Acoustic coding of an enhancement frame having a shorter time length than a base frame

• US 7,996,233 B2
• Filed: 08/12/2003
• Issued: 08/09/2011
• Est. Priority Date: 09/06/2002
• Status: Expired due to Fees

• Current Assignee: Panasonic Intellectual Property Corporation of America (Panasonic Holdings Corporation)
• Original Assignee: Panasonic Corporation (Panasonic Holdings Corporation)
• Inventors: Oshikiri, Masahiro
• Primary Examiner(s): SHAH, PARAS D
• Application Number:
  US10/526,566
• Publication Number:
  US 20050252361A1
• Time in Patent Office: 2,919 Days
• Field of Search: 704/219, 704/220, 704/207, 704/500, 704/501, 704/200.1, 704/200
• US Class Current: 704/500
• CPC Class Codes: G10L 19/16 Vocoder architecture
• CPC Class Codes: G10L 19/22 Mode decision, i.e. based on audio signal content versus external parameters
• CPC Class Codes: G10L 19/24 Variable rate codecs, e.g. for generating different qualities using a scalable representation such as hierarchical encoding or layered encoding

Abstract

A downsampler 101 converts input data having a sampling rate 2*FH to a sampling rate 2*FL which is lower than the sampling rate 2*FH. A base layer coder 102 encodes the input data having the sampling rate 2*FL in predetermined base frame units. A local decoder 103 decodes a first coded code. An upsampler 104 increases the sampling rate of the decoded signal to 2*FH. A subtractor 106 subtracts the decoded signal from the input signal and regards the subtraction result as a residual signal. A frame divider 107 divides the residual signal into enhancement frames having a shorter time length than that of the base frame. An enhancement layer coder 108 encodes the residual signal divided into the enhancement frames and outputs a second coded code obtained by this coding to a multiplexer 109.

Claims

1. An acoustic coding apparatus comprising a processor comprising:

a base layer coding section that encodes an input signal per base frame and obtains a base layer coded code;

a decoding section that decodes the base layer coded code and obtains a decoded signal;

a subtraction section that obtains a residual signal between the input signal and the decoded signal;

a frame division section that divides the residual signal into a plurality of residual signals in units of an enhancement frame having a shorter time length than the base frame;

an enhancement layer coding section performed by the processor that encodes the plurality of residual signals and obtains an enhancement layer coded code; and

a multiplexing section that multiplexes the base layer coded code and the enhancement layer coded code to output a multiplexed code,

wherein the enhancement layer coding section comprises;

a frequency domain transform section that transforms the plurality of residual signals in the frequency domain and obtains a plurality of frequency domain transform coefficients represented on a two dimensional plane comprised of a time axis and a frequency axis;

a domain divider that divides the plurality of frequency domain transform coefficients into a plurality of domains on the two dimensional plane such that each domain includes at least a plurality of frequency domain transform coefficients which are grouped continuously along a time axis;

a quantization domain determining section that determines a part of the plurality of domains in each base frame to be quantization targets based on power spectrum values of the frequency domain transform coefficients within each domain and outputs domain information showing the part of the plurality of domains; and

a quantization domain coding section that encodes the domain information and the frequency domain transform coefficients within the part of the plurality of domains shown by the domain information, and obtains the enhancement layer coded code.

2. The acoustic coding apparatus according to claim 1, wherein the base layer coding section encodes the input signal using a code excited linear prediction coding.

3. The acoustic coding apparatus according to claim 1, wherein the enhancement layer coding section transforms the residual signal from the time domain to the frequency domain using a modified discrete cosine transform.

4. The acoustic coding apparatus according to claim 3, wherein the enhancement layer coding section encodes only part of a band, shown by the domain information, of the residual signal transformed to the frequency domain.

5. The acoustic coding apparatus according to claim 3, further comprising a perceptual masking section that calculates perceptual masking expressing an amplitude value which does not affect auditory perception, wherein the enhancement layer coding section does not regard signals in the perceptual masking as coding targets.

6. The acoustic coding apparatus according to claim 5, wherein the enhancement layer coding section calculates a difference between the perceptual masking and the residual signal, regards a residual signal for which the difference is relatively large as a coding target and encodes positions in the time domain and the frequency domain in which the residual signal exists on the two dimensional plane.

7. An acoustic decoding apparatus comprising a processor comprising:

a demultiplexing section that demultiplexes a code coded by an acoustic coding apparatus into a base layer coded code and an enhancement layer coded code;

a base layer decoding section that decodes the base layer coded code which is generated at a coding side in predetermined base frame units and obtains a base layer signal; and

an enhancement layer decoding section performed by the processor that decodes the enhancement layer coded code which is generated at the coding side in units of an enhancement frame having a shorter time length than a base frame and obtains an enhancement layer signal;

wherein the enhancement layer decoding section comprises;

a domain divider that, on a two dimensional plane comprised of a time axis and a frequency axis, divides a plurality of frequency domain transform coefficients into a plurality of domains such that each domain includes at least a plurality of frequency transform coefficients which are grouped continuously along a time axis;

a quantization domain determining section that generates domain information from enhancement layer coded code in each base frame showing domains which are quantization targets and determined based on power spectrum values of the frequency domain transform coefficients within each domain at the coding side and determines quantization target domains from the plurality of domains using the domain information; and

a transform coefficient decoding section that generates the frequency domain transform coefficients included in the quantization target domains from enhancement layer coded code and obtains the enhancement layer signal.

8. The acoustic decoding apparatus according to claim 7, wherein the base layer decoding section decodes the base layer coded code using code excited linear prediction coding.

9. The acoustic decoding apparatus according to claim 7, wherein the transform coefficient decoding section transforms the frequency domain transform coefficients in a time domain signal using an inverse modified discrete cosine transform.

10. An acoustic coding method executed by a processor, the method comprising:

a base layer coding step of encoding an input signal per base frame and generating a base layer coded code in a base layer coding section;

a decoding step of decoding the base layer coded code and generating a decoded signal in a decoding section;

a subtracting step of generating a residual signal between the input signal and the decoded signal in a subtracting section; and

a frame division step of dividing the residual signal into a plurality of residual signals in units of an enhancement frame having a shorter time length than the base frame in a frame division section;

an enhancement layer coding step of encoding the plurality of residual signals and generating an enhancement layer coded code in an enhancement layer coding section; and

a multiplexing step of multiplexing the base layer coded code and the enhancement layer coded code to output a multiplexed code in a multiplexing section,

wherein the enhancement layer coding step comprises;

a frequency domain transform step of transforming the plurality of residual signals in the frequency domain and generating a plurality of frequency domain transform coefficients represented on a two dimensional plane comprised of a time axis and a frequency axis;

a domain division step of dividing the plurality of frequency domain transform coefficients into plurality of domains on the two dimensional plane such that each domain includes at least a plurality of frequency domain transform coefficients which are grouped continuously along a time axis;

a quantization domain determining step of determining a part of the plurality of domains in each base frame to be quantization targets based on power spectrum values of the frequency domain transform coefficients within each domain and outputting domain information showing the part of the plurality of domains; and

a quantization domain coding step of encoding the domain information and the frequency domain transform coefficients within the part of the plurality of domains shown by the domain information and generating the enhancement layer coded code.

11. An acoustic decoding method executed by a processor, the method comprising;

a demultiplexing step of demultiplexing a code coded by an acoustic coding apparatus into a base layer coded code and an enhancement layer coded code in a demultiplexing section;

a base layer decoding step of decoding the base layer coded code which is generated at a coding side in predetermined base frame units and generating a base layer signal in a base layer decoding section; and

an enhancement layer decoding step of decoding the enhancement layer coded code which is generated at the coding side in units of an enhancement frame having a shorter time length than the base frame and generating an enhancement layer signal in an enhancement layer decoding section;

wherein the enhancement layer decoding step comprises;

a domain division step of, on a two dimensional plane comprised of a time axis and a frequency axis, dividing a plurality of frequency domain transform coefficients into a plurality of domains such that each domain includes at least a plurality of frequency domain transform coefficients which are grouped continuously along a time axis;

a quantization domain determining step of generating domain information from enhancement layer coded code in each base frame showing domains which are quantization targets and determined based on power spectrum values of the frequency domain transform coefficients within each domain at the coding side and determines quantization target domains from the plurality of domains using the domain information; and

a transform coefficient decoding step of generating the frequency domain transform coefficients included in the quantization target domains from enhancement layer coded code and generating the enhancement layer signal.