Half-rate vocoder

• US 8,359,197 B2
• Filed: 04/01/2003
• Issued: 01/22/2013
• Est. Priority Date: 04/01/2003
• Status: Active Grant

• Current Assignee: Digital Voice Systems, Inc.
• Original Assignee: Digital Voice Systems, Inc.
• Inventors: Hardwick, John C.
• Primary Examiner(s): Armstrong, Angela A
• Application Number:
  US10/402,938
• Publication Number:
  US 20050278169A1
• Time in Patent Office: 3,584 Days
• Field of Search: 704/230, 704/223, 704219-220
• US Class Current: 704/223
• CPC Class Codes: G10L 19/005 Correction of errors induced by the transmission channel, if related to the coding algorithm
• CPC Class Codes: G10L 19/087 using mixed excitation models, e.g. MELP, MBE, split band LPC or HVXC
• CPC Class Codes: G10L 19/167 Audio streaming, i.e. formatting and decoding of an encoded audio signal representation into a data stream for transmission or storage purposes

Abstract

Encoding a sequence of digital speech samples into a bit stream includes dividing the digital speech samples into one or more frames, computing model parameters for a frame, and quantizing the model parameters to produce pitch bits conveying pitch information, voicing bits conveying voicing information, and gain bits conveying signal level information. One or more of the pitch bits are combined with one or more of the voicing bits and one or more of the gain bits to create a first parameter codeword that is encoded with an error control code to produce a first FEC codeword that is included in a bit stream for the frame. The process may be reversed to decode the bit stream.

Claims

1. A method of encoding a sequence of digital speech samples into a bit stream, the method comprising:

dividing the digital speech samples into one or more frames;

computing model parameters for a frame;

quantizing the model parameters to produce pitch bits conveying pitch information, voicing bits conveying voicing information, and gain bits conveying signal level information, wherein the pitch bits, the voicing bits and the gain bits are included in quantizer bits for the frame;

combining one or more of the pitch bits with one or more of the voicing bits and one or more of the gain bits to create a first parameter codeword that includes less than all of the quantizer bits for the frame;

encoding the first parameter codeword with an error control code to produce a first FEC (“

forward error control”

) codeword; and

including the first FEC codeword in a bit stream for the frame.

2. The method of claim 1, wherein computing the model parameters for the frame include computing a fundamental frequency parameter, one or more of voicing decisions, and a set of spectral parameters.

3. The method of claim 2, wherein computing the model parameters for a frame includes using the Multi-Band Excitation speech model.

4. The method of claim 2, wherein quantizing the model parameters comprises producing the pitch bits by applying a logarithmic function to the fundamental frequency parameter.

5. The method of claim 3, wherein quantizing the model parameters comprises producing the voicing bits by jointly quantizing voicing decisions for the frame.

6. The method of claim 5, wherein:

the voicing bits represent an index into a voicing codebook, and

the value of the voicing codebook is the same for two or more different values of the index.

7. The method of claim 1, wherein the first parameter codeword comprises twelve bits.

8. The method of claim 7, wherein the first parameter codeword is formed by combining four of the pitch bits, plus four of the voicing bits, plus four of the gain bits.

9. The method of claim 8, wherein the first parameter codeword is encoded with a Golay error control code.

10. The method of claim 8, wherein:

the spectral parameters include a set of logarithmic spectral magnitudes, and

the gain bits are produced at least in part by computing the mean of the logarithmic spectral magnitudes.

11. The method of claim 10, further comprising:

quantizing the logarithmic spectral magnitudes into spectral bits; and

combining a plurality of the spectral bits to create a second parameter codeword; and

encoding the second parameter codeword with a second error control code to produce a second FEC codeword,

wherein the second FEC codeword is also included in the bit stream for the frame.

12. The method of claim 11, wherein:

the pitch bits, voicing bits, gain bits and spectral bits are each divided into more important bits and less important bits,

the more important pitch bits, voicing bits, gain bits, and spectral bits are included in the first parameter codeword and the second parameter codeword and encoded with error control codes, and

the less important pitch bits, voicing bits, gain bits, and spectral bits are included in the bit stream for the frame without encoding with error control codes.

13. The method of claim 12, wherein:

there are 7 pitch bits divided into 4 more important pitch bits and 3 less important pitch bits,

there are 5 voicing bits divided into 4 more important voicing bits and 1 less important voicing bit, and

there are 5 gain bits divided into 4 more important gain bits and 1 less important gain bit.

14. The method of claim 13, wherein the second parameter code comprises twelve more important spectral bits which are encoded with a Golay error control code to produce the second FEC codeword.

15. The method of claim 14, further comprising:

computing a modulation key from the first parameter codeword;

generating a scrambling sequence from the modulation key;

combining the scrambling sequence with the second FEC codeword to produce a scrambled second FEC codeword; and

including the scrambled second FEC codeword in the bit stream for the frame.

16. The method of claim 8, further comprising:

detecting certain tone signals; and

if a tone signal is detected for a frame, then including tone identifier bits and tone amplitude bits in the first parameter codeword, wherein the tone identifier bits allow the bits for the frame to be identified as corresponding to a tone signal.

17. The method of claim 16, wherein:

if a tone signal is detected for a frame then additional tone index bits are included in the bit stream for the frame, and

the tone index bits determine frequency information for the tone signal.

18. The method of claim 17, wherein the tone identifier bits correspond to a disallowed set of pitch bits to permit the bits for the frame to be identified as corresponding to a tone signal.

19. The method of claim 18, wherein the first parameter codeword comprises six tone identifier bits and six tone amplitude bits if a tone signal is detected for a frame.

20. The method of claim 7, wherein the first parameter codeword is encoded with a Golay error control code.

21. The method of claim 7, further comprising:

detecting certain tone signals; and

if a tone signal is detected for a frame, then including tone identifier bits and tone amplitude bits in the first parameter codeword, wherein the tone identifier bits allow the bits for the frame to be identified as corresponding to a tone signal.

22. The method of claim 21, wherein:

if a tone signal is detected for a frame then additional tone index bits are included in the bit stream for the frame, and

the tone index bits determine frequency information for the tone signal.

23. The method of claim 22, wherein the tone identifier bits correspond to a disallowed set of pitch bits to permit the bits for the frame can be identified as corresponding to a tone signal.

24. The method of claim 23, wherein the first parameter codeword comprises six tone identifier bits and six tone amplitude bits if a tone signal is detected for a frame.

25. The method of claim 6, wherein:

the spectral parameters include a set of logarithmic spectral magnitudes, and

the gain bits are produced at least in part by computing the mean of the logarithmic spectral magnitudes.

26. The method of claim 25, further comprising:

quantizing the logarithmic spectral magnitudes into spectral bits; and

combining a plurality of the spectral bits to create a second parameter codeword; and

encoding the second parameter codeword with a second error control code to produce a second FEC codeword,

wherein the second FEC codeword is also included in the bit stream for the frame.

27. The method of claim 26, wherein:

the pitch bits, voicing bits, gain bits and spectral bits are each divided into more important bits and less important bits,

the more important pitch bits, voicing bits, gain bits, and spectral bits are included in the first parameter codeword and the second parameter codeword and encoded with error control codes, and

the less important pitch bits, voicing bits, gain bits, and spectral bits are included in the bit stream for the frame without encoding with error control codes.

28. The method of claim 27, wherein:

there are 7 pitch bits divided into 4 more important pitch bits and 3 less important pitch bits,

there are 5 voicing bits divided into 4 more important voicing bits and 1 less important voicing bit, and

there are 5 gain bits divided into 4 more important gain bits and 1 less important gain bit.

29. The method of claim 28, wherein the second parameter code comprises twelve more important spectral bits which are encoded with a Golay error control code to produce the second FEC codeword.

30. The method of claim 29, further comprising:

computing a modulation key from the first parameter codeword;

generating a scrambling sequence from the modulation key;

combining the scrambling sequence with the second FEC codeword to produce a scrambled second FEC codeword; and

including the scrambled second FEC codeword in the bit stream for the frame.

31. The method of claim 2, wherein:

the spectral parameters include a set of logarithmic spectral magnitudes, and

the gain bits are produced at least in part by computing the mean of the logarithmic spectral magnitudes.

32. The method of claim 31, further comprising:

quantizing the logarithmic spectral magnitudes into spectral bits; and

combining a plurality of the spectral bits to create a second parameter codeword; and

encoding the second parameter codeword with a second error control code to produce a second FEC codeword,

wherein the second FEC codeword is also included in the bit stream for the frame.

33. The method of claim 32, wherein:

the pitch bits, voicing bits, gain bits and spectral bits are each divided into more important bits and less important bits,

the more important pitch bits, voicing bits, gain bits, and spectral bits are included in the first parameter codeword and the second parameter codeword and encoded with error control codes, and

the less important pitch bits, voicing bits, gain bits, and spectral bits are included in the bit stream for the frame without encoding with error control codes.

34. The method of claim 33, wherein:

there are 7 pitch bits divided into 4 more important pitch bits and 3 less important pitch bits,

there are 5 voicing bits divided into 4 more important voicing bits and 1 less important voicing bit, and

there are 5 gain bits divided into 4 more important gain bits and 1 less important gain bit.

35. The method of claim 34, wherein the second parameter code comprises twelve more important spectral bits which are encoded with a Golay error control code to produce the second FEC codeword.

36. The method of claim 35, further comprising:

computing a modulation key from the first parameter codeword;

generating a scrambling sequence from the modulation key;

combining the scrambling sequence with the second FEC codeword to produce a scrambled second FEC codeword; and

including the scrambled second FEC codeword in the bit stream for the frame.

37. The method of claim 1, wherein the first parameter codeword is encoded with a Golay error control code.

38. The method of claim 1, further comprising:

detecting certain tone signals; and

if a tone signal is detected for a frame, then including tone identifier bits and tone amplitude bits in the first parameter codeword, wherein the tone identifier bits allow the bits for the frame to be identified as corresponding to a tone signal.

39. The method of claim 38, wherein:

if a tone signal is detected for a frame then additional tone index bits are included in the bit stream for the frame, and

the tone index bits determine frequency information for the tone signal.

40. The method of claim 39, wherein the tone identifier bits correspond to a disallowed set of pitch bits to permit the bits for the frame can be identified as corresponding to a tone signal.

41. The method of claim 40, wherein the first parameter codeword comprises six tone identifier bits and six tone amplitude bits if a tone signal is detected for a frame.

42. A method for decoding digital speech samples from a bit stream, the method comprising:

dividing the bit stream into one or more frames of bits;

extracting a first FEC (“

forward error control”

) codeword from a frame of bits;

error control decoding the first FEC codeword to produce a first parameter codeword;

extracting pitch bits, voicing bits and gain bits from the first parameter codeword, the extracted pitch bits, voicing bits and gain bits including less than all of a set of quantizer bits for the frame;

using the extracted pitch bits to at least in part reconstruct pitch information for the frame;

using the extracted voicing bits to at least in part reconstruct voicing information for the frame;

using the extracted gain bits to at least in part reconstruct signal level information for the frame; and

using the reconstructed pitch information, voicing information and signal level information for one or more frames to compute digital speech samples.

43. The method of claim 42, wherein the pitch information for a frame includes a fundamental frequency parameter, and the voicing information for a frame includes one or more voicing decisions.

44. The method of claim 43, wherein the voicing decisions for the frame are reconstructed by using the voicing bits as an index into a voicing codebook.

45. The method of claim 44, wherein the value of the voicing codebook is the same for two or more different indices.

46. The method of claim 44, further comprising reconstructing spectral information for a frame.

47. The method of claim 46, wherein:

the spectral information for a frame comprises at least in part a set of logarithmic spectral magnitude parameters, and

the signal level information is used to determine the mean value of the logarithmic spectral magnitude parameters.

48. The method of claim 47, wherein:

the first FEC codeword is decoded with a Golay decoder, and

four pitch bits, plus four voicing bits, plus four gain bits are extracted from the first parameter codeword.

49. The method of claim 47, further comprising:

generating a modulation key from the first parameter codeword;

computing a scrambling sequence from the modulation key;

extracting a second FEC codeword from the frame of bits;

applying the scrambling sequence to the second FEC codeword to produce a descrambled second FEC codeword;

error control decoding the descrambled second FEC codeword to produce a second parameter codeword;

computing an error metric from the error control decoding of the first FEC codeword and from the error control decoding of the descrambled second FEC codeword; and

applying frame error processing if the error metric exceeds a threshold value.

50. The method of claim 49, wherein the frame error processing includes repeating the reconstructed model parameter from a previous frame for the current frame.

51. The method of claim 50, wherein the error metric uses the sum of the number of errors corrected by error control decoding the first FEC codeword and by error control decoding the descrambled second FEC codeword.

52. The method of claim 50, wherein the spectral information for a frame is reconstructed at least in part from the second parameter codeword.

53. The method of claim 43, further comprising reconstructing spectral information for a frame.

54. The method of claim 53, wherein:

the spectral information for a frame comprises at least in part a set of logarithmic spectral magnitude parameters, and

the signal level information is used to determine the mean value of the logarithmic spectral magnitude parameters.

55. The method of claim 54, wherein:

the first FEC codeword is decoded with a Golay decoder, and

four pitch bits, plus four voicing bits, plus four gain bits are extracted from the first parameter codeword.

56. The method of claim 54, further comprising:

generating a modulation key from the first parameter codeword;

computing a scrambling sequence from the modulation key;

extracting a second FEC codeword from the frame of bits;

applying the scrambling sequence to the second FEC codeword to produce a descrambled second FEC codeword;

error control decoding the descrambled second FEC codeword to produce a second parameter codeword;

computing an error metric from the error control decoding of the first FEC codeword and from the error control decoding of the descrambled second FEC codeword; and

applying frame error processing if the error metric exceeds a threshold value.

57. The method of claim 56, wherein the frame error processing includes repeating the reconstructed model parameter from a previous frame for the current frame.

58. The method of claim 57, wherein the error metric uses the sum of the number of errors corrected by error control decoding the first FEC codeword and by error control decoding the descrambled second FEC codeword.

59. The method of claim 57, wherein the spectral information for a frame is reconstructed at least in part from the second parameter codeword.

60. A method for decoding digital signal samples from a bit stream, the method comprising:

dividing the bit stream into one or more frames of bits;

extracting a first FEC (“

forward error control”

) codeword from a frame of bits;

error control decoding the first FEC codeword to produce a first parameter codeword;

using the first parameter codeword to determine whether the frame of bits corresponds to a tone signal;

extracting tone amplitude bits from the first parameter codeword if the frame of bits is determined to correspond to a tone signal, otherwise extracting pitch bits, voicing bits, and gain bits from the first codeword if the frame of bits is determined to not correspond to a tone signal, the extracted pitch bits, voicing bits and gain bits including less than all of a set of quantizer bits for the frame; and

using either the tone amplitude bits or the pitch bits, voicing bits and gain bits to compute digital signal samples.

61. The method of claim 60, further comprising:

generating a modulation key from the first parameter codeword;

computing a scrambling sequence from the modulation key;

extracting a second FEC codeword from the frame of bits;

applying the scrambling sequence to the second FEC codeword to produce a descrambled second FEC codeword;

error control decoding the descrambled second FEC codeword to produce a second parameter codeword; and

computing digital signal samples using the second parameter codeword.

62. The method of claim 61, further comprising:

summing the number of errors corrected by the error control decoding of the first FEC codeword and by the error control decoding of the descrambled second FEC codeword to compute an error metric; and

applying frame error processing if the error metric exceeds a threshold, wherein the frame error processing includes repeating the reconstructed model parameter from a previous frame.

63. The method of claim 61, wherein additional spectral bits are extracted from the second parameter codeword and used to reconstruct the digital signal samples.

64. The method of claim 63, wherein the spectral bits include tone index bits if the frame of bits is determined to correspond to a tone signal.

65. The method of claim 64, wherein the frame of bits is determined to correspond to a tone signal if some of the bits in the first parameter codeword equal a known tone identifier value which corresponds to a disallowed value of the pitch bits.

66. The method of claim 64, wherein the tone index bits are used to identify whether the frame of bits corresponds to a signal frequency tone, a DTMF tone, a Knox tone or a call progress tone.

67. The method of claim 64, wherein:

the spectral bits are used to reconstruct a set of logarithmic spectral magnitude parameters for the frame, and

the gain bits are used to determine the mean value of the logarithmic spectral magnitude parameters.

68. The method of claim 67, wherein the voicing bits are used as an index into a voicing codebook to reconstruct voicing decisions for the frame.

69. The method of claim 67, wherein:

the first FEC codeword is decoded with a Golay decoder, and

four pitch bits, plus four voicing bits, plus four gain bits are extracted from the first parameter codeword.

70. The method of claim 63, wherein the voicing bits are used as an index into a voicing codebook to reconstruct voicing decisions for the frame.

71. The method of claim 60, wherein the voicing bits are used as an index into a voicing codebook to reconstruct voicing decisions for the frame.

72. A method for decoding a frame of bits into speech samples, the method comprising:

determining the number of bits in the frame of bits;

extracting spectral bits from the frame of bits;

using one or more of the spectral bits to form a spectral codebook index, wherein the index is determined at least in part by the number of bits in the frame of bits;

reconstructing spectral information using the spectral codebook index; and

computing speech samples using the reconstructed spectral information.

73. The method of claim 72, wherein pitch bits, voicing bits and gain bits are also extracted from the frame of bits.

74. The method of claim 73, wherein the voicing bits are used as an index into a voicing codebook to reconstruct voicing information which is also used to compute the speech samples.

75. The method of claim 74, wherein the frame of bits is determined to correspond to a tone signal if some of the pitch bits and some of the voicing bits equal a known tone identifier value.

76. The method of claim 75, wherein:

the spectral information includes a set of logarithmic spectral magnitude parameters, and

the gain bits are used to determine the mean value of the logarithmic spectral magnitude parameters.

77. The method of claim 76, wherein the logarithmic spectral magnitude parameters for a frame are reconstructed using the extracted spectral bits for the frame combined with the reconstructed logarithmic spectral magnitude parameters from a previous frame.

78. The method of claim 76, wherein the mean value of the logarithmic spectral magnitude parameters for a frame is determined from the extracted gain bits for the frame and from the mean value of the logarithmic spectral magnitude parameters of a previous frame.

79. The method of claim 76, wherein the frame of bits includes 7 pitch bits representing the fundamental frequency, 5 voicing bits representing voicing decisions, and 5 gain bits representing the signal level.

80. The method of claim 74, wherein:

the spectral information includes a set of logarithmic spectral magnitude parameters, and

the gain bits are used to determine the mean value of the logarithmic spectral magnitude parameters.

81. The method of claim 80, wherein the logarithmic spectral magnitude parameters for a frame are reconstructed using the extracted spectral bits for the frame combined with the reconstructed logarithmic spectral magnitude parameters from a previous frame.

82. The method of claim 80, wherein the mean value of the logarithmic spectral magnitude parameters for a frame is determined from the extracted gain bits for the frame and from the mean value of the logarithmic spectral magnitude parameters of a previous frame.

83. The method of claim 80, wherein the frame of bits includes 7 pitch bits representing the fundamental frequency, 5 voicing bits representing voicing decisions, and 5 gain bits representing the signal level.

84. The method of claim 73, wherein:

the spectral information includes a set of logarithmic spectral magnitude parameters, and

the gain bits are used to determine the mean value of the logarithmic spectral magnitude parameters.

85. The method of claim 84, wherein the logarithmic spectral magnitude parameters for a frame are reconstructed using the extracted spectral bits for the frame combined with the reconstructed logarithmic spectral magnitude parameters from a previous frame.

86. The method of claim 84, wherein the mean value of the logarithmic spectral magnitude parameters for a frame is determined from the extracted gain bits for the frame and from the mean value of the logarithmic spectral magnitude parameters of a previous frame.

87. The method of claim 84 wherein the frame of bits includes 7 pitch bits representing the fundamental frequency, 5 voicing bits representing voicing decisions, and 5 gain bits representing the signal level.