Coding method and apparatus for multiple channels of audio information representing three-dimensional sound fields

• US 6,021,386 A
• Filed: 03/09/1999
• Issued: 02/01/2000
• Est. Priority Date: 01/08/1991
• Status: Expired due to Fees

• Current Assignee: Dolby Laboratories Licensing Corporation (Dolby Laboratories Incorporated)
• Original Assignee: Dolby Laboratories Licensing Corporation (Dolby Laboratories Incorporated)
• Inventors: Todd, Craig Campbell, Davis, Mark Franklin
• Primary Examiner(s): Dorvil, Richemond
• Application Number:
  US09/264,781
• Time in Patent Office: 329 Days
• Field of Search: 704/200, 704/229, 704/230, 704/205, 704/206, 704/500, 704/501, 704/502, 704/503, 704/504
• US Class Current: 704/229
• CPC Class Codes: G10L 19/008 Multichannel audio signal coding or decoding using interchannel correlation to reduce redundancy, e.g. joint-stereo, intensity-coding or matrixing
• CPC Class Codes: G11B 20/00 Signal processing not specific to the method of recording or reproducing; Circuits therefor
• CPC Class Codes: G11B 20/10527 Audio or video recording; Data buffering arrangements G11B20/12 - G11B20/18 take precedence
• CPC Class Codes: H04H 20/91 broadcasting computer programmes
• CPC Class Codes: H04S 3/008 in which the audio signals are in digital form, i.e. employing more than two discrete digital channels data reduction aspects thereof based on psychoacoustics G10L19/02

Abstract

In an encoder, multiple channels of audio information representing multidimensional sound fields are split into subband signals and the subband signals in one or more subbands are combined to form composite signals. The composite signals, the subband signals not combined into a composite signal and information describing the spectral levels of subband signals combined into composite signals are assembled into an encoded output signal. The spectral level information conveys either the amplitude or power of the combined subband signals or the apparent direction of the sound field represented by the combined subband signals. In digital implementations, adaptive bit allocation may be used to reduce the informational requirements of the encoded signal.

Claims

1. An apparatus for encoding audio information for a plurality of audio channels comprising:

one or more input terminals receiving subband signals for the plurality of audio channels,

a coarse-level quantizer coupled to the one or more input terminals, wherein the coarse-level quantizer derives coarse measures of subband signal contents,

a subband-allocation estimator coupled to the coarse-level quantizer, wherein the subband-allocation estimator derives an estimated number of bits to quantize the subband signals in response to the coarse measure of subband signal contents,

a steering controller coupled to the one or more input terminals and to the subband-allocation estimator, wherein the steering controller generates composite representations of subband signals in respective frequency subbands for a plurality of the audio channels in response to the estimated number of bits, and generates steering control information representing levels of the subband signals represented by the composite representations,

an adaptive-bit allocator coupled to the coarse-level quantizer and to the steering controller, wherein the adaptive-bit allocator allocates bits to the composite representations and to subband signals not represented by the composite representations in response to the coarse measure of subband contents and the steering control information, and

a quantizer coupled to the one or more input terminals and to the steering controller, wherein the quantizer quantizes the composite representations and quantizes subband signals not represented by the composite representations.

2. An apparatus according to claim 1 wherein the adaptive-bit allocator allocates the bits to the composite representations and to subband signals not represented by the composite representations from a common pool of bits.

3. An apparatus according to claim 1 wherein the steering controller, in response to the estimated number of bits, adaptively selects the respective subbands for which the composite representations are generated.

4. An apparatus according to claim 1 wherein the steering controller generates the composite representations in a manner that reduces signal loss caused by subband signals being out of phase.

5. An apparatus according to claim 1 wherein the quantizer introduces a noise-like component having an amplitude matched to an expected quantizing error.

6. An apparatus for decoding input information representing a plurality of audio channels comprising:

one or more input terminals receiving the input information, wherein the input information comprises coarse measures of frequency subband content, quantized subband signals for a respective frequency subband for the plurality of audio channels, a quantized composite representation of subband signals in another respective frequency subband for the plurality of audio channels, and steering control information,

an adaptive bit-allocation calculator coupled to the one or more input terminals, wherein the adaptive bit-allocation calculator determines numbers of bits to allocate to the quantized subband signals and the quantized composite representation of subband signals in response to the steering control information and the coarse measures of frequency subband content,

a dequantizer coupled to the one or more input terminals and to the adaptive bit-allocation calculator, wherein the dequantizer dequantizes the quantized subband signals and the quantized composite representation of subband signals according to the numbers of bits determined by the adaptive bit-allocation calculator to generate dequantized subband signals and a dequantized composite representation of subband signals, respectively, and

an inverse steering controller coupled to the dequantizer and to the one or more input terminals, wherein the inverse steering controller generates a plurality of subband signals in response to the dequantized composite representation and the steering control information.

7. An apparatus according to claim 6 wherein the adaptive bit-allocation calculator determines the numbers of bits to allocate to the quantized subband signals and the quantized composite representation from a common pool of bits.

8. An apparatus according to claim 6 wherein the input information comprises a phase-cancellation compensation signal and the inverse steering controller adapts its operation in response to the phase-cancellation compensation signal to generate the plurality of subband signals in a manner that introduces out-of-phase conditions.

9. An apparatus according to claim 6 wherein the dequantizer introduces a noise-like component having an amplitude matched to an expected quantizing error.

10. An apparatus for encoding audio information, wherein the apparatus comprises:

means for generating a discrete representation of subband signals in each respective frequency subband in a first set of frequency subbands for three or more channels of audio information representing a three-dimensional sound field,

means for generating an aggregate representation of subband signals in each respective frequency subband in a second set of frequency subbands for the three or more of channels of audio information, wherein the aggregate representation for a respective frequency subband includes a composite audio-information signal and a steering control signal, and wherein the composite audio-information signal and the steering signal correspond to levels of the subband signals in the respective frequency subband, and

means for encoding the discrete representations of subband signals and the aggregate representations of subband signals.

11. An apparatus according to claim 10 that comprises means for adaptively selecting the frequency subbands in the first and the second sets of frequency subbands according to estimates of information capacity requirements of the three or more channels of audio information.

12. An apparatus according to claim 10 wherein the aggregate representations are generated in a manner that reduces signal loss caused by subband signals being out of phase.

13. An apparatus for decoding input information, wherein the apparatus comprises:

means for receiving and decoding the input information to obtain discrete representations of subband signals in each respective frequency subband in a first set of frequency subbands for three or more channels of audio information representing a three-dimensional soundfield, and to obtain aggregate representations of subband signals in each respective frequency subband in a second set of frequency subbands for the three or more channels of audio information,

means for generating discrete subband signals in the second set of frequency subbands for the three or more channels of audio information from the aggregate representations, wherein the aggregate representation for a respective frequency subband includes a composite audio-information signal and a steering control signal, and wherein the composite audio-information signal and the steering signal correspond to levels of subband signals in the respective frequency subband for three or more audio channels, and

means for generating three or more channels of audio information representing a three dimensional sound field from discrete subband signals for the first set of frequency subbands and from discrete subband signals for the second set of frequency subbands.

14. An apparatus according to claim 13 that comprises means for receiving and decoding the input information to obtain a phase-cancellation compensation signal, wherein the means for generating the discrete subband signals in the second set of frequency subbands adapts its operation to compensate for techniques used to reduce signal loss caused by subband signals being out of phase.

15. A method for encoding audio information, the method comprising:

receiving input signals conveying three or more channels of audio information representing a three-dimensional sound field,

applying one or more filters or one or more transforms to the input signals to generate a plurality of subband signals for each of the channels of audio information,

combining subband signals to generate composite signals, wherein a respective composite signal is generated by combining subband signals in a respective frequency subband and at least one composite signal is generated by combining subband signals for at least three of the channels of audio information,

generating a steering control signal for each respective composite signal representing levels of the subband signals that are combined to generate that respective composite signal,

allocating a number of bits to and quantizing the composite signals and the subband signals not combined into composite signals to generate encoded information, and

assembling the encoded information and information representing the steering control signals into an output signal.

16. A method according to claim 15 that allocates the bits to the composite signals and the subband signals not combined into composite signals from a common bit pool.

17. A method according to claim 15 that comprises reducing effects of out-of-phase cancellation by controlling levels of the subband signals combined to generate the composite signals.

18. A method according to claim 15 that comprises reducing effects of out-of-phase cancellation by controlling phase of one or more subband signals combined to generate the composite signals.

19. A method according to claim 15 that comprises excluding one or more subband signals for a selected channel of audio information from being combined into respective composite signals such that the number of bits to be allocated is less than or equal to an allocation limit.

20. A method according to claim 15 that comprises selecting in which frequency subbands composite signals are generated such that the number of bits to be allocated is less than or equal to an allocation limit.

21. A method according to claim 15 wherein the steering control signal for a respective frequency subband conveys either subtractive differences in levels or multiplicative ratios of levels for subband signals that are combined to generate the composite signal for that respective subband.

22. A method for decoding an input signal conveying an encoded representation of audio information, the method comprising:

disassembling an input signal to obtain encoded information and steering control signals,

allocating a number of bits to and dequantizing the encoded information to obtain composite signals and subband signals,

generating subband signals from the composite signals, wherein subband signals are generated for three or more audio channels in a respective frequency subband from a respective composite signal for that respective frequency subband according to a respective steering control signal that represents levels of subband signals that were combined to generate that respective composite signal, and

applying one or more synthesis filters or one or more inverse transforms to the subband signals obtained from the input signal and generated from composite signals to generate three or more channels of audio information representing a three dimensional sound field.

23. A method according to claim 22 that allocates bits the encoded information from a common bit pool.

24. A method according to claim 22 that comprises compensating for techniques used to reduce effects of out-of-phase cancellation by controlling levels of one or more subband signals generated from the composite signals.

25. A method according to claim 22 that comprises compensating for techniques used to reduce effects of out-of-phase cancellation by controlling phase of one or more subband signals generated from the composite signals.

26. A method according to claim 22 wherein the steering control signal for a respective frequency subband conveys either subtractive differences in levels or multiplicative ratios of levels for subband signals combined to generate the composite signal for that respective subband.