METHOD AND APPARATUS FOR PROCESSING AUDIO SIGNALS AT LOW COMPLEXITY

• US 20130275140A1
• Filed: 02/04/2013
• Published: 10/17/2013
• Est. Priority Date: 02/02/2012
• Status: Abandoned Application

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: KIM, Mi-young, POROV, Anton Viktorovich, OSIPOV, Konstantin Sergeevich
• Application Number:
  US13/758,054
• Publication Number:
  US 20130275140A1
• Time in Patent Office: Days
• US Class Current: 704/500
• CPC Class Codes: G10L 19/002 Dynamic bit allocation for perceptual audio coders G10L19/032
• CPC Class Codes: G10L 19/0208 Subband vocoders
• CPC Class Codes: G10L 19/032 Quantisation or dequantisation of spectral components

Abstract

An audio encoding apparatus is provided that includes a transform unit to transform an audio signal in the time domain into an audio spectrum in the frequency domain; a bit allocation unit to determine the number of allocated bits by using spectral energy in predetermined frequency band units for the audio spectrum; and an encoding unit to determine the number of unit magnitude pulses for factorial pulse coding based on the number of allocated bits for the audio spectrum and to perform factorial pulse coding in the frequency band units for the audio spectrum by using the determined number of unit magnitude pulses.

Claims

1. A spectrum encoding method comprising:

determining a number of unit magnitude pulses for factorial pulse coding based on a number of allocated bits in frequency band units for a spectrum; and

performing factorial pulse coding in the frequency band units for the spectrum by using the determined number of unit magnitude pulses.

2. The spectrum encoding method of claim 1, wherein the determining of the number of unit magnitude pulses comprises estimating a maximum value of the number of unit magnitude pulses with respect to the number of allocated bits and determining the final number of unit magnitude pulses of each frequency band by performing a binary search within a range from a minimum value to the maximum value.

3. The spectrum encoding method of claim 1, wherein the minimum value is 0.

4. The spectrum encoding method of claim 1, wherein the minimum value is a value obtained by subtracting a predetermined constant from the maximum value according to a length of a frequency band.

5. The spectrum encoding method of claim 1, wherein the determining of the number of unit magnitude pulses comprises estimating a maximum value of the number of unit magnitude pulses with respect to the number of allocated bits and determining the final number of unit magnitude pulses of each frequency band by selectively performing a binary search method and a linear decrement method within a range of a minimum value and the maximum value.

6. The spectrum encoding method of claim 5, wherein the minimum value is 0.

7. The spectrum encoding method of claim 5, wherein the minimum value is a value obtained by subtracting a predetermined constant from the maximum value according to a length of a frequency band.

8. The spectrum encoding method of claim 5, wherein one of a binary search method and a linear decrement method is used for a predetermined number of unit magnitude pulses by using a difference value between the number of bits required for factorial pulse coding and the number of allocated bits.

9. The spectrum encoding method of claim 1, wherein the number of unit magnitude pulses of a corresponding frequency band is allocated to be 0 when the number of allocated bits is less than the number of bits required for coding at least one pulse.

21. A non-transitory computer-readable storage medium having stored therein program instructions, which when executed by a computer, perform the method of claim 1.

10. An audio encoding apparatus comprising:

a transform unit to transform an audio signal in a time domain into an audio spectrum in a frequency domain;

a bit allocation unit to determine a number of allocated bits by using spectral energy in predetermined frequency band units for the audio spectrum; and

an encoding unit to determine a number of unit magnitude pulses for factorial pulse coding based on the number of allocated bits for the audio spectrum and to perform factorial pulse coding in the frequency band units for the audio spectrum by using the determined number of unit magnitude pulses.

11. A spectrum decoding method comprising:

determining a number of unit magnitude pulses for factorial pulse coding based on a number of allocated bits in frequency band units for a spectrum; and

performing factorial pulse decoding in the frequency band units for the spectrum by using the determined number of unit magnitude pulses.

12. The spectrum decoding method of claim 11, wherein the determining of the number of unit magnitude pulses comprises estimating a maximum value of the number of unit magnitude pulses with respect to the number of allocated bits and determining the final number of unit magnitude pulses of each frequency band by performing a binary search within a range from a minimum value to the maximum value.

13. The spectrum decoding method of claim 11, wherein the minimum value is 0.

14. The spectrum decoding method of claim 11, wherein the minimum value is a value obtained by subtracting a predetermined constant from the maximum value according to a length of a frequency band.

15. The spectrum decoding method of claim 11, wherein the determining of the number of unit magnitude pulses comprises estimating a maximum value of the number of unit magnitude pulses with respect to the number of allocated bits and determining the final number of unit magnitude pulses of each frequency band by selectively using a binary search method and a linear decrement method with respect to a range from a minimum value to the maximum value.

16. The spectrum decoding method of claim 15, wherein the minimum value is 0.

17. The spectrum decoding method of claim 15, wherein the minimum value is a value obtained by subtracting a predetermined constant from the maximum value according to a length of a frequency band.

18. The spectrum decoding method of claim 15, wherein one of the binary search method and the linear decrement method is used for a predetermined number of unit magnitude pulses by using a difference value between the number of bits required for factorial pulse coding and the number of allocated bits.

19. The spectrum decoding method of claim 11, wherein the number of unit magnitude pulses of a corresponding frequency band is allocated to be 0 when the number of allocated bits is less than the number of bits required for coding at least one pulse.

22. A non-transitory computer-readable storage medium having stored therein program instructions, which when executed by a computer, perform the method of claim 11.

20. An audio decoding apparatus comprising:

a bit allocation unit to determine a number of allocated bits by using spectral energy in predetermined frequency band units for an audio spectrum included in a bitstream;

a decoding unit to determine a number of unit magnitude pulses for factorial pulse decoding based on the number of allocated bits in frequency band units for the audio spectrum and to perform factorial pulse decoding in the frequency band units for the audio spectrum by using the determined number of unit magnitude pulses; and

an inverse transform unit to transform the audio spectrum decoded by the decoding unit into an audio signal in a time domain.