Methods and apparatus for quantizing, encoding and decoding and recording media therefor

• US 5,774,844 A
• Filed: 06/22/1995
• Issued: 06/30/1998
• Est. Priority Date: 11/09/1993
• Status: Expired due to Term

• Current Assignee: Sony Corporation (Sony Group Corp.)
• Original Assignee: Sony Corporation (Sony Group Corp.)
• Inventors: Akagiri, Kenzo
• Primary Examiner(s): Hudspeth, David R.
• Assistant Examiner(s): Chawan, Vijay B.
• Application Number:
  US08/464,787
• Time in Patent Office: 1,104 Days
• Field of Search: 395/239, 395/2.24, 395/2.35, 395/2.38, 395/2.92, 395/2.95
• US Class Current: 704/230
• CPC Class Codes: H04B 14/046 Systems or methods for reducing noise or bandwidth

Abstract

In this invention, in quantizing time region samples or frequency region samples of an input signal, an audio signal, etc., an approach is employed to allow a MDCT coefficient as the frequency region sample to undergo block floating processing. Thereafter, the processed MDCT coefficient is quantized to generate a first quantized value which is utilized calculate a quantization error between the first quantized value and the MDCT coefficient each quantization error is individually generated using one sample. The quantization error is then utilized to generate a second quantized value to thereby decompose the MDCT coefficient into at least two words. Thus, in accordance with this invention, degradation of sound quality can be minimized, and backward compatibility can be ensured.

Claims

1. A quantizing apparatus for quantizing time domain samples or frequency domain samples of an input signal, comprising:

samples thereafter to further quantize, individually by one sample, each quantization error at quantization of the preceding stage to thereby decompose the time region sample or the frequency region

at least one quantizing means for quantizing the samples; and

means for determining a quantization error based upon the samples and the quantized samples, wherein the at least one quantizing means also quantizes the quantization error to thereby decompose the samples into at least two words.

2. A quantizing apparatus as set forth in claim 1,

wherein the at least one quantizing means performs a round-off operation.

3. A quantizing apparatus as set forth in claim 1 or 2,

wherein an output bit rate of at least one of the quantizing means is a fixed bit rate in predetermined time units.

4. A quantizing apparatus as set forth in claim 3,

wherein output bit rates of all the quantizing means are a fixed bit rate in predetermined time units.

5. A quantizing apparatus as set forth in claim 1,

wherein the samples undergo block floating processing for every block of plural samples to determine, from a scale factor for a quantized output of at least a preceding stage, a scale factor for quantizing the quantization error of the preceding stage.

6. A quantizing apparatus as set forth in claim 5,

wherein the scale factor for quantizing the quantization error of the preceding stage is determined from the scale factor and a word length for the quantized output of the at least the preceding stage.

7. A quantizing apparatus as set forth in claim 5,

wherein, with respect to samples within blocks subdivided with respect to time and frequency, quantization having a same block floating and word length is carried out within the subdivided blocks.

8. A decoding apparatus for decoding signals generated by quantizing time domain samples or frequency domain samples, the apparatus comprising:

at least one inverse normalization means for inverse normalizing the quantized samples; and

synthesis means for combining into one sample at least two inverse normalized quantized partial samples, the at least two inverse normalized quantized partial samples relating to samples in the same domain.

12. A decoding apparatus for decoding signals obtained by quantizing time domain samples or frequency domain samples, the apparatus comprising:

at least one inverse normalization means for inverse normalizing the quantized samples; and

synthesis means for combining into one sample at least two inverse normalized quantized partial samples, the at least two inverse normalized quantized partial samples relating to samples in the same domain, wherein the samples in the same domain are of an entire bandwidth region.

9. A decoding apparatus as set forth in claim 8, 12 or 13;

wherein, in the quantization, at least one quantizing means quantizes samples, and thereafter quantizes, individually by one sample, each quantization error of a preceding stage to thereby decompose the sample into at least two words, and

wherein an output bit rate of at least one of the quantizing means is a fixed bit rate in predetermined time units.

10. A decoding apparatus as set forth in claim 9,

wherein, in the quantization, at least one quantizing means quantizes samples, and thereafter quantizes, individually by one sample, each quantization error of a preceding stage to thereby decompose each sample into at least two words, and

wherein output bit rates of all the quantizing means are a fixed bit rate in predetermined time units.

11. A decoding apparatus as set forth in claim 10,

wherein, in the quantization, samples undergo block floating processing every block of plural samples, and

wherein a scale factor for sample data obtained by quantizing each quantization error of the preceding stage is determined from a scale factor for a quantized output of at least the preceding stage.

13. A decoding apparatus adapted for decoding signals obtained by quantizing time domain samples or frequency domain samples, the apparatus comprising:

at least one inverse normalization means for inverse normalizing the quantized samples; and

synthesis means for combining into one sample at least two inverse normalized quantized partial samples, the at least two inverse normalized quantized partial samples relating to samples in the same domain, wherein the samples in the same domain are individual time domain samples of partial bandwidth regions, wherein the combined one sample is a time domain sample of an entire bandwidth region.

14. A quantizing method for quantizing a predetermined sample in the time domain or the frequency domain of an input signal, the method comprising the steps of:

quantizing the predetermined sample to generate a first quantized value;

calculating a quantization error between the predetermined sample and the first quantized value;

quantizing each quantization error individually by one sample to generate a second quantized value; and

outputting the first and second quantized values.

15. A quantizing method as set forth in claim 14,

wherein at least one of the first quantized value and the second quantized value is generated by quantization based upon a round-off operation.

16. A quantizing method as set forth in claim 14,

wherein at least one of the first quantized value and the second quantized value is output at a fixed bit rate.

17. A quantizing method as set forth in claim 14,

wherein the first quantized value and the second quantized value are output so that a total bit rate thereof becomes equal to a fixed bit rate.

18. A quantizing method as set forth in claim 14, the method further comprising the steps of:

allowing the predetermined samples to undergo block floating processing by using a first scale factor for every block comprised of plural samples; and

normalizing the quantization errors by a second scale factor determined on the basis of the first scale factor.

19. A quantizing method as set forth in claim 18,

wherein the second scale factor is determined on the basis of a word length in generating the first scale factor and the first quantized value.

20. A quantizing method as set forth in claim 14,

wherein, with respect to sample data within blocks subdivided with respect to time and frequency, normalization and quantization are performed using a same scale factor and word length within the subdivided blocks.

21. An encoding apparatus for allocating bits to time domain samples or frequency domain samples of an input signal comprising:

at least one quantizing means for quantizing the samples; and

means for determining a quantization error based upon the samples and the quantized samples, wherein the at least one quantizing means also quantizes the quantization error to thereby decompose the samples into at least two words.

22. An encoding apparatus as set forth in claim 21,

wherein the at least one quantizing means performs a round-off operation.

23. An encoding apparatus as set forth in claim 21 or 22,

wherein an output bit rate of at least one of the quantizing means is a fixed bit rate in predetermined time units.

24. An encoding apparatus as set forth in claim 23,

wherein output bit rates of all the quantizing means are a fixed bit rate in predetermined time units.

25. An encoding apparatus as set forth in claim 21,

wherein the samples undergo block floating processing for every block of plural samples to determine, from a scale factor for a quantized output of at least the preceding stage, a scale factor for quantizing the quantization error of the preceding stage.

26. An encoding apparatus as set forth in claim 25,

wherein the scale factor for sample data to quantize the quantization error of the preceding stage is determined from the scale factor and a word length for a quantized output of at least the preceding stage.

27. An encoding apparatus as set forth in claim 21,

wherein, with respect to sample data within blocks subdivided with respect to time and frequency, quantization having a same block floating or the same block floating and a word length is carried out within the subdivided blocks.

28. An encoding apparatus adapted for carrying out bit allocation for time domain samples or frequency domain samples of an input signal to encode the input signal for recording the encoded signal onto or into a recording medium, comprising:

at least one quantizing means for quantizing the samples;

means for determining a quantization error based upon the samples and the quantized samples, wherein the at least one quantizing means also quantizes the quantization error to thereby decompose the samples into at least two words; and

means for separately recording in predetermined time units, output samples of all the quantizing means into one sync block.

29. An encoding apparatus adapted for carrying out bit allocation for time domain samples or frequency domain samples of an input signal to encode the input signal to record the encoded signal onto or into a recording medium, comprising:

at least one quantizing means for quantizing the samples;

means for determining a quantization error based upon the samples and the quantized samples, wherein the at least one quantizing means also quantizes the quantization error to thereby decompose the samples into at least two words; and

means for alternately recording, into one sync block, output samples of respective ones of the quantizing means in order of one of time and frequency for every respective quantizing means.

30. A code decoding apparatus for decoding signals encoded after undergoing bit allocation for time domain samples or frequency domain samples, comprising:

at least one inverse normalization means for inverse normalizing the quantized samples; and

synthesis means for combining into one sample at least two inverse normalized quantized partial samples, the at least two inverse normalized quantized partial samples relating to samples of the same domain.

31. A code decoding apparatus as set forth in claim 30, 34 or 35,

wherein a bit rate of at least one of the encoded signals is a fixed bit rate in predetermined time units.

32. A code decoding apparatus as set forth in claim 31,

wherein bit rates of all the encoded signals are a fixed bit rate in predetermined time units.

33. A code decoding apparatus as set forth in claim 32,

wherein, with respect to the encoded signal, domain samples undergo block floating processing for every block of plural samples to determine, from a scale factor for a quantized output obtained by quantizing at least one domain sample, a scale factor for sample data obtained by quantizing a quantization error of the preceding stage.

36. A code decoding apparatus as set forth in claim 30,

wherein sample data which undergo quantization having a same block floating or the same block floating and a word length within blocks subdivided with respect to time and frequency is decoded.

34. A code decoding apparatus adapted for decoding signals encoded after undergoing bit allocation for time domain samples or frequency domain samples, comprising:

at least one inverse normalization means for inverse normalizing the quantized samples; and

synthesis means for combining into one sample at least two partial samples, the at least two inverse normalized partial samples relating to samples of the same domain, wherein the samples in the same domain are of an entire bandwidth region.

35. A code decoding apparatus adapted for decoding signals encoded after undergoing bit allocation for time domain samples or frequency domain samples, comprising:

at least one inverse normalization means for inverse normalizing the quantized samples; and

synthesis means for combining into one sample at least two inverse normalized quantized partial samples, the at least two inverse normalized quantized partial samples relating to samples of the same domain, wherein the samples in the same domain are individual samples of partial bandwidth regions, wherein the combined one sample is a time domain sample of an entire bandwidth region.

37. An encoding method for encoding an input signal, comprising the steps of:

dividing a total bit allocation into a basic bit allocation and an additional bit allocation;

quantizing a predetermined sample of at least one of a time domain and a frequency domain of the input signal by a word length based on the basic bit allocation to generate a first quantized value;

calculating a quantization error between the predetermined sample and the first quantized value; and

quantizing each quantization error individually by one sample to generate a second quantized value as the additional bit allocation.

38. An encoding method as set forth in claim 37,

wherein at least one of the first quantized value and the second quantized value is generated by quantization based on a round-off operation.

39. An encoding method as set forth in claim 37,

wherein at least one of the first quantized value and the second quantized value is output at a fixed bit rate.

40. An encoding method as set forth in claim 37,

wherein the first quantized value and the second quantized value are output so that a total bit rate thereof becomes a fixed bit rate.

41. An encoding method as set forth in claim 37, the method further comprising the steps of:

allowing the predetermined samples to undergo block floating processing by using a first scale factor for every block comprised of plural samples; and

normalizing the quantization errors by a second scale factor determined on the basis of the first scale factor.

42. An encoding method as set forth in claim 37,

wherein the second scale factor is determined on the basis of a word length in generating the first scale factor and the first quantized value.

43. An encoding method as set forth in claim 37,

wherein, with respect to sample data within blocks subdivided with respect to time and frequency, normalization and quantization are performed using a same scale factor and word length within the subdivided blocks.