Electrical audio signal processing systems and devices – One-way audio signal program distribution – Public address system
Patent
1988-07-21
1990-11-20
Shaw, Dale M.
Electrical audio signal processing systems and devices
One-way audio signal program distribution
Public address system
G10L 702
Patent
active
049724843
DESCRIPTION:
BRIEF SUMMARY
BACKGROUND OF THE INVENTION
The invention relates to a method of transmitting or storing digitalized audio signals, wherein an audio signal is digitally represented by a plurality of spectral quantized components.
In connection with the transmission cf digital audio signals it is known from the following publications: ("Journal of Audio Engineering Society," November, 1979, Volume 27, No. 11, pages 855-865; "The Bell System Technical Journal," September, 1981, pages 1633-1653; and "IEEE Transactions on Acoustics, Speech and Signal Processing, " Vol. ASSP-30, No. 5, October 1982, pages 751-765), to subdivide a broadband digital audio signal by means of a QMF (quadrature mirror filter) filter bank into a number of sub-band signals according to a linear quantization and to subject the resulting sub-band signals to a data reduction, for example, by means of adaptive PCM or DPCM coding.
It is also known from German Patent No. 3,440,613 to select the quantization of the useful information within each sub-band signal so that the quantizing noise is just covered by the useful information in the same sub-band, which can also result in data reduction. The data reduction factor realizable with the prior art method has approximately a value of four, i.e. the information flow of a high-quality digital audio signal is reduced from about 500 kbit/s to about 125 kbit/s without any subjective reduction in quality occurring.
To realize an even further data reduction, it is also known from (European Patent No. 0,193,143, DE-OS 3,506,912 and "Rundfunktechnische Mitteilungen" [Radio Engineering News], Volume 30 (1986), No. 3, pages 117-123), to perform a spectral analysis of the broadband audio signal with the aid of a discrete Fourier transformation (for example, by means of a fast-Fourier transformation) and to code certain relevant spectral values within different frequency groups according to magnitude and phase so that greater data reduction is realized by considering masking properties of the human auditory system defined by masking thresholds and according to different quality criteria.
The analysis time window required for the Fourier transformation, however, is about 25 ms. This value constitutes a compromise to meet the requirements, on the one hand, for spectral resolution and, on the other hand, for temporal resolution of the human auditory system. The spectral resolution that can be realized with this analysis time window is merely 40 Hz so that, in the range of low frequencies, where the frequency group width of the human auditory system is about 100 Hz, only two spectral values can be transmitted. The resulting sidebands therefore lie in adjacent frequency groups so that perceptible reductions in quality cannot be excluded. On the other hand, the 25 ms analysis time window selected as a compromise is too long for the time resolution of the human auditory system. Since for pulses containing useful signals, this inaccuracy in the time domain leads to noticeable distortions. The amplitude values of the spectral components preceding in time must be raised in order to reduce distortion, but this does not lead to the desired success in all cases. Moreover, in digital audio studio technology, block lengths of about 5 ms must not be exceeded so that inaudible cuts are possible when digitalized audio signals are edited. Additionally, processor expenditures, particularly in the receiver, for the retransformation of signals in the high frequency range transformed at the transmitter are unnecessarily high since the consideration of psychoacoustic criteria occurs only by frequency groups.
Additionally, in the prior art method discussed last, the recovery of spectral values in the receiver according to magnitude and phase and the inverse Fourier transformation in the receiver requires the transmission of secondary information. The secondary information represents a relatively high percentage of the entire net information flow and requires particularly effective error protection which correspondingly increases the flow of inf
REFERENCES:
patent: 4538234 (1985-08-01), Honda et al.
patent: 4569075 (1986-02-01), Nussbaumer
patent: 4831636 (1989-05-01), Taniguchi et al.
Crochiere, "Sub-Band Coding", The Bell System Technical Journal, vol. 60, No. 7, 9/81, pp. 1633-1653.
Journal of the Audio Engineering Society, 1979 Nov., vol. 27, No. 11 "An All-Digital `Commentary Grade` Subband Coder", J. D. Johnston and R. E. Crochiere, pp. 855-865.
IEEE Transactions on Acoustics, Speech, and Signal Processing, vol. ASSP-30, No. 5, Oct. 1982.
"Subband Coder Design Incorporating Recursive Quadrature Filters and Optimum ADPCM Coders", Thomas P. Barnwell, III, pp. 751-765.
RTM, vol. 30, No. 3, May/Jun. 1986, Ein Verfahren Zur Datenredukton Bei Diytalen Audiosignalen Unter Ausnutzung Psychookustischer Phanomene, von Derlefkrahe, pp. 117-123.
8089 IEEE Transactions on Communications, Com-33(1985), Oct. No. 10, New York, U.S.A. "Digital Speech Interpolation for Variable Rate Coders with Application to Subband Coding", Kuel Young Kou et al., pp. 1100-1108.
Link Martin
Stoll Gerhard
Theile Gunther
Bayerische Rundfunkwerbung GmbH
Merecki John A.
Shaw Dale M.
LandOfFree
Method of transmitting or storing masked sub-band coded audio si does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method of transmitting or storing masked sub-band coded audio si, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method of transmitting or storing masked sub-band coded audio si will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-460471