Low Bitrate Audio Encoding/Decoding Scheme Having Cascaded Switches

• US 20110202354A1
• Filed: 01/11/2011
• Published: 08/18/2011
• Est. Priority Date: 07/11/2008
• Status: Active Grant

• Current Assignee: Fraunhofer Gesellschaft Zur Foerderung Der Angewandten Forsching E.V., VoiceAge Corporation
• Original Assignee: Fraunhofer Gesellschaft Zur Foerderung Der Angewandten Forsching E.V., VoiceAge Corporation
• Inventors: Salami, Redwan, Lecomte, Jeremie, Geiger, Ralf, Neuendorf, Max, Popp, Harald, Hilpert, Johannes, Bessette, Bruno, Grill, Bernhard, Fuchs, Guillaume, Multrus, Markus, Bayer, Stefan, Kraemer, Ulrich, Lapierre, Jimmy, Rettelbach, Nikolaus, Geyersberger, Stefan, Lefebvre, Roch, Gournay, Philippe
• Granted Patent: US 8,930,198 B2
• Time in Patent Office: Days
• US Class Current: 704/500
• CPC Class Codes: G10L 19/0017 Lossless audio signal coding; Perfect reconstruction of coded audio signal by transmission of coding error G10L19/24 takes precedence
• CPC Class Codes: G10L 19/008 Multichannel audio signal coding or decoding using interchannel correlation to reduce redundancy, e.g. joint-stereo, intensity-coding or matrixing
• CPC Class Codes: G10L 19/0212 using orthogonal transformation
• CPC Class Codes: G10L 19/173 Transcoding, i.e. converting between two coded representations avoiding cascaded coding-decoding
• CPC Class Codes: G10L 19/18 Vocoders using multiple modes
• CPC Class Codes: G10L 2019/0008 Algebraic codebooks

Abstract

An audio encoder has a first information sink oriented encoding branch such as a spectral domain encoding branch, a second information source or SNR oriented encoding branch such as an LPC-domain encoding branch, and a switch for switching between the first encoding branch and the second encoding branch, wherein the second encoding branch has a converter into a specific domain different from the spectral domain such as an LPC analysis stage generating an excitation signal, and wherein the second encoding branch furthermore has a specific domain coding branch such as LPC domain processing branch, and a specific spectral domain coding branch such as LPC spectral domain processing branch, and an additional switch for switching between the specific domain coding branch and the specific spectral domain coding branch. An audio decoder has a first domain decoder such as a spectral domain decoding branch, a second domain decoder such as an LPC domain decoding branch for decoding a signal such as an excitation signal in the second domain, and a third domain decoder such as an LPC-spectral decoder branch and two cascaded switches for switching between the decoders.

Claims

1. Audio encoder for encoding an audio input signal, the audio input signal being in a first domain, comprising:

a first coding branch for encoding an audio signal using a first coding algorithm to acquire a first encoded signal;

a second coding branch for encoding an audio signal using a second coding algorithm to acquire a second encoded signal, wherein the first coding algorithm is different from the second coding algorithm; and

a first switch for switching between the first coding branch and the second coding branch so that, for a portion of the audio input signal, either the first encoded signal or the second encoded signal is in an encoder output signal,

wherein the second coding branch comprises;

a converter for converting the audio signal into a second domain different from the first domain,

a first processing branch for processing an audio signal in the second domain to acquire a first processed signal;

a second processing branch for converting a signal into a third domain different from the first domain and the second domain and for processing the signal in the third domain to acquire a second processed signal; and

a second switch for switching between the first processing branch and the second processing branch so that, for a portion of the audio signal input into the second coding branch, either the first processed signal or the second processed signal is in the second encoded signal.

2. Audio encoder in accordance with claim 1, in which the first coding algorithm in the first coding branch is based on an information sink model, or in which the second coding algorithm in the second coding branch is based on an information source or a signal to noise ratio model.

3. Audio encoder in accordance with claim 1, in which the first coding branch comprises a converter for converting the audio input signal into a fourth domain different from the first domain, the second domain, and the third domain.

4. Audio encoder in accordance with claim 1, in which the first domain is the time domain, the second domain is an LPC domain acquired by an LPC filtering the first domain signal, the third domain is an LPC spectral domain acquired by converting an LPC filtered signal into a spectral domain, and the fourth domain is a spectral domain acquired by frequency domain converting the first domain signal.

5. Audio encoder in accordance with claim 1, further comprising a controller for controlling the first switch or the second switch in a signal adaptive way,

wherein the controller is operative to analyze a signal input into the first switch or output by the first coding branch or the second coding branch or a signal acquired by decoding an output signal of the first coding branch or the second coding branch with respect to a target function, or

wherein the controller is operative to analyze a signal input into the second switch or output by the first processing branch or the second processing branch or signals acquired by inverse processing output signals from the first processing branch and the second processing branch with respect to a target function.

6. Audio encoder in accordance with claim 1, in which the first coding branch or the second processing branch of the second coding branch comprises an aliasing introducing time/frequency converter and a quantizer/entropy coder stage and wherein the first processing branch of the second coding branch comprises a quantizer or entropy coder stage without an aliasing introducing conversion.

7. Audio encoder in accordance with claim 6, in which the aliasing introducing time/frequency converter comprises a windower for applying an analysis window and a modified discrete cosine transform algorithm, the windower being operative to apply the window function to subsequent frames in an overlapping manner so that a sample of an input signal into the windower occurs in at least two subsequent frames.

8. Audio encoder in accordance with claim 1, in which the first processing branch comprises the LPC excitation coding of an algebraic code excited linear prediction coder and the second processing branch comprises an MDCT spectral converter and a quantizer for quantizing spectral components to acquire quantized spectral components, wherein each quantized spectral component is zero or is defined by one quantization index of a plurality of quantization indices.

9. Audio encoder in accordance with claim 5, in which the controller is operative to control the first switch in an open loop manner and to control the second switch in a closed loop manner.

10. Audio encoder in accordance with claim 1, in which the first coding branch and the second coding branch are operative to encode the audio signal in a block wise manner, wherein the first switch or the second switch are switching in a block-wise manner so that a switching action takes place, at the minimum, after a block of a predefined number of samples of a signal, the predefined number of samples forming a frame length for the corresponding switch.

11. Audio encoder in accordance with claim 10, in which the frame length for the first switch is at least double the size of the frame length of the second switch.

12. Audio encoder in accordance with claim 5, in which the controller is operative to perform a speech/music discrimination in such a way that a decision to speech is favored with respect to a decision to music so that a decision to speech is taken even when a portion less than 50% of a frame for the first switch is speech and a portion more than 50% of the frame for the first switch is music.

13. Audio encoder in accordance with claim 5, in which a frame for the second switch is smaller than a frame for the first switch, and in which the controller is operative to take a decision to speech when only a portion of the first frame which comprises a length which is more than 50% of the length of the second frame is found out to comprise music.

14. Audio encoder in accordance with claim 1, in which the first encoding branch or the second processing branch of the second coding branch comprises a variable time warping functionality.

15. Method of encoding an audio input signal, the audio input signal being in a first domain, comprising:

encoding an audio signal using a first coding algorithm to acquire a first encoded signal;

encoding an audio signal using a second coding algorithm to acquire a second encoded signal, wherein the first coding algorithm is different from the second coding algorithm; and

switching between encoding using the first coding algorithm and encoding using the second coding algorithm so that, for a portion of the audio input signal, either the first encoded signal or the second encoded signal is in an encoded output signal,

wherein encoding using the second coding algorithm comprises;

converting the audio signal into a second domain different from the first domain,

processing an audio signal in the second domain to acquire a first processed signal;

converting a signal into a third domain different from the first domain and the second domain and processing the signal in the third domain to acquire a second processed signal; and

switching between processing the audio signal and converting and processing so that, for a portion of the audio signal encoded using the second coding algorithm, either the first processed signal or the second processed signal is in the second encoded signal.

16. Decoder for decoding an encoded audio signal, the encoded audio signal comprising a first coded signal, a first processed signal in a second domain, and a second processed signal in a third domain, wherein the first coded signal, the first processed signal, and the second processed signal are related to different time portions of a decoded audio signal, and wherein a first domain, the second domain and the third domain are different from each other, comprising:

a first decoding branch for decoding the first encoded signal based on the first coding algorithm;

a second decoding branch for decoding the first processed signal or the second processed signal,

wherein the second decoding branch comprises

a first inverse processing branch for inverse processing the first processed signal to acquire a first inverse processed signal in the second domain;

a second inverse processing branch for inverse processing the second processed signal to acquire a second inverse processed signal in the second domain;

a first combiner for combining the first inverse processed signal and the second inverse processed signal to acquire a combined signal in the second domain; and

a converter for converting the combined signal to the first domain; and

a second combiner for combining the converted signal in the first domain and the first decoded signal output by the first decoding branch to acquire a decoded output signal in the first domain.

17. Decoder of the claim 16, in which the first combiner or the second combiner comprises a switch comprising a cross fading functionality.

18. Decoder of claim 16 or 17, in which the first domain is a time domain, the second domain is an LPC domain, the third domain is an LPC spectral domain, or the first encoded signal is encoded in a fourth domain, which is a time-spectral domain acquired by time/frequency converting a signal in the first domain.

19. Decoder in accordance with claim 16, in which the first decoding branch comprises an inverse coder and a de-quantizer and a frequency domain time domain converter, or

the second decoding branch comprises an inverse coder and a de-quantizer in the first inverse processing branch or an inverse coder and a de-quantizer and an LPC spectral domain to LPC domain converter in the second inverse processing branch.

20. Decoder of claim 19, in which the first decoding branch or the second inverse processing branch comprises an overlap-adder for performing a time domain aliasing cancellation functionality.

21. Decoder in accordance with claim 16, in which the first decoding branch or the second inverse processing branch comprises a de-warper controlled by a warping characteristic comprised in the encoded audio signal.

22. Decoder in accordance with claim 16, in which the encoded signal comprises, as side information, an indication whether a coded signal is to be coded by a first encoding branch or a second encoding branch or a first processing branch of the second encoding branch or a second processing branch of the second encoding branch, and

which further comprises a parser for parsing the encoded signal to determine, based on the side information, whether a coded signal is to be processed by the first decoding branch, or the second decoding branch, or the first inverse processing branch of the second decoding branch or the second inverse processing branch of the second decoding branch.

23. Method of decoding an encoded audio signal, the encoded audio signal comprising a first coded signal, a first processed signal in a second domain, and a second processed signal in a third domain, wherein the first coded signal, the first processed signal, and the second processed signal are related to different time portions of a decoded audio signal, and wherein a first domain, the second domain and the third domain are different from each other, comprising:

decoding the first encoded signal based on a first coding algorithm;

decoding the first processed signal or the second processed signal,

wherein the decoding the first processed signal or the second processed signal comprises;

inverse processing the first processed signal to acquire a first inverse processed signal in the second domain;

inverse processing the second processed signal to acquire a second inverse processed signal in the second domain;

combining the first inverse processed signal and the second inverse processed signal to acquire a combined signal in the second domain; and

converting the combined signal to the first domain; and

combining the converted signal in the first domain and the decoded first signal to acquire a decoded output signal in the first domain.

24. Encoded audio signal comprising:

a first coded signal encoded or to be decoded using a first coding algorithm,

a first processed signal in a second domain, and a second processed signal in a third domain, wherein the first processed signal and the second processed signal are encoded using a second coding algorithm,

wherein the first coded signal, the first processed signal, and the second processed signal are related to different time portions of a decoded audio signal,

wherein a first domain, the second domain and the third domain are different from each other, and

side information indicating whether a portion of the encoded signal is the first coded signal, the first processed signal or the second processed signal.

25. Computer program for performing, when running on the computer, the method of encoding an audio signal, the audio input signal being in a first domain, comprising:

encoding an audio signal using a first coding algorithm to acquire a first encoded signal;

encoding an audio signal using a second coding algorithm to acquire a second encoded signal, wherein the first coding algorithm is different from the second coding algorithm; and

switching between encoding using the first coding algorithm and encoding using the second coding algorithm so that, for a portion of the audio input signal, either the first encoded signal or the second encoded signal is in an encoded output signal,

wherein encoding using the second coding algorithm comprises;

converting the audio signal into a second domain different from the first domain,

processing an audio signal in the second domain to acquire a first processed signal;

converting a signal into a third domain different from the first domain and the second domain and processing the signal in the third domain to acquire a second processed signal; and

switching between processing the audio signal and converting and processing so that, for a portion of the audio signal encoded using the second coding algorithm, either the first processed signal or the second processed signal is in the second encoded signal.

26. Computer program for performing, when running on the computer, the method of decoding an encoded audio signal, the encoded audio signal comprising a first coded signal, a first processed signal in a second domain, and a second processed signal in a third domain, wherein the first coded signal, the first processed signal, and the second processed signal are related to different time portions of a decoded audio signal, and wherein a first domain, the second domain and the third domain are different from each other, comprising:

decoding the first encoded signal based on a first coding algorithm;

decoding the first processed signal or the second processed signal,

wherein the decoding the first processed signal or the second processed signal comprises;

inverse processing the first processed signal to acquire a first inverse processed signal in the second domain;

inverse processing the second processed signal to acquire a second inverse processed signal in the second domain;

combining the first inverse processed signal and the second inverse processed signal to acquire a combined signal in the second domain; and

converting the combined signal to the first domain; and

combining the converted signal in the first domain and the decoded first signal to acquire a decoded output signal in the first domain.