Binaural dissimilarity and optimum ceilings for concert halls: More lateral sound diffusion

Ceiling shapes for concert halls are proposed that, on the basis of prior extensive subjective evaluations, should result in high listener preference of the hall’s acpistoca; response to music. These shapes are based on the premise that as little as possible of the early sound energy should arrive at a listener’s ears in the ’’median’’ plane (the vertical symmetry plane through the listener). While this goal is inherently approached in old‐style, high‐and‐narrow halls, its realization in modern, low‐ceiling halls requires special ceiling shapes and surface structures to keep early, median‐plane sound away from the listener’s ears.

Improved quantizer for adaptive predictive coding of speech signals at low bit rates

Adaptive predictive coding of speech signals at bit rates lower than 10 kbits/sec often requires the use of 2-level (1 bit) quantization of the samples of the prediction residual. Such a coarse quantization of the prediction residual can produce audible quantizing noise in the reproduced speech signal at the receiver. This paper describes a new method of quantization for improving the speech quality. The improvement is obtained by center clipping the prediction residual and by fine quantization of the high-amplitude portions of the prediction residual. The threshold of center clipping is adjusted to provide encoding of the prediction residual at a specified bit rate. This method of quantization not only improves the speech quality by accurate quantization of the prediction residual when its amplitude is large but also allows encoding of the prediction residual at bit rates below 1 bit/sample.

Vocoders: Analysis and synthesis of speech

Techniques for analysis and synthesis of speech signals are reviewed with emphasis on vocoders and related devices for more efficient transmission and storage of speech. Selected applications of speech coding methods as sensory aids to the handicapped are described.

Model for mechanical to neural transduction in the auditory receptor

We describe a model for transduction of displacement of the basilar membrane to activity of auditory nerve fibers. The model is physiologically oriented. Quanta (“vesicles”) are added to a population inside an urn (hair cell) at a fixed average rate. They are removed from the population with a probability proportional to the number of quanta in the population and related in a simple manner to the instantaneous amplitude of the input signal (displacement). The removal of a quantum results in an event (action potential) with a probability related to time elapsed since the preceding event (refractory period). Apart from refractoriness, the model is completely specified by three parameters which determine spontaneous and maximum firing rates and the time constant of recovery after intense stimulation. The model is more compatible with observed activity of the auditory nerve than threshold‐crossing models. An important feature of auditory‐nerve data reproduced by the model is the level normalization observed at moderate to high stimulus intensities. The model reproduces features of period, PST, and interval histograms in response to pure and complex tones, tone bursts, and noise. We present results both from analysis and from computer simulation of the model.

The ‘‘Schroeder frequency’’ revisited

It is noted that the cross‐over frequency that marks the transition from individual resonances of a multimode system to overlapping normal modes, when expressed as a cross‐over wavelength, equals—within a numerical constant—the diffuse‐field distance in both three‐ and two‐dimensional resonators.

Phase effects in masking related to dispersion in the inner ear

Phase effects in masking experiments using multitone maskers are usually associated with strong variations in the masker envelope. In this article, psychoacoustic experiments with such maskers that lead to phase-dependent threshold variations of up to 20 dB, although the phase transformation leaves the envelope unchanged, are described. However, after filtering the maskers with a realistic basilar membrane model, the envelopes are different owing to the models phase-dispersive properties. Comparison of model outputs with the experimental results reveals a strong correlation between the two for a wide range of parameters, provided one makes the additional assumption that the ear has a minimum integration time of a few milliseconds.

Direct (nonrecursive) relations between cepstrum and predictor coefficients

Direct, i.e., nonrecursive, relations are derived for the cepstrum in terms of the predictor coefficients and vice versa. Connections with algebraic roots, symmetric functions, statistical moments, and cumulants are pointed out. Some implications for pitch detection are also discussed.