Low bit-rate spatial coding method and system

• US 6,016,473 A
• Filed: 04/07/1998
• Issued: 01/18/2000
• Est. Priority Date: 04/07/1998
• Status: Expired due to Term

• Current Assignee: Dolby Laboratories Licensing Corporation (Dolby Laboratories Incorporated)
• Original Assignee: Ray M. Dolby
• Inventors: Dolby, Ray M.
• Primary Examiner(s): Hudspeth, David R.
• Assistant Examiner(s): ABEBE, DANIEL DEMELASH
• Application Number:
  US09/056,503
• Time in Patent Office: 651 Days
• Field of Search: 704/500, 704/501
• US Class Current: 704/500
• 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: H04B 1/667 using a division in frequency subbands for TV signals H04N19/63
• CPC Class Codes: H04S 2420/03 Application of parametric coding in stereophonic audio systems
• 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

A spatial audio coding system, including an encoder and a decoder, operates at very low bit-rates and is useful for audio via the Internet. The listener or listeners preferably are located within a predictable listening area, for example, users of a personal computer or television viewers. An encoder produces a composite audio-information signal representing the soundfield to be reproduced and a directional vector or "steering control signal." The composite audio-information signal has its frequency spectrum broken into a number of subbands, preferably commensurate with the critical bands of the human ear. The steering control signal has a component relating to the dominant direction of the soundfield in each of the subbands. Because the system is based on the premise that only sound from a single direction is heard at any instant, the decoder need not apply a signal to more than two sound transducers at any instant.

Claims

1. A low bit-rate spatial coding system for encoding a plurality of audio streams representing a soundfield into an encoded signal and decoding said encoded signal, said system including an encoder and a decoder, said encoder comprising

means for generating a plurality of subband signals in response to said plurality of audio streams, each subband signal representing a respective frequency subband of a respective one of said audio streams,

means for generating a composite signal representing the combination of subband signals in respective frequency subbands,

means for generating a steering control signal for said composite signal indicating the principal direction of said soundfield in respective subbands,

means for generating encoded information by allocating bits to said composite signal and said steering control signal, and

means for assembling said encoded information into an encoded signal, and

said decoder comprising

means for deriving the composite signal and steering control signal from said encoded signal,

means for deriving subband signals in response to said composite signal and said steering control signal,

means for supplying reproduction information describing the number of output channels of said decoder and the location or virtual location of sound transducers connected to the respective output channels, wherein there are three or more output channels, and

means for generating an audio stream in no more than two output channels at any instant in response to said subband signals and reproduction information.

4. A low bit-rate spatial coding system according to claim 1 or claim 2 further comprising a spatializer including an acoustic-crossfeed canceller for processing audio streams in said output channels, the spatializer providing a reduced number of output channels having audio streams with characteristics such that when coupled to sound transducers provide the auditory impression of a soundfield similar to that which would have resulted if the original number of output channels were coupled to a larger number of sound transducers.

5. A low bit-rate spatial coding system according to claim 1 or claim 2 wherein said steering control signal comprises a directional vector for each subband.

6. A low bit-rate spatial coding system according to claim 1 or claim 2 wherein said respective frequency subband has a bandwidth commensurate with a respective critical band of the human auditory system.

7. A low bit-rate spatial coding system according to claim 1 or claim 2 wherein said bits are allocated according to psychoacoustic principles.

2. A low bit-rate spatial coding system for encoding a plurality of audio streams representing a soundfield into an encoded signal, decoding said encoded signal, and reproducing an auditory likeness of said soundfield, said system including an encoder and a decoder, said encoder comprising

means for generating a plurality of subband signals in response to said plurality of audio streams, each subband signal representing a respective frequency subband of a respective one of said audio streams,

means for generating a composite signal representing the combination of subband signals in each frequency subband,

means for generating a steering control signal for said composite signal indicating the principal direction of said soundfield in each subband,

means for generating encoded information by allocating bits to said composite signal and said steering control signal, and

means for assembling said encoded information into an encoded signal,

said decoder comprising

means for deriving the composite signal and steering control signal from said encoded signal,

means for deriving subband signals in response to said composite signal and said steering control signal,

means for supplying reproduction information describing the number of output channels of said decoder and the location or virtual location of sound transducers connected to the respective output channels, and

means for generating an audio stream in one or more output channels in response to said subband signals and reproduction information,

and further comprising

a plurality of sound transducers coupled to the output channels of said decoder and arranged so as to generate an auditory likeness of said soundfield to a listener or listeners within a spatial coding sweet-spot listening area.

3. A low bit-rate spatial coding system according to claim 2 wherein there are three or more output channels, and said means for generating an audio stream in one of more output channels in response to said subband signals and reproduction information generates an audio stream in no more than two output channels at any instant.

8. A low bit-rate spatial coding system according to claim 2 wherein said decoder further comprises a spatializer including an acoustic-crossfeed canceller for processing audio streams in said output channels, whereby the auditory likeness of said soundfield generated within the spatial coding sweet-spot listening area renders the perception of a wide-stage or surround-sound soundfield.

9. A decoder for use in a low bit-rate spatial coding system for decoding an encoded signal derived from a plurality of audio streams representing a soundfield by generating a plurality of subband signals in response to the plurality of audio streams, each subband signal representing a respective frequency subband of a respective one of said audio streams, generating a composite signal representing the combination of subband signals in respective frequency subbands, generating a steering control signal for the composite signal indicating the principal direction of said soundfield in respective subbands, generating encoded information by allocating bits to the composite signal and the steering control signal, and assembling the encoded information into an encoded signal, comprising

means for deriving the composite signal and steering control signal from said encoded signal,

means for deriving subband signals in response to said composite signal and said steering control signal,

means for supplying reproduction information describing the number of output channels of said decoder and the location or virtual location of sound transducers connected to the respective output channels, wherein there are three or more output channels, and

means for generating an audio stream in no more than two output channels at any instant in response to said subband signals and reproduction information.

10. A decoder and reproduction system for use in a low bit-rate spatial coding system for decoding and reproducing an encoded signal derived from a plurality of audio streams representing a soundfield by generating a plurality of subband signals in response to the plurality of audio streams, each subband signal representing a respective frequency subband of a respective one of said audio streams, generating a composite signal representing the combination of subband signals in respective frequency subbands, generating a steering control signal for the composite signal indicating the principal direction of said soundfield in respective subbands, generating encoded information by allocating bits to the composite signal and the steering control signal, and assembling the encoded information into an encoded signal, comprising

means for deriving the composite signal and steering control signal from said encoded signal,

means for deriving subband signals in response to said composite signal and said steering control signal,

means for supplying reproduction information describing the number of output channels of said decoder and the location or virtual location of sound transducers connected to the respective output channels, and

means for generating an audio stream in one or more output channels in response to said subband signals and reproduction information, and

a plurality of sound transducers coupled to the output channels of said decoder and arranged so as to generate an auditory likeness of said soundfield to a listener or listeners within a spatial coding sweet-spot listening area.

11. The combination of claim 10 wherein there are three or more output channels, and said means for generating an audio stream in one of more output channels in response to said subband signals and reproduction information generates an audio stream in no more than two output channels at any instant.

12. The combination of claim 9 or claim 10 further comprising a spatializer including an acoustic-crossfeed canceller for processing audio streams in said output channels, the spatializer providing a reduced number of output channels having audio streams with characteristics such that when coupled to sound transducers provide the auditory impression of a soundfield similar to that which would have resulted if the original number of output channels were coupled to a larger number of sound transducers.

13. The combination of claim 9 or claim 10 wherein said steering control signal comprises a directional vector for each subband.

14. The combination of claim 9 or claim 10 wherein said respective frequency subband has a bandwidth commensurate with a respective critical band of the human auditory system.

15. The combination of claim 9 or claim 10 wherein said bits are allocated according to psychoacoustic principles.

16. The combination of claim 10 wherein said decoder further comprises a spatializer including an acoustic-crossfeed canceller for processing audio streams in said output channels, whereby the auditory likeness of said soundfield generated within the spatial coding sweet-spot listening area renders the perception of a wide-stage or surround-sound soundfield.

17. A low bit-rate spatial coding system for encoding a plurality of audio streams representing a soundfield into an encoded signal and decoding said encoded signal, said system including an encoder and a decoder, said encoder comprising

a subband signal generator generating a plurality of subband signals in response to said plurality of audio streams, each subband signal representing a respective frequency subband of a respective one of said audio streams,

a signal combiner generating a composite signal representing the combination of subband signals in respective frequency subbands,

a soundfield direction detector generating a steering control signal for said composite signal indicating the principal direction of said soundfield in respective subbands,

an encoder and bit allocator generating encoded information by allocating bits to said composite signal and said steering control signal, and

a formatter assembling said encoded information into an encoded signal, and

said decoder comprising

a deformatter deriving the composite signal and steering control signal from said encoded signal,

an inverse subband generator deriving subband signals in response to said composite signal and said steering control signal,

an information input describing the number of output channels of said decoder and the location or virtual location of sound transducers connected to the respective output channels, wherein there are three or more output channels, and

a signal generator generating an audio stream in no more than two output channels at any instant in response to said subband signals and reproduction information.

20. A low bit-rate spatial coding system according to claim 17 or claim 18 further comprising a spatializer including an acoustic-crossfeed canceller for processing audio streams in said output channels, the spatializer providing a reduced number of output channels having audio streams with characteristics such that when coupled to sound transducers provide the auditory impression of a soundfield similar to that which would have resulted if the original number of output channels were coupled to a larger number of sound transducers.

21. A low bit-rate spatial coding system according to claim 17 or claim 18 wherein said steering control signal comprises a directional vector for each subband.

22. A low bit-rate spatial coding system according to claim 17 or claim 18 wherein said respective frequency subband has a bandwidth commensurate with a respective critical band of the human auditory system.

23. A low bit-rate spatial coding system according to claim 17 or claim 18 wherein said bits are allocated according to psychoacoustic principles.

18. A low bit-rate spatial coding system for encoding a plurality of audio streams representing a soundfield into an encoded signal, decoding said encoded signal, and reproducing an auditory likeness of said soundfield, said system including an encoder and a decoder, said encoder comprising

a subband signal generator generating a plurality of subband signals in response to said plurality of audio streams, each subband signal representing a respective frequency subband of a respective one of said audio streams,

a signal combiner generating a composite signal representing the combination of subband signals in each frequency subband,

a soundfield direction detector generating a steering control signal for said composite signal indicating the principal direction of said soundfield in each subband,

an encoder and bit allocator generating encoded information by allocating bits to said composite signal and said steering control signal, and

a formatter assembling said encoded information into an encoded signal,

said decoder comprising

a deformatter deriving the composite signal and steering control signal from said encoded signal,

an inverse subband generator deriving subband signals in response to said composite signal and said steering control signal,

an information input describing the number of output channels of said decoder and the location or virtual location of sound transducers connected to the respective output channels, and

a signal generator generating an audio stream in one or more output channels in response to said subband signals and reproduction information,

and further comprising

a plurality of sound transducers coupled to the output channels of said decoder and arranged so as to generate an auditory likeness of said soundfield to a listener or listeners within a spatial coding sweet-spot listening area.

19. A low bit-rate spatial coding system according to claim 18 wherein there are three or more output channels and said signal generator generates an audio stream in no more than two output channels at any instant.

24. A low bit-rate spatial coding system according to claim 18 wherein said decoder further comprises a spatializer including an acoustic-crossfeed canceller for processing audio streams in said output channels, whereby the auditory likeness of said soundfield generated within the spatial coding sweet-spot listening area renders the perception of a wide-stage or surround-sound soundfield.

25. A decoder for use in a low bit-rate spatial coding system for decoding an encoded signal derived from a plurality of audio streams representing a soundfield by generating a plurality of subband signals in response to the plurality of audio streams, each subband signal representing a respective frequency subband of a respective one of said audio streams, generating a composite signal representing the combination of subband signals in respective frequency subbands, generating a steering control signal for the composite signal indicating the principal direction of said soundfield in respective subbands, generating encoded information by allocating bits to the composite signal and the steering control signal, and assembling the encoded information into an encoded signal, comprising

a deformatter deriving the composite signal and steering control signal from said encoded signal,

an inverse subband generator deriving subband signals in response to said composite signal and said steering control signal,

an information input describing the number of output channels of said decoder and the location or virtual location of sound transducers connected to the respective output channels, wherein there are three or more output channels, and

a signal generator generating an audio stream in no more than two output channels at any instant in response to said subband signals and reproduction information.

28. The combination of claim 25 or claim 26 further comprising a spatializer including an acoustic-crossfeed canceller for processing audio streams in said output channels, the spatializer providing a reduced number of output channels having audio streams with characteristics such that when coupled to sound transducers provide the auditory impression of a soundfield similar to that which would have resulted if the original number of output channels were coupled to a larger number of sound transducers.

29. The combination of claim 25 or claim 26 wherein said steering control signal comprises a directional vector for each subband.

30. The combination of claim 25 or claim 26 wherein said respective frequency subband has a bandwidth commensurate with a respective critical band of the human auditory system.

31. The combination of claim 25 or claim 26 wherein said bits are allocated according to psychoacoustic principles.

26. A decoder and reproduction system for use in a low bit-rate spatial coding system for decoding and reproducing an encoded signal derived from a plurality of audio streams representing a soundfield by generating a plurality of subband signals in response to the plurality of audio streams, each subband signal representing a respective frequency subband of a respective one of said audio streams, generating a composite signal representing the combination of subband signals in respective frequency subbands, generating a steering control signal for the composite signal indicating the principal direction of said soundfield in respective subbands, generating encoded information by allocating bits to the composite signal and the steering control signal, and assembling the encoded information into an encoded signal, comprising

a deformatter deriving the composite signal and steering control signal from said encoded signal,

an inverse subband generator deriving subband signals in response to said composite signal and said steering control signal,

an information input describing the number of output channels of said decoder and the location or virtual location of sound transducers connected to the respective output channels, and

a signal generator generating an audio stream in one or more output channels in response to said subband signals and reproduction information, and

a plurality of sound transducers coupled to the output channels of said decoder and arranged so as to generate an auditory likeness of said soundfield to a listener or listeners within a spatial coding sweet-spot listening area.

27. The combination of claim 26 wherein there are three or more output channels, and said generator generates an audio stream in no more than two output channels at any instant.

32. The combination of claim 26 wherein said decoder further comprises a spatializer including an acoustic-crossfeed canceller for processing audio streams in said output channels, whereby the auditory likeness of said soundfield generated within the spatial coding sweet-spot listening area renders the perception of a wide-stage or surround-sound soundfield.

33. A low bit-rate spatial coding method for encoding a plurality of audio streams representing a soundfield into an encoded signal and decoding said encoded signal, said method including encoding and decoding, said encoding comprising

generating a plurality of subband signals in response to said plurality of audio streams, each subband signal representing a respective frequency subband of a respective one of said audio streams,

generating a composite signal representing the combination of subband signals in respective frequency subbands,

generating a steering control signal for said composite signal indicating the principal direction of said soundfield in respective subbands,

generating encoded information by allocating bits to said composite signal and said steering control signal, and

assembling said encoded information into an encoded signal, and

said decoding comprising

deriving the composite signal and steering control signal from said encoded signal,

deriving subband signals in response to said composite signal and said steering control signal,

supplying reproduction information describing the number of output channels of said decoder and the location or virtual location of sound transducers connected to the respective output channels, wherein there are three or more output channels, and

generating an audio stream in no more than two output channels at any instant in response to said subband signals and reproduction information.

36. A low bit-rate spatial coding method according to claim 33 or claim 34 further comprising processing audio streams in said output channels with a spatializer including an acoustic-crossfeed canceller, said processing providing a reduced number of output channels having audio streams with characteristics such that when coupled to sound transducers provide the auditory impression of a soundfield similar to that which would have resulted if the original number of output channels were coupled to a larger number of sound transducers.

37. A low bit-rate spatial coding method according to claim 33 or claim 34 wherein said steering control signal comprises a directional vector for each subband.

38. A low bit-rate spatial coding method according to claim 33 or claim 34 wherein said respective frequency subband has a bandwidth commensurate with a respective critical band of the human auditory system.

39. A low bit-rate spatial coding method according to claim 33 or claim 34 wherein said bits are allocated according to psychoacoustic principles.

34. A low bit-rate spatial coding method for encoding a plurality of audio streams representing a soundfield into an encoded signal, decoding said encoded signal, and reproducing an auditory likeness of said soundfield, said method including an encoder and a decoder, said encoding comprising

generating a plurality of subband signals in response to said plurality of audio streams, each subband signal representing a respective frequency subband of a respective one of said audio streams,

generating a composite signal representing the combination of subband signals in each frequency subband,

generating a steering control signal for said composite signal indicating the principal direction of said soundfield in each subband,

generating encoded information by allocating bits to said composite signal and said steering control signal, and

assembling said encoded information into an encoded signal,

said decoding comprising

deriving the composite signal and steering control signal from said encoded signal,

deriving subband signals in response to said composite signal and said steering control signal,

supplying reproduction information describing the number of output channels of said decoder and the location or virtual location of sound transducers connected to the respective output channels, and

generating an audio stream in one or more output channels in response to said subband signals and reproduction information,

and further comprising

coupling said output channels to a plurality of sound transducers arranged so as to generate an auditory likeness of said soundfield to a listener or listeners within a spatial coding sweet-spot listening area.

35. A low bit-rate spatial coding method according to claim 34 wherein there are three or more output channels, and said generating an audio stream in one of more output channels in response to said subband signals and reproduction information generates an audio stream in no more than two output channels at any instant.

40. A low bit-rate spatial coding method according to claim 34 wherein said decoder further comprises processing audio streams in said output channels with a spatializer including an acoustic-crossfeed canceller, whereby the auditory likeness of said soundfield generated within the spatial coding sweet-spot listening area renders the perception of a wide-stage or surround-sound soundfield.

41. A low bit-rate spatial coding decoding method for decoding an encoded signal derived from a plurality of audio streams representing a soundfield by generating a plurality of subband signals in response to the plurality of audio streams, each subband signal representing a respective frequency subband of a respective one of said audio streams, generating a composite signal representing the combination of subband signals in respective frequency subbands, generating a steering control signal for the composite signal indicating the principal direction of said soundfield in respective subbands, generating encoded information by allocating bits to the composite signal and the steering control signal, and assembling the encoded information into an encoded signal, comprising

deriving the composite signal and steering control signal from said encoded signal,

deriving subband signals in response to said composite signal and said steering control signal,

supplying reproduction information describing the number of output channels of said decoder and the location or virtual location of sound transducers connected to the respective output channels, wherein there are three or more output channels, and

generating an audio stream in no more than two output channels at any instant in response to said subband signals and reproduction information.

44. The method of claim 41 or claim 42 further comprising processing audio streams in said output channels with a spatializer including an acoustic-crossfeed canceller, said processing providing a reduced number of output channels having audio streams with characteristics such that when coupled to sound transducers provide the auditory impression of a soundfield similar to that which would have resulted if the original number of output channels were coupled to a larger number of sound transducers.

45. The method of claim 41 or claim 42 wherein said steering control signal comprises a directional vector for each subband.

46. The method of claim 41 or claim 42 wherein said respective frequency subband has a bandwidth commensurate with a respective critical band of the human auditory system.

47. The method of claim 41 or claim 42 wherein said bits are allocated according to psychoacoustic principles.

42. A low bit-rate spatial coding decoding and reproduction method for decoding and reproducing an encoded signal derived from a plurality of audio streams representing a soundfield by generating a plurality of subband signals in response to the plurality of audio streams, each subband signal representing a respective frequency subband of a respective one of said audio streams, generating a composite signal representing the combination of subband signals in respective frequency subbands, generating a steering control signal for the composite signal indicating the principal direction of said soundfield in respective subbands, generating encoded information by allocating bits to the composite signal and the steering control signal, and assembling the encoded information into an encoded signal, comprising

deriving the composite signal and steering control signal from said encoded signal,

deriving subband signals in response to said composite signal and said steering control signal,

supplying reproduction information describing the number of output channels of said decoder and the location or virtual location of sound transducers connected to the respective output channels, and

generating an audio stream in one or more output channels in response to said subband signals and reproduction information, and

coupling a plurality of sound transducers to the output channels of said decoder, the sound transducers arranged so as to generate an auditory likeness of said soundfield to a listener or listeners within a spatial coding sweet-spot listening area.

43. The method of claim 42 wherein there are three or more output channels, and said generating an audio stream in one of more output channels in response to said subband signals and reproduction information generates an audio stream in no more than two output channels at any instant.

48. The method of claim 42 further comprising processing audio streams in said output channels with a spatializer including an acoustic-crossfeed canceller, whereby the auditory likeness of said soundfield generated within the spatial coding sweet-spot listening area renders the perception of a wide-stage or surround-sound soundfield.