Frequency-domain audio coding supporting transform length switching

Patent metadata

Field	Value
Publication number	US-11862182-B2
Application number	US-202117227178-A
Country	US
Kind code	B2
Filing date	Apr 9, 2021
Priority date	Jul 22, 2013
Publication date	Jan 2, 2024
Grant date	Jan 2, 2024

Abstract

A frequency-domain audio codec is provided with the ability to additionally support a certain transform length in a backward-compatible manner, by the following: the frequency-domain coefficients of a respective frame are transmitted in an interleaved manner irrespective of the signalization signaling for the frames as to which transform length actually applies, and additionally the frequency-domain coefficient extraction and the scale factor extraction operate independent from the signalization. By this measure, old-fashioned frequency-domain audio coders/decoders, insensitive for the signalization, would be able to nevertheless operate without faults and with reproducing a reasonable quality. Concurrently, frequency-domain audio coders/decoders able to support the additional transform length would offer even better quality despite the backward compatibility. As far as coding efficiency penalties due to the coding of the frequency domain coefficients in a manner transparent for older decoders are concerned, same are of comparatively minor nature due to the interleaving.

First claim

The invention claimed is:

1. An audio decoder, comprising an electronic circuit, programmed computer or microprocessor configured to:
extract a sequence of frequency-domain coefficients relating to a frame of an audio signal from a data stream;
extract scale factors for the frame from the data stream;
subject the frequency-domain coefficients, scaled according to the scale factors, to inverse transformation to obtain a time-domain portion of the audio signal corresponding to the frame;
subject the time-domain portion to an overlap-add process to obtain the audio signal,
wherein the audio decoder is responsive to a signalization within the data stream for the frame of the audio signal so as to, depending on the signalization,
form one transform out of the sequence of frequency-domain coefficients by maintaining a sequential order of the frequency-domain coefficients in the sequence of frequency-domain coefficients and subject the one transform, scaled according to the scale factors, to an inverse transformation of a first transform length, or
form more than one transform by de-interleaving the frequency-domain coefficients from the sequence of frequency-domain coefficients and subject, scaled according to the scale factors, each of the more than one transforms to an inverse transformation of a second transform length, shorter than the first transform length,
wherein the audio decoder is configured to extract the sequence of frequency-domain coefficients and the scale factors from the data stream so that a number of the sequence of frequency-domain coefficients and a number of the scale factors and an order at which the sequence of frequency-domain coefficients are extracted from the data stream and an order at which the scale factors are extracted from the data stream are independent from the signalization.

2. The audio decoder of claim 1 , configured to use context based entropy decoding to extract the sequence of frequency-domain coefficients from the data stream, with assigning, for each frequency-domain coefficient, a context to the respective frequency-domain coefficient in a manner independent from the signalization.

3. The audio decoder of claim 1 , configured to subject the frequency-domain coefficients to scaling according to the scale factors at a spectral resolution independent from the signalization.

4. The audio decoder of claim 1 , configured to subject the sequence of frequency-domain coefficients to noise filling, at a spectral resolution which is independent from the signalization.

5. The audio decoder of claim 1 , configured to:
in the formation of the one transform, apply inverse temporal noise shaping filtering on the sequence of frequency-domain coefficients using the sequential order, and
in the formation of the more than one transforms, apply inverse temporal noise shaping filtering on the sequence of frequency-domain coefficients by de-interleaving the frequency-domain coefficients of the more than one transforms, concatenating of the more than one transforms spectrally transform-wise to yield a concatenation of the more than one transforms and applying the inverse temporal noise shaping filtering on the concatenation of the more than one transforms.

6. The audio decoder of claim 1 , configured to support joint-stereo coding with or without inter-channel stereo prediction and to use the sequence of frequency-domain coefficients as a sum or difference spectrum or prediction residual of the inter-channel stereo prediction.

7. The audio decoder of claim 1 , wherein the number of the more than one transforms equals 2, and the first transform length is twice the second transform length.

8. The audio decoder of claim 1 , wherein the inverse transformation is an inverse modified discrete cosine transform (MDCT).

9. An audio encoder, comprising an electronic circuit, programmed computer or microprocessor configured to:
subject time-domain portions of an audio signal to transformation to obtain, for each time-domain portion, frequency-domain coefficients;
inversely scale the frequency-domain coefficients according to scale factors,
wherein the audio encoder configured to switch, for a predetermined frame, between
performing one transform of a first transform length, and
performing more than one transform of a second transform length, shorter than the first transform length;

signal the switching for the predetermined frame by a signalization within the data stream for the predetermined frame,
wherein the audio encoder is configured to:
arrange, for the predetermined frame, the frequency-domain coefficients, inversely scaled according to scale factors, at a sequential order so as to obtain a sequence of frequency-domain coefficients by, depending on the signalization,
sequentially arranging the frequency-domain coefficients of the one transform in case of one transform performed for the respective frame, and
by interleaving the frequency-domain coefficients of the more than one transform of the respective frame in case of more than one transform performed for the respective frame, so that spectrally corresponding frequency-domain coefficients of the more than two transforms immediately follow each other; and

insert the sequence of frequency domain coefficients and scale factors into the data stream so that the sequence of frequency-domain coefficients and the scale factors are inserted into data stream so that a number of the sequence of frequency-domain coefficients and a number of the scale factors and an order at which the sequence of frequency-domain coefficients are inserted the data stream and an order at which the scale factors are inserted into the data stream are independent from the signalization.

10. A method for audio decoding, comprising:
extracting a sequence of frequency-domain coefficients relating to a frame of an audio signal from a data stream;
extracting scale factors for the frame from the data stream;
subjecting the frequency-domain coefficients, scaled according to scale factors, to inverse transformation to obtain a time-domain portion of the audio signal; and
subjecting the time-domain portion to an overlap-add process to obtain the audio signal,
wherein the subjection to inverse transformation is responsive to a signalization within the data stream for the frame so as to, depending on the signalization, comprise:
forming one transform out of the sequence of frequency-domain coefficients by maintaining a sequential order of the frequency-domain coefficients in the sequence of frequency-domain coefficients and subjecting the one transform, scaled according to the scale factors, to an inverse transformation of a first transform length, or
forming more than one transform by de-interleaving the frequency-domain coefficients from the sequence of frequency-domain coefficients and subjecting, scaled according to the scale factors, each of the more than one transforms to an inverse transformation of a second transform length, shorter than the first transform length,
wherein the extraction of the frequency-domain coefficients and the extraction of the scale factors are performed so that a number of the sequence of frequency-domain coefficients and a number of the scale factors and an order at which the sequence of frequency-domain coefficients are extracted from the data stream and an order at which the scale factors are extracted from the data stream are independent from the signalization.

11. A method for audio encoding, comprising:
subjecting time-domain portions of an audio signal to transformation to obtain, for each time-domain portion, frequency-domain coefficients;
inversely scaling the frequency-domain coefficients according to

Assignees

• Fraunhofer Ges Forschung

Inventors

• Dick Sascha
• Helmrich Christian
• Hoelzer Andreas

Classifications

• H03M7/30

  Compression (speech analysis-synthesis for redundancy reduction G10L19/00; for image communication H04N); Expansion; Suppression of unnecessary data, e.g. redundancy reduction · CPC title

• G10L19/022 Primary

  Blocking, i.e. grouping of samples in time; Choice of analysis windows; Overlap factoring · CPC title

• G10L19/03

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

• G10L19/008

  Multichannel audio signal coding or decoding using interchannel correlation to reduce redundancy, e.g. joint-stereo, intensity-coding or matrixing · CPC title

• G10L19/028

  Noise substitution, i.e. substituting non-tonal spectral components by noisy source (comfort noise for discontinuous speech transmission G10L19/012) · CPC title