Method and apparatus for processing an audio signal, audio decoder, and audio encoder to filter a discontinuity by a filter which depends on two fir filters and pitch lag

Patent metadata

Field	Value
Publication number	US-12165665-B2
Application number	US-202318537655-A
Country	US
Kind code	B2
Filing date	Dec 12, 2023
Priority date	Jul 28, 2014
Publication date	Dec 10, 2024
Grant date	Dec 10, 2024

Abstract

A method is described that processes an audio signal. A discontinuity between a filtered past frame and a filtered current frame of the audio signal is removed using linear predictive filtering.

First claim

The invention claimed is:

1. A method for decoding an audio signal, the method comprising: receiving an encoded audio signal; and
generating a decoded audio signal;
wherein generating the decoded the audio signal comprises:
decoding the encoded audio signal to acquire the decoded audio signal,
filtering audio frames of the decoded audio signal using the following filter:
H ( z )=(1− B ( z ))/(1− C ( z )· z −T )

where B(z) and C(z) are two FIR filters and the H(z) filter parameters are the coefficients of the FIR filters B(z) and C(z), and T indicates the pitch lag, and
removing a discontinuity between a filtered past frame and a filtered current frame by modifying a beginning portion of the filtered current frame by a signal acquired by linear predictive filtering a predefined signal with initial states of a linear predictive filter, the initial states defined on basis of a last part of an unfiltered past frame filtered using a set of current frame filter parameters for filtering the current frame.

2. The method of claim 1 , further comprising estimating the linear predictive filter on the filtered or non-filtered audio signal.

3. The method of claim 2 , wherein estimating the linear predictive filter comprises estimating the filter based on the past and/or current frame of the audio signal or based on the past filtered frame of the audio signal using the Levinson-Durbin algorithm.

4. The method claim 1 , wherein the linear predictive filter comprises a linear predictive filter of an audio codec.

5. A non-transitory digital storage medium having a computer program stored thereon to perform the method for decoding an audio signal, the method comprising:
receiving an encoded audio signal;
and generating a decoded audio signal;
wherein generating the decoded audio signal comprises:
decoding the encoded audio signal to acquire the decoded audio signal,
filtering audio frames of the decoded audio signal using the following filter:
H ( z )=(1− B ( z ))/(1− C ( z )· z −T )

where B(z) and C(z) are two FIR filters and the H(z) filter parameters are the coefficients of the FIR filters B(z) and C(z), and T indicates the pitch lag, and
removing a discontinuity between a filtered past frame and a filtered current frame by modifying a beginning portion of the filtered current frame by a signal acquired by linear predictive filtering a predefined signal with initial states of a linear predictive filter, the initial states defined on basis of a last part of an unfiltered past frame filtered using a set of current frame filter parameters for filtering the current frame,
when said computer program is run by a computer.

6. A decoder for decoding an audio signal, the apparatus comprising:
an input for receiving an encoded audio signal; and
a processor configured to generate a decoded audio signal,
wherein, for generating the decoded audio signal, the processor is configured to:
decode the encoded audio signal to acquire the decoded audio signal,
filter audio frames of the decoded audio signal using the following filter:
H ( z )=(1− B ( z ))/(1− C ( z )· z −T )

where B(z) and C(z) are two FIR filters and the H(z) filter parameters are the coefficients of the FIR filters B(z) and C(z), and T indicates the pitch lag, and
remove a discontinuity between a filtered past frame and a filtered current frame by modifying a beginning portion of the filtered current frame by a signal acquired by linear predictive filtering a predefined signal with initial states of a linear predictive filter, the initial states defined on basis of a last part of an unfiltered past frame filtered using a set of current frame filter parameters for filtering the current frame.

Assignees

• Fraunhofer Ges Forschung

Inventors

• Ravelli Emmanuel
• Jander Manuel
• Pietrzyk Grzegorz
• Dietz Martin
• Gayer Marc

Classifications

• G10L19/12

  the excitation function being a code excitation, e.g. in code excited linear prediction [CELP] vocoders · CPC title

• H04B1/1027

  assessing signal quality or detecting noise/interference for the received signal · CPC title

• G10L19/03

  Spectral prediction for preventing pre-echo; Temporary noise shaping [TNS], e.g. in MPEG2 or MPEG4 · CPC title

• G10L19/005

  Correction of errors induced by the transmission channel, if related to the coding algorithm · CPC title

• G11B27/038

  Cross-faders therefor · CPC title