Method of transmitting or storing digitalized, multi-channel audio signals

• US 5,682,461 A
• Filed: 11/21/1994
• Issued: 10/28/1997
• Est. Priority Date: 03/24/1992
• Status: Expired due to Fees

• Current Assignee: Institut Fur Rundfunktechnik GmbH
• Original Assignee: Institut Fur Rundfunktechnik GmbH
• Inventors: Silzle, Andreas, Theile, Guenther, Stoll, Gerhard, Link, Martin
• Primary Examiner(s): Safourek, Benedict V.
• Assistant Examiner(s): NGO, RICKY QUOC
• Application Number:
  US08/307,665
• Time in Patent Office: 1,072 Days
• Field of Search: 395/2.09, 395/2.1, 395/2.14, 395/2.38, 395/2.39, 395/2.12, 395/2.15, 381/36-37, 381/39, 381/41-50, 381/35
• US Class Current: 704/205
• CPC Class Codes: H04B 1/665 using psychoacoustic properties of the ear, e.g. masking effect
• CPC Class Codes: H04H 20/88 Stereophonic broadcast systems multiplex system in general H04J

Abstract

In order to transmit or to store digitalized, multi-channel audio signals which are digitally represented by a plurality of spectral subband signals, at an encoder, subband signals of different audio channels but having the same frequency position are combined across channels according to a dynamic control signal. This control signal is derived from several audio channels by an audio signal analysis based on a psycho-acoustic binaural model. At a decoder, the subband signals of several audio channels but having the same frequency position are dissociated across channels according to a control value derived from the dynamic control signal, transmitted or stored therewith.

Claims

1. A method of transmitting or storing digitalized, multi-channel audio signals, comprising the steps of:

(a) encoding the multi-channel audio signals, step (a) including

(1) for each audio channel, generating a number of spectral subband signals;

(2) analyzing the subband signals to obtain

a dynamic control signal which is oriented toward a binaural, psychoacoustic model,

combiner mode information which is derived from the dynamic control signal,

bit allocation information, and

for each subband signal, a scale factor that classes a peak value of the respective subband signal within a certain time block;

(3) combining the subband signals of different audio channels, but which have the same frequency position, to obtain combined subband signals, step (a-3) being conducted in an interchannel manner according to the dynamic control signal;

(4) normalizing the combined subband signals using the scale factors to obtain normalized subband signals; and

(5) quantizing the normalized subband signals to obtain encoded subband signals in accordance with a channel-specific, intra-channel perceptual threshold of human hearing, step (a-5) being controlled by the bit allocation information;

(b) transmitting or storing the encoded subband signals, the bit allocation information, the scale factors, and the combiner mode information; and

(c) decoding and processing the encoded subband signals, step (c) including

(1) decoding the encoded subband signals according to the bit allocation information to obtain decoded subband signals;

(2) decombining the decoded subband signals of different audio channels, but the same frequency position, to obtain decombined subband signals, step (c-2) being conducted as a function of the combiner mode information;

(3) denormalizing the decombined subband signals according to the scale factors, to obtain denormalized subband signals; and

(4) generating a broadband digital audio signal from the denormalized subband signals.

2. A method according to claim 1, wherein step (a) comprises effecting puncture coding with the aid of at least one of the dynamic control signal and the bit allocation information, so that encoded subband signals for relevant subbands need not be transmitted or stored in step (b), wherein step (c) comprises substituting the non-transmitted or non-stored encoded subband signals in the relevant subbands by transmitted or stored encoded subband signals in the subbands of other audio channels, but of the same frequency position, according to the combiner mode information, and wherein the substituted encoded subband signals are adapted with respect to their level to original level conditions of the relevant subbands.

3. A method according to claim 2, further comprising the step of calculating weighting factors, with which the substituted encoded subband signals are weighted, from the scale factors for the level adaptation.

4. A method according to claim 2, wherein the puncture coding is effected as a function of a determination of a sequence of a transient and a stationary signal state of the same audio channel or adjacent audio channels.

5. A method according to claim 2, wherein the puncture coding occurs more frequently in subbands of higher frequency position than in subbands of lower frequency position.

6. A method according to claim 2, wherein an entire audio signal is comprised of the multi-channel audio signals, and the puncture coding occurs more frequently in subbands in which the signal energy is relatively low with respect to the signal energy of the entire audio signal.

7. A method according to claim 2, wherein the puncture coding is effected with the utilization of the inertia effect of human hearing.

8. A method according to claim 2, wherein the puncture coding is effected with the utilization of the psychoacoustic masking of adjacent audio channels.

9. A method according to claim 2, wherein the puncture coding is effected with the utilization of redundancy in subbands of the same frequency position of adjacent audio channels, and wherein during step (c) the substituted encoded subband signals are only substituted in their original audio channel during recombination.

10. A method according to claim 2, wherein the step (a) further comprises the step of determining whether alias distortions exists, and wherein the puncture coding is only effected when alias distortions are irrelevant with respect to properties of human hearing.

11. A method according to claim 2, wherein puncture coding is not effected when a negative correlation exists between subband signals of the same frequency position.

12. A method of transmitting or storing digitalized audio signals for a plurality of intercorrelated audio channels, the intercorrelated audio channels including a left channel, a right channel, and at least one further channel, said method comprising the steps of:

(a) encoding the audio signals, step (a) including

(1) for each audio channel, generating a number of spectral subband signals;

(2) analyzing the subband signals to obtain

bit allocation information which is derived from a common perceptual threshold, the common perceptual threshold being formed from all of the audio channels, and

for each subband signal, a scale factor that classes a peak value of the respective subband signal within a certain time block;

(3) normalizing signals which are derived from the subband signals, using the scale factors, to obtain normalized subband signals; and

(4) quantizing the normalized subband signals to obtain encoded subband signals in accordance with a channel-specific, intra-channel perceptual threshold of human hearing, step (a-4) being controlled by the bit allocation information;

(b) transmitting or storing the encoded subband signals, the bit allocation information, and the scale factors; and

(c) decoding and processing the encoded subband signals, step (c) including

(1) decoding the encoded subband signals as a function of the bit allocation information to obtain decoded subband signals;

(2) denormalizing signals derived from the decoded subband signals according to the scale factors, to obtain denormalized subband signals; and

(3) generating a broadband digital audio signal from the denormalized subband signals.

13. A method according to claim 12, wherein the common perceptual threshold is calculated with consideration of a masking difference in playback.

14. A method according to claim 12, wherein the at least one further channel includes a center channel, a left surround channel, and a right surround channel.

15. A method of transmitting or storing digitalized, multi-channel audio signals, comprising the steps of:

(a) encoding the multi-channel audio signals, step (a) including;

(1) for each audio channel, generating a number of spectral subband signals;

(2) analyzing the subband signals to obtain

bit allocation information, and

for each subband signal, a scale factor that classes a peak value of there spective subband signal within a certain time block;

(3) combining the subband signals of different audio channels, but which have the same frequency position, to obtain combined subband signals, step (a-3) being conducted as a function of an audio representation format of the audio channel corresponding to a system configuration;

(4) normalizing the combined subband signals using the scale factors to obtain normalized subband signals; and

(5) quantizing the normalized subband signals to obtain encoded subband signals in accordance with a channel-specific, intra-channel perceptual threshold of human hearing, step (a-5) being controlled by the bit allocation information;

(b) transmitting or storing the encoded subband signals, the bit allocation information, the scale factors, and a control signal which is derived from the audio representation format; and

(c) decoding and processing the encoded subband signals, step (c) including

(1) decoding the encoded subband signals according to the bit allocation information to obtain decoded subband signals;

(2) decombining the decoded subband signals of different audio channels, but the same frequency position, to obtain decombined subband signals, step (c-2) being conducted as a function of the control signal;

(3) denormalizing the decombined subband signals according to the scale factors, to obtain denormalized subband signals; and

(4) generating a broadband digital audio signal from the denormalized subband signals.