Method for coding an audio signal

• US 6,424,939 B1
• Filed: 10/06/1999
• Issued: 07/23/2002
• Est. Priority Date: 07/14/1997
• Status: Expired due to Term

• Current Assignee: Fraunhofer Gesellschaft Zur Foerderung Der Angewandten Forsching E.V.
• Original Assignee: Fraunhofer Gesellschaft Zur Foerderung Der Angewandten Forsching E.V.
• Inventors: Ehret, Andreas, Herre, Jürgen, Dietz, Martin, Kunz, Oliver, Teichmann, Bodo, Brandenburg, Karlheinz, Gerhauser, Heinz, Gbur, Uwe
• Primary Examiner(s): Korzuch, William
• Assistant Examiner(s): MCFADDEN, SUSAN IRIS
• Application Number:
  US09/402,684
• Time in Patent Office: 1,021 Days
• Field of Search: 704/219, 704/220, 704/226, 704/227, 704/228, 704/230, 704/800
• US Class Current: 704/219
• CPC Class Codes: G10L 19/028 Noise substitution, i.e. substituting non-tonal spectral components by noisy source comfort noise for discontinuous speech transmission G10L19/012
• CPC Class Codes: H04B 1/665 using psychoacoustic properties of the ear, e.g. masking effect

Abstract

A method for coding or decoding an audio signal combines the advantages of TNS processing and noise substitution. A time-discrete audio signal is initially transformed to the frequency domain in order to obtain spectral values of the temporal audio signal. Subsequently, a prediction of the spectral values in relation to frequency is carried out in order to obtain spectral residual values. Within the spectral residual values, areas are detected encompassing spectral residual values with noise properties. The spectral residual values in the noise areas are noise-substituted, whereupon information concerning the noise areas and noise substitution is incorporated into side information pertaining to a coded audio signal. Thus, considerable bit savings in case of transient signals can be achieved.

Claims

1. A method for coding an audio signal, comprising the following steps:

transforming a temporal audio signal to the frequency domain in order to obtain spectral values of the temporal audio signal;

performing a prediction of the spectral values in relation to frequency in order to obtain spectral residual values;

detecting noise areas in the spectral residual values;

noise-substituting the spectral residual values in the noise areas; and

introducing information concerning the noise areas and the noise-substitution in side information of a coded audio signal.

2. The method of claim 1, wherein the prediction or inverse prediction, respectively, is carried out only for a specific range of the spectral values.

3. The method of claim 1, wherein a hint that prediction was used, prediction coefficients and, optionally, frequency range information for the prediction are contained in the side information of the coded audio signal.

4. The method of claim 1, wherein filter coefficients generated by prediction are introduced into side information of the coded audio signal.

5. The method of claim 1, wherein the noise substitution is carried out in the form of scale factor bands.

6. The method of claim 1, wherein the step of noise substitution comprises calculating of the energy of the spectral residual values in a detected noise area having a scale factor band, the information contained in the side information of the coded audio signal and concerning noise substitution being contained at the location provided in the bit flow for the scale factor(s) for the one scale factor band and comprising the energy of the error residual values in a corresponding noise area.

7. The method of claim 1, wherein, during coding, the spectral residual values not located in noise areas are quantized in consideration of psychoacoustic masking, and wherein, during decoding, the quantized spectral residual values not located in noise areas are requantized and then subjected to inverse prediction in order to obtain the spectral values.

8. A method for decoding a coded audio signal, comprising the following steps:

receiving the coded audio signal;

detecting information in the side information that relates to noise substitution and to noise areas of the spectral residual values;

generating spectral noise residual values on the basis of the detected information in the noise areas;

performing an inverse prediction in relation to frequency in order to obtain spectral values from the noise-substituted spectral noise residual values; and

transforming the spectral values to the time domain in order to obtain a decoded audio signal.

9. The method of claim 8, wherein the prediction or inverse prediction, respectively, is carried out only for a specific range of the spectral values.

10. The method of claim 8, wherein a hint that prediction was used, prediction coefficients and, optionally, frequency range information for the prediction are contained in the side information of the coded audio signal.

11. The method of claim 8, wherein filter coefficients generated by prediction are introduced into side information of the coded audio signal.

12. The method of claim 8, wherein the noise substitution is carried out in the form of scale factor bands.

13. The method of claim 8, wherein the step of noise substitution comprises calculating of the energy of the spectral residual values in a detected noise area having a scale factor band, the information contained in the side information of the coded audio signal and concerning noise substitution being contained at the location provided in the bit flow for the scale factor(s) for the one scale factor band and comprising the energy of the error residual values in a corresponding noise area.

14. The method of claim 8, wherein, during coding, the spectral residual values not located in noise areas are quantized in consideration of psychoacoustic masking, and wherein, during decoding, the quantized spectral residual values not located in noise areas are requantized and then subjected to inverse prediction in order to obtain the spectral values.