Coded data generation or conversion – Analog to or from digital conversion – Differential encoder and/or decoder
Reexamination Certificate
2002-08-09
2003-11-11
JeanPierre, Peguy (Department: 2819)
Coded data generation or conversion
Analog to or from digital conversion
Differential encoder and/or decoder
C341S155000, C341S120000
Reexamination Certificate
active
06646582
ABSTRACT:
BACKGROUND OF THE INVENTION
This invention relates generally to the error control or management of digital signals as in the playback of digitized audio signals, and pertains more specifically to a method of, and means for, compensating for uncorrectable or uncorrected errors in sigma-delta-modulated audio signals or the like in which each sample of the original analog signal is expressed by one bit of digital data, that is, by either a binary 0 or 1.
The so-called “super audio compact disk,” with its greatly extended range of frequencies compared to that of the more conventional compact disk, has recently been developed by Sony and Philips and introduced to the market the world over. Recorded on the SACD is what is known as the “direct stream digital” (DSD) signal. The DSD is essentially equivalent to delta-sigma (or sigma-delta) modulation, in which each sample of the analog audio signal is translated into either of the binary digits 0 and 1, as will be later explained in more detail. Japanese Unexamined Patent Publication No. 2000-114971 is hereby cited as dealing with sigma-delta modulation.
Whereas each sample of the audio signal is translated into a plurality of, sixteen for example, bits in the more conventional pulse-code modulation, the delta-sigma modulation employs but either of the binary digits 0 and 1 for expressing each sample of the analog signal. Let it be assumed that a stream of audio signal is now both delta-sigma and pulse-code modulated into digital signals that are the same in the total number of bits. Then the sampling frequency of delta-sigma modulation can be made very much, sixteen times for example, higher than that of pulse-code modulation. A higher-resolution analog-to-digital conversion is therefore possible by delta-sigma modulation without an increase in the total number of bits for a given signal volume.
Read errors are, however, almost unavoidable in the playback of the delta-sigma modulated SACD as in that of the regular CD but, again as with the regular CD, detectable by error-detecting codes such as the familiar cyclic redundancy checks. Error-correcting codes are also available for correcting any errors that are correctable, in delta-sigma modulated data. Uncorrectable errors have been circumvented. As far as the applicant is aware, the only known solution for such errors in the SACD has been to cause the analog output signal to fade out immediately before each error location and to fade in just afterward.
In the case of pulse-code modulated data, the datum missing from an uncorrectable error location is estimated to be the same as the preceding datum, and this preceding datum is interpolated in the error location. It might be contemplated to apply this interpolation method to delta-sigma modulated audio signals. This remedy is totally unacceptable, as it would give rise to audible noise instead of the fading method discussed above. For, in delta-sigma modulation, the logical 0 or 1 immediately before the error does not reflect the amplitude of the analog signal recorded. The trouble would occur if the datum before the error was 1. Should the uncorrectable error datum be estimated to be 1 on the basis of this preceding datum, the result might be the production of a formidable noise. More will be said later about this noise production with reference to the attached drawings.
SUMMARY OF THE INVENTION
The present invention seeks to compensate, by interpolation and without fading, for read or transmission errors that occur as in the play-back of delta-sigma or like pulse-density modulated data, just as noiselessly as the familiar error compensation of pulse-code modulated data by interpolation.
Briefly, the present invention may be summarized as a noise control system for use on a transmission path for a digital signal that is constituted of primary data in which each sample of an analog signal is expressed by one bit of information, and of error correction data for use in correction of error in the primary data. The noise control system comprises an error-detecting and -correcting circuit for detecting and, if possible, correcting errors in the primary data of the digital signal with use of the error correction data. The error-detecting and -correcting circuit has a first output for delivery of the digital signal, and a second output for production of an uncorrectable error signal indicative of the presence or absence of an uncorrectable or uncorrected error in or from the digital signal. Also included is an interpolation data generator for generating a predefined interpolation datum for insertion in each uncorrectable or uncorrected error location in the digital signal. An interpolation circuit is connected to both the error-detecting and -correcting circuit and the interpolation data generator for inserting, when the uncorrectable error signal from the second output of the error-detecting and -correcting circuit indicates the presence of an uncorrectable or uncorrected error, the predefined interpolation datum in the location of the uncorrectable or uncorrected error in the digital signal issuing from the first output of the error-detecting and -correcting circuit.
The invention further specifies that the interpolation datum for insertion in each uncorrectable or uncorrected error location be constituted of a plurality of binary 0s and a plurality of binary 1s, and that the 0s and 1s in each interpolation datum be equal in number. Furthermore, the 0s and 1s in each interpolation datum should be so arranged as to represent a frequency higher than the upper limit of audio frequencies, which commonly is believed to be 20 kilohertz. An example of interpolation datum meeting all these requirements is [1010] or an integral multiple thereof.
When the digital signal with an uncorrectable or uncorrected error compensated for by the teachings of the invention is subsequently converted into an analog signal, the analog equivalent of the interpolation datum maintains on the average the signal value preceding the error location. Any uncorrectable or uncorrected errors in delta-sigma or like pulse-density modulated signals can be compensated for by the method of this invention in essentially the same way as errors in pulse-code modulated signals are conventionally compensated for by interpolation of the preceding values. This remedy is obviously more agreeable to the ear than the conventional fading and certainly far less obnoxious than the interpolation of previous values.
The above and other objects, features and advantages of this invention will become more apparent, and the invention itself will best be understood, from a study of the following detailed description and appended claims, with reference had to the attached drawings showing a preferred embodiment of the invention together with the associated prior art.
REFERENCES:
patent: 5014274 (1991-05-01), Higurashi et al.
patent: 5774481 (1998-06-01), Meaney et al.
patent: 5936561 (1999-08-01), Lee
patent: 00/114971 (2000-04-01), None
Jean-Pierre Peguy
TEAC Corporation
Woodcock & Washburn LLP
LandOfFree
Error compensation for sigma-delta modulated digital signals... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Error compensation for sigma-delta modulated digital signals..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Error compensation for sigma-delta modulated digital signals... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3161088