CN101552008B - Voice coding method, coding device, decoding method and decoding device - Google Patents
Voice coding method, coding device, decoding method and decoding device Download PDFInfo
- Publication number
- CN101552008B CN101552008B CN2008100891481A CN200810089148A CN101552008B CN 101552008 B CN101552008 B CN 101552008B CN 2008100891481 A CN2008100891481 A CN 2008100891481A CN 200810089148 A CN200810089148 A CN 200810089148A CN 101552008 B CN101552008 B CN 101552008B
- Authority
- CN
- China
- Prior art keywords
- present frame
- envelope information
- temporal envelope
- frame
- information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 title claims abstract description 83
- 230000002123 temporal effect Effects 0.000 claims description 314
- 238000001228 spectrum Methods 0.000 claims description 56
- 230000009466 transformation Effects 0.000 claims description 42
- 239000000284 extract Substances 0.000 claims description 38
- 238000010606 normalization Methods 0.000 claims description 32
- 238000000605 extraction Methods 0.000 claims description 21
- 230000005540 biological transmission Effects 0.000 claims description 20
- 238000011002 quantification Methods 0.000 claims description 13
- 230000007704 transition Effects 0.000 claims description 12
- 238000004321 preservation Methods 0.000 claims description 5
- 238000012937 correction Methods 0.000 claims description 4
- 230000008569 process Effects 0.000 description 23
- 230000008447 perception Effects 0.000 description 11
- 238000006243 chemical reaction Methods 0.000 description 9
- 238000010586 diagram Methods 0.000 description 6
- 230000000717 retained effect Effects 0.000 description 6
- 238000004891 communication Methods 0.000 description 4
- 238000001914 filtration Methods 0.000 description 4
- 238000012545 processing Methods 0.000 description 3
- 230000009467 reduction Effects 0.000 description 3
- 230000001771 impaired effect Effects 0.000 description 2
- 230000000873 masking effect Effects 0.000 description 2
- 239000000654 additive Substances 0.000 description 1
- 230000000996 additive effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000012856 packing Methods 0.000 description 1
- 238000013139 quantization Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L19/00—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis
- G10L19/04—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis using predictive techniques
Landscapes
- Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Compression, Expansion, Code Conversion, And Decoders (AREA)
Abstract
The invention relates to a data coding technique and discloses a coding method, a coding device, a decoding method and a decoding device. The coding method comprises the following steps: obtaining quantified time domain enveloping information of a current frame; dividing the quantified time domain enveloping information into at least two parts including a current frame part and a subsequent frame;combining the current frame part and the subsequent frame of a frame before to obtain time domain enveloping information of the current frame to be coded and transmitted; and coding and transmitting the time domain enveloping information of the current frame to be coded and transmitted. Using the technical scheme provided by the invention, partial information of the current frame can be utilized to carry out decoding when the current frame is lost or damaged.
Description
Technical field
The present invention relates to data encoding technique, be specifically related to coding method and device, coding/decoding method and device.
Background technology
In communication system, can set up a loop between two conversational terminals, the packet that will comprise voice messaging by this loop transmitting terminal is sent to receiving end, again at the receiving end synthetic speech signal.If losing of packet in transport process, occurred, need recover the voice messaging that lost data packets comprises by packet or the additive method that received in the past, otherwise can bring the sense of discomfort of the sense of hearing.Along with development of Communication Technique, now a lot of voice communications much are to connect, transmit by Internet protocol (IP:Internet Protocol) network, wireless network, and under this communication environment, the probability that data-bag lost occurs is bigger.
In narrowband speech scrambler, wideband acoustic encoder, adopted various forward error corrections (FEC:Forward Error Correction) strategy at present, the problem that voice quality descends when preferably resolving LOF.But signal for the ultra broadband part, because signal is the bigger difference that has of the feature in the bandwidth, parameter etc. and arrowband, broadband part in this section, the ultra broadband part effect that is used for the FEC strategy that is used for arrowband, broadband part differs and reserves, and the signal for the ultra broadband part does not then also have a kind of strategy of FEC preferably at present.
Existing a kind of disposal route for ultra-broadband signal is such: if LOF or impaired situation repeat the decoded parameter of former frame, promptly use the parameter of the decoded parameter of former frame as present frame, carry out subsequent decoding again and handle.
Though voice signal has smooth performance in short-term, the signal of adjacent two frames exists certain correlativity, but said method directly utilizes the parameter of former frame to decode as the parameter of present frame, promptly do not comprise the due information of any present frame, the signal that then final decoding obtains still has more tangible discontinuous sense, and auditory perception is uncomfortable.
Summary of the invention
The embodiment of the invention provides coding method and device, coding/decoding method and device, and the technical scheme of using the embodiment of the invention to provide can use the partial information of present frame to decode when present frame is lost or be impaired.
The purpose of the embodiment of the invention is achieved through the following technical solutions:
The embodiment of the invention provides a kind of voice coding method, comprising:
Obtain the time-domain signal of present frame, extract the temporal envelope information of described time-domain signal, the temporal envelope information of extracting is quantized to obtain quantified time domain enveloping information;
Described quantified time domain enveloping information is divided into two parts at least, comprises present frame part and subsequent frame part;
Make up the described present frame part and the subsequent frame part of frame before, obtain the temporal envelope information that present frame need be encoded and be transmitted;
Coding also sends the temporal envelope information that described present frame need be encoded and be transmitted.
The embodiment of the invention also provides a kind of tone decoding method, comprising:
Decoding obtains the temporal envelope information that the present frame coding transmits;
From the temporal envelope information that described present frame coding transmits, extract the present frame part in the temporal envelope information of described present frame;
Decoding obtains the temporal envelope information of the subsequent frame coding transmission of described present frame; From the temporal envelope information that described subsequent frame coding transmits, extract the subsequent frame part in the temporal envelope information of described present frame;
Make up the temporal envelope information that described present frame part and described subsequent frame partly obtain described present frame.
The embodiment of the invention also provides a kind of tone decoding method, comprising:
Decoding obtains the temporal envelope information of the subsequent frame coding transmission of present frame;
From the temporal envelope information that described subsequent frame coding transmits, extract the subsequent frame part in the temporal envelope information of described present frame;
Use subsequent frame in the temporal envelope information of described present frame partly to obtain the temporal envelope information of described present frame.
The embodiment of the invention also provides a kind of sound encoding device, comprising:
The temporal envelope information acquisition unit is used to obtain the time-domain signal of present frame, extracts the temporal envelope information of described time-domain signal, and the temporal envelope information of extracting is quantized to obtain quantified time domain enveloping information;
Split cells is used for the quantified time domain enveloping information that described temporal envelope information acquisition unit is obtained is divided into two parts at least, comprises present frame part and subsequent frame part;
Assembled unit is used to make up the described present frame part and the subsequent frame part of frame data before, obtains the temporal envelope information that present frame need be encoded and be transmitted;
The coding transmitting element is used to encode and sends described assembled unit and makes up the temporal envelope information that the described present frame that obtains need be encoded and be transmitted.
The embodiment of the invention also provides a kind of audio decoding apparatus, comprising:
Decoding unit, being used to decode obtains the temporal envelope information that present frame coding transmits;
Extraction unit is used for the temporal envelope information that transmits from described present frame coding, extracts the present frame part in the temporal envelope information of described present frame;
Acquiring unit, the temporal envelope information that the subsequent frame coding that obtains described present frame of being used to decode transmits from the temporal envelope information that described subsequent frame coding transmits, is extracted the subsequent frame part in the temporal envelope information of described present frame;
Assembled unit is used to make up the temporal envelope information that described present frame part and described subsequent frame partly obtain described present frame.
The embodiment of the invention also provides a kind of audio decoding apparatus, comprising:
Decoding unit, the temporal envelope information that the subsequent frame coding that obtains described present frame of being used to decode transmits;
Extraction unit is used for the temporal envelope information that transmits from described subsequent frame coding, extracts the subsequent frame part in the temporal envelope information of described present frame;
The temporal envelope information acquisition unit is used for using the subsequent frame of the temporal envelope information of described present frame partly to obtain the temporal envelope information of described present frame.
The above technical scheme that provides from the embodiment of the invention as can be seen, because present embodiment will be worked as each frame and all will be divided into two parts at least, comprise present frame part and subsequent frame part, and present frame part and subsequent frame partly are placed in the coding of at least two frames and divide transmission, have only present frame partly to belong to present frame in the temporal envelope information that coding sends like this, and remainder is placed in the temporal envelope information of subsequent frame and sends, if thereby decoding end can not be correctly decoded present frame, also can from subsequent frame, extract the temporal envelope information that belongs to present frame, and then can use the temporal envelope information that belongs to present frame to decode, because the decoding that the temporal envelope information that belongs to present frame that is to use is carried out, therefore the continuity of the holding signal that can try one's best makes auditory perception comfortable.
Description of drawings
Fig. 1 is the process flow diagram of coding method embodiment one in the embodiment of the invention;
Fig. 2 is the process flow diagram of coding method embodiment two in the embodiment of the invention;
Fig. 3 is the process flow diagram of coding/decoding method embodiment one in the embodiment of the invention;
Fig. 4 is the process flow diagram of coding/decoding method embodiment two in the embodiment of the invention;
Fig. 5 is the process flow diagram of coding/decoding method embodiment four in the embodiment of the invention;
Fig. 6 is the process flow diagram of coding/decoding method embodiment five in the embodiment of the invention;
Fig. 7 is the structural drawing of code device embodiment in the embodiment of the invention;
Fig. 8 is the structural drawing of decoding device embodiment one in the embodiment of the invention;
Fig. 9 is the structural drawing of decoding device embodiment two in the embodiment of the invention.
Embodiment
For make purpose of the present invention, technical scheme, and advantage clearer, below with reference to the accompanying drawing embodiment that develops simultaneously, the present invention is described in more detail.
Fig. 1 has described the flow process of the coding method embodiment one that the embodiment of the invention provides, comprising:
In obtaining the quantified time domain enveloping information process, need to obtain earlier the time-domain signal of present frame, extract the temporal envelope information of time-domain signal then, the temporal envelope information of extracting is quantized, thereby obtain quantified time domain enveloping information;
The described time-domain signal of the embodiment of the invention can be the time-domain signal of arrowband part, the time-domain signal of broadband part, the time-domain signal of ultra broadband part etc.;
The extracting mode of temporal envelope information can have multiple, and the embodiment of the invention does not limit the concrete mode of extracting temporal envelope information; Temporal envelope information is one group of information, and the quantity of this group information can be predefined, and the quantity of this group information of each frame is all identical; The quantity of this group information also can be according to the input signal characteristics determined of different frame, and this moment, this quantity of organizing information of different frame may be different;
The quantification manner of quantification acquisition quantified time domain enveloping information also can have multiple, and the embodiment of the invention does not limit;
Quantified time domain enveloping information is one group of signal, and the embodiment of the invention does not limit how to be divided into present frame part and subsequent frame part, and does not also limit the quantity of the subsequent frame part that is divided into; But the embodiment of the invention is preferably with the quantified time domain enveloping information separated into two parts, i.e. a present frame part and a subsequent frame part; With the quantified time domain enveloping information separated into two parts time, the embodiment of the invention preferably is divided into present frame part and subsequent frame part according to the mode of odd even with quantified time domain enveloping information, for example the present frame part can be the odd number part or the even number part of quantified time domain enveloping information, and remainder is as the subsequent frame part.With temporal envelope information separated into two parts the time, have only a subsequent frame can carry subsequent frame part in the temporal envelope information of present frame; When being divided into temporal envelope information more than two parts, then can there be at least two frames to carry subsequent frame part in the temporal envelope information of present frame.
Step 103, combination present frame part and the subsequent frame part of frame data before obtain the temporal envelope information that present frame need be encoded and be transmitted;
With the present frame of present frame part and before the subsequent frame of frame when partly making up, can combined crosswise, also can segmentation directly make up, the embodiment of the invention does not limit specifically and how to make up;
In the quantity of temporal envelope information is during according to the input signal characteristics determined, can determine the quantity of the temporal envelope information of each frame in order to make decoding end, can be when combination obtains present frame and need encode the temporal envelope information that transmits, with the quantity of the temporal envelope information of the quantity of the temporal envelope information of present frame and next frame, also be combined in present frame and need encode in the temporal envelope information that transmits;
When present frame is the 1st frame, because the subsequent frame part of frame data before not have, therefore can set in advance the subsequent frame part of frame data before of the 1st frame data, for example can be set at complete 0 or complete 1 etc.;
Step 104, coding also send the temporal envelope information that present frame need be encoded and be transmitted.
From the above, present embodiment will be worked as each frame and all will be divided into two parts at least, comprise present frame part and subsequent frame part, and present frame part and subsequent frame partly are placed in the coding of at least two frames and divide transmission, have only present frame partly to belong to present frame in the temporal envelope information that coding sends like this, and remainder is placed in the temporal envelope information of subsequent frame and sends, if thereby decoding end can not be correctly decoded present frame, also can from subsequent frame, extract the temporal envelope information that belongs to present frame, and then can use the temporal envelope information that belongs to present frame to decode, because the decoding that the temporal envelope information that belongs to present frame that is to use is carried out, therefore the continuity of the holding signal that can try one's best makes auditory perception comfortable.
Further, for the signal that makes transmission is more accurate, after obtaining quantified time domain enveloping information, can further obtain and send frequency-region signal and frequency domain spectra parameter, so the embodiment of the invention also provides the embodiment two of coding method, Fig. 2 has described the flow process of this embodiment two, comprising:
The temporal envelope information of step 202, extraction time-domain signal;
Step 203, quantification temporal envelope information obtain quantified time domain enveloping information;
Step 204, quantified time domain enveloping information is divided into two parts at least, comprises present frame part and subsequent frame part;
Step 205, combination present frame part and the subsequent frame part of frame data before obtain the temporal envelope information that present frame need be encoded and be transmitted;
The frequency-region signal of step 206, use quantified time domain enveloping information acquisition present frame;
Concrete, can carry out level and smooth interpolation to quantified time domain enveloping information and handle, obtain the temporal envelope after level and smooth; Use the temporal envelope after level and smooth that time-domain signal is carried out normalized, obtain the time-domain signal after the normalization; Time-domain signal after the time-domain signal after the use normalization and the normalization of previous frame data constitutes array; Array is carried out the frequency-region signal that transition coding obtains described present frame;
Wherein, the embodiment of the invention does not limit the method for level and smooth interpolating method, normalized and the mode of transition coding; The preferred transition coding mode of the embodiment of the invention is the discrete cosine transform (MDCT:Modified Discrete Cosine Transform) of 50% overlapping correction;
The frequency domain envelope information of step 207, extraction frequency-region signal is chosen the frequency domain spectra parameter;
After obtaining frequency-region signal, can extract the frequency domain envelope information of frequency-region signal, the embodiment of the invention does not limit frequency domain envelope information how to extract frequency-region signal;
When choosing the frequency domain spectra parameter, specifically can select according to people's ear masking effect;
Step 208, quantize frequency domain envelope information and described frequency domain spectra parameter, the frequency domain spectra parameter after frequency domain envelope information after obtaining quantizing and the quantification;
Step 209, coding also send temporal envelope information, the frequency domain envelope information after the quantification and the frequency domain spectra parameter after the quantification that present frame need be encoded and be transmitted.
From the above, present embodiment will be worked as each frame and all will be divided into two parts at least, comprise present frame part and subsequent frame part, and present frame part and subsequent frame partly are placed in the coding of at least two frames and divide transmission, have only present frame partly to belong to present frame in the temporal envelope information that coding sends like this, and remainder is placed in the temporal envelope information of subsequent frame and sends, if thereby decoding end can not be correctly decoded present frame, also can from subsequent frame, extract the temporal envelope information that belongs to present frame, and then can use the temporal envelope information that belongs to present frame to decode, because the decoding that the temporal envelope information that belongs to present frame that is to use is carried out, therefore the continuity of the holding signal that can try one's best makes auditory perception comfortable.
The embodiment of the invention also provides the embodiment three of coding method, the process that the embodiment of the invention is encoded to ultra-broadband signal has been described, what present embodiment was described is the situation that temporal envelope information to be quantified is divided into a present frame part and a subsequent frame part, and flow process is as follows:
Suppose that present frame is the k frame, and the ultra broadband of k frame input time-domain signal is { y_hi
k(0), y_hi
k(1), L L, y_hi
k(319) }, extract the one group envelope information { A of this ultra-broadband signal in time domain
0, k, A
1, k, L L A
N, k; The number of the temporal envelope that extracts can be fixed, and each frame is all identical; Also can determine according to current input signal feature.The embodiment of the invention does not limit the account form of temporal envelope, supposes N=9 in the present embodiment, is 2ms corresponding to the temporal resolution of temporal envelope, and then temporal envelope information can adopt
Or
Calculate etc. mode.
This group temporal envelope information of extracting is quantized, can obtain quantified time domain enveloping information { A
0, k q, A
1, k q, L L A
9, k q.
The quantified time domain enveloping information of k frame signal is divided into odd number part { A
0, k q, A
2, k q, A
4, k q, A
6, k q, A
8, k qAnd even number part { A
1, k q, A
3, k q, A
5, k q, A
7, k q, A
9, k q; need to prove; quantified time domain enveloping information is divided into odd number part and even number partly is a kind of more excellent implementation that the embodiment of the invention provides, but the embodiment of the invention does not limit and quantified time domain enveloping information can only be divided into odd number partly and the even number part.Wherein, with even number part as the present frame part, with the odd number part as the subsequent frame part.Even number part { A with the k frame signal
0, k q, A
2, k q, A
4, k q, A
6, k q, A
8, k qAnd the previous frame signal be kept at odd number part { A in the encoding state device
1, k-1 q, A
3, k-1 q, A
5, k-1 q, A
7, k-1 q, A
9, k-1 qTo reconfigure be the temporal envelope information { A that the k frame need be encoded and be sent
0, k q, A
2, k q, A
4, k q, A
6, k q, A
8, k q, A
1, k-1 q, A
3, k-1 q, A
5, k-1 q, A
7, k-1 q, A
9, k-1 q.Odd number part { A with this frame signal
1, k q, A
3, k q, A
5, k q, A
7, k q, A
9, k qBe kept in the encoding state register, be used for the temporal envelope coding of k+1 frame signal.
To this frame signal quantified time domain enveloping information { A
0, k q, A
1, k qL L A
9, k qCarry out level and smooth interpolation and handle, obtain the temporal envelope after level and smooth
Wherein, the embodiment of the invention does not limit the concrete grammar of level and smooth interpolation, and a kind of level and smooth interpolating method that the embodiment of the invention provides is as follows
J=0L 319, i=-1L N, wherein
Utilize level and smooth temporal envelope information afterwards that time-domain signal is carried out normalized again, can adopt following formula to carry out normalized:
Thereby obtain the time-domain signal { y_hi after the normalization
k' (0), y_hi
k' (1), L L, y_hi
k' (319) }.
With the time-domain signal { y_hi after the normalization
k' (0), y_hi
k' (1), L L, y_hi
k' (319) } and be kept at k-1 frame signal { y_hi in the encoding state register
K-1' (0), y_hi
K-1' (1), L L, y_hi
K-1' (319) } constitute one group 640 array:
{ y_hi
K-1' (0), y_hi
K-1' (1), L L, y_hi
K-1' (319), y_hi
k' (0), y_hi
k' (1), L L, y_hi
k' (319) }, then this array is carried out 50% overlapping MDCT conversion, obtain the frequency-region signal { y_swb (0), y_swb (1), L L, y_swb (319) } of k frame.
Extract the frequency domain envelope information of the frequency-region signal { y_swb (0), y_swb (1), L L, y_swb (319) } of k frame, select one group of frequency domain spectra parameter according to people's ear masking effect; Frequency domain envelope information and frequency domain spectra parameter are quantized and encode.
K frame behind coding temporal envelope information, frequency domain envelope information, the frequency domain spectra parameter that transmits of need encoding added packing transmission in the last code stream.
Further, in the quantity of temporal envelope information during by the input signal characteristics determined, when combination obtained present frame and need encode the temporal envelope information that sends, the quantity of the temporal envelope information of two frames that the temporal envelope information that the needs coding can be sent is related also was combined in present frame and need encodes in the temporal envelope information that sends;
Suppose that the quantity of the temporal envelope information of k frame is L
k, the quantity of the temporal envelope information of k-1 frame is L
K-1, then with the { A of this frame signal
0, k q, A
2, k q, Λ Λ } and indication present frame extraction temporal envelope number L
kBe kept at { A in the encoding state device with the previous frame signal
1, k-1 q, A
3, k-1 q, Λ Λ } and indication previous frame extraction temporal envelope number L
K-1Reconfigure and be { L
k, A
0, k q, A
2, k q, Λ Λ, L
K-1, A
1, k-1 q, A
3, k-1 q, Λ Λ }, encode as the present frame temporal envelope information that transmits of need encoding.{ L with this frame signal
k, A
1, k q, A
3, k q, Λ Λ, } be kept in the encoding state register, be used for the temporal envelope coding of next frame signal.
Be understandable that need the encode array mode of the temporal envelope information that transmits of above-mentioned present frame is a kind of embodiment that the embodiment of the invention proposes, and is not to limit to adopt aforesaid way to make up; As long as the different combined information of the differentiation that decoding end can be correct just can not influence the realization of the embodiment of the invention.The above-mentioned situation that temporal envelope information is divided into a present frame part and a subsequent frame part of only having described, the treatment scheme that temporal envelope information is divided into a present frame part and a plurality of subsequent frame part is similar, repeats no more herein.
Fig. 3 has described the flow process of the coding/decoding method embodiment one that the embodiment of the invention provides, comprising:
Step 301, decoding obtain the temporal envelope information that the present frame coding transmits;
Step 302, from the temporal envelope information that present frame coding transmits, extract the present frame part in the temporal envelope information of present frame;
The temporal envelope information that present frame in embodiment of the invention coding transmits be comprise present frame part in the temporal envelope information of present frame and present frame before subsequent frame part in the temporal envelope information of frame; Therefore, can be from the temporal envelope information that the present frame coding transmits, the present frame part in the temporal envelope information of extraction present frame;
Further, if the quantity of the temporal envelope information of present frame is by the input signal characteristics determined, then the temporal envelope information that transmits of present frame coding also can comprise the quantity of temporal envelope information of present frame and present frame before the quantity of temporal envelope information of frame; At this moment, can be according to the quantity of the temporal envelope information of described present frame, from the temporal envelope information that described present frame coding transmits, extract the present frame part in the temporal envelope information of described present frame;
Step 303, obtain the subsequent frame part of the temporal envelope information of present frame;
The subsequent frame part of the temporal envelope information of present frame can obtain by following flow process: decoding obtains the temporal envelope information of the subsequent frame coding transmission of present frame; From the temporal envelope information that the subsequent frame coding transmits, the subsequent frame part in the temporal envelope information of extraction present frame; When temporal envelope information only has been divided into two parts, can from a subsequent frame, obtain the subsequent frame part; When temporal envelope information has been divided into more than two parts, can from least two subsequent frames, obtain the subsequent frame part;
If present frame is a last frame, when promptly not having subsequent frame, the subsequent frame of present frame does not partly send, can use preset mode to obtain the subsequent frame part of the temporal envelope information of present frame this moment, specifically can use the present frame of the temporal envelope information of present frame partly to obtain the whole temporal envelope information of present frame, for example can also can use the information of present frame part to carry out the whole temporal envelope information that the interpolation processing obtains present frame in modes such as the position of subsequent frame temporal envelope information partly fillings 0 or 1;
Step 304, combination present frame part and subsequent frame partly obtain the temporal envelope information of present frame.
From the above, have only present frame partly to belong to present frame in the temporal envelope information of the present frame that present embodiment receives, remainder is transmitted by other frames, even therefore present frame can not be correctly decoded or lose, also can from other frames, obtain the remainder of the temporal envelope information of present frame, and then can use the remainder of the temporal envelope information that belongs to present frame to decode, because the decoding that the temporal envelope information that belongs to present frame that is to use is carried out, therefore the continuity of the holding signal that can try one's best makes auditory perception comfortable; Simultaneously, when the present frame normal transmission, also can the information of present frame be correctly decoded, can not influence the continuity of signal.
Fig. 4 has described the flow process of the coding/decoding method embodiment two that the embodiment of the invention provides, comprising:
Step 401, decoding obtain temporal envelope information, frequency domain envelope information and the frequency domain spectra parameter information that the present frame coding transmits;
Step 402, from the temporal envelope information that present frame coding transmits, extract the present frame part in the temporal envelope information of present frame;
This step specifically can comprise: generate the random noise spectrum, use the frequency domain envelope information that the random noise spectrum is carried out the energy adjustment, calculate the spectrum envelope information through energy adjustment random noise spectrum; The frequency domain spectra parameter information is substituted spectrum envelope information, obtain the reconstructed frequency domain signal of present frame;
When the random noise spectrum is carried out the energy adjustment, the spectrum envelope information of random noise spectrum can be adjusted to the envelope information that meets original signal;
After having obtained the reconstructed frequency domain signal of present frame, can further preserve the reconstructed frequency domain signal of present frame;
After obtaining the reconstruct time-domain signal, the M signal of the present frame that obtains in the time of can further preserving inverse transformation reconstruct;
Carry out after the inverse transformation reconstruct M signal of the present frame that obtains can be with inverse transformation reconstruct the time each time and preserve, be convenient to use when next frame carries out inverse transformation reconstruct;
The mode of inverse transformation reconstruct is corresponding with the mode of the transition coding that when coding adopts, and what for example adopt when coding is 50% MDCT conversion, then when inverse transformation reconstruct, adopt corresponding anti-MDCT conversion and 50% overlap-add;
Step 407, the temporal envelope information of present frame is carried out the operation of level and smooth interpolation, obtain the temporal envelope information after level and smooth;
Step 408, use the reconstruct time-domain signal and level and smooth after temporal envelope information, carry out the present frame information that anti-normalization operation obtains reconstruct.
From the above, have only partial information to belong to present frame in the temporal envelope information of the present frame that present embodiment receives, remainder is transmitted by other frames, even therefore present frame can not be correctly decoded or lose, also can from other frames, obtain the remainder of the temporal envelope information of present frame, and then can use the remainder of the temporal envelope information that belongs to present frame to decode, because the decoding that the temporal envelope information that belongs to present frame that is to use is carried out, therefore the continuity of the holding signal that can try one's best makes auditory perception comfortable; Simultaneously, when the present frame normal transmission, also can the information of present frame be correctly decoded, can not influence the continuity of signal.
The embodiment of the invention also provides the embodiment three of coding/decoding method, the situation when having described with temporal envelope information separated into two parts, and flow process is as follows:
Decoding obtains the temporal envelope information { A of current k frame from the code stream that receives
0, k q, A
2, k q, A
4, k q, A
6, k q, A
8, k q, A
1, k-1 q, A
3, k-1 q, A
5, k-1 q, A
7, k-1 q, A
9, k-1 q, frequency domain envelope information and frequency domain spectra parameter information since in decoding end in order to carry out the MDCT inverse transformation, have the time delay of a frame, so the temporal envelope information { A that when decoding end is decoded the temporal envelope information of k frame, can decode and obtain the k+1 frame
0, k+1 q, A
2, k+1 q, A
4, k+1 q, A
6, k+1 q, A
8, k+1 q, A
1, k q, A
3, k q, A
5, k q, A
7, k q, A
9, k q.
Generate one group of random noise spectrum, noise spectrum is carried out the energy adjustment, make the spectrum envelope of noise spectrum meet the envelope information of original signal, calculate the spectrum envelope information of the noise spectrum of adjusting through energy according to the frequency domain envelope information that decoding obtains.
The spectrum envelope information of the noise spectrum of the frequency domain spectra parameter substitution relevant position that decoding is obtained obtains the reconstructed frequency domain signal { y_swb of present frame
k(0), y_swb
k(1), L L, y_swb
k(319) }; Further, this can be organized signal and be retained in the decoded state register, use when carrying out LOF and maybe can not decode for needs.With reconstructed frequency domain signal { y_swb
k(0), y_swb
k(1), L L, y_swb
k(319) } and the previous frame signal be retained in one group of signal { imdct in the decoded state register when carrying out inverse transformation reconstruct
K-1(0), imdct
K-1(1), L L, imdct
K-1(319) }, carry out inverse transformation reconstruct, obtain reconstruct time-domain signal { y_hi
k' (0), y_hi
k' (1), L L, y_hi
k' (319) }, the transition coding mode that inverse transformation reconstruct mode and when coding adopt is corresponding, and what adopt when for example encoding is the MDCT conversion, and the inverse transformation reconstruct during decoding can be adopted anti-MDCT conversion.Further, one group of M signal { imdct that this frame signal can be produced when the anti-MDCT conversion
k(0), imdct
k(1), L L, imdct
k(319) } be retained in the decoded state register, use when carrying out inverse transformation reconstruct for the next frame signal.
Temporal envelope information { A from the k frame
0, k q, A
2, k q, A
4, k q, A
6, k q, A
8, k q, A
1, k-1 q, A
3, k-1 q, A
5, k-1 q, A
7, k-1 q, A
9, k-1 qAnd the temporal envelope information { A of k+1 frame
0, k+1 q, A
2, k+1 q, A
4, k+1 q, A
6, k+1 q, A
8, k+1 q, A
1, k q, A
3, k q, A
5, k q, A
7, k q, A
9, k qIn extract the actual time domain envelope information { A of present frame
0, k q, A
1, k q, L L A
9, k q.To { A
0, k q, A
1, k q, L L A
9, k qCarry out identical with coding side level and smooth interpolation operation, obtain the temporal envelope information after level and smooth
Temporal envelope after utilizing smoothly is to the time-domain signal { y_hi of reconstruct
k' (0), y_hi
k' (1), L L, y_hi
k' (319) } carry out anti-normalization operation, obtain the ultra-broadband signal { y_hi of reconstruct
k(0), y_hi
k(1), L L, y_hi
k(319) }.Further, after carrying out anti-normalization operation, can carry out some aftertreatments, as back Filtering Processing, back filtering enhancement process etc.
Fig. 5 has described the flow process of the coding/decoding method embodiment four that the embodiment of the invention provides, comprising:
Step 501, decoding obtain the temporal envelope information of the subsequent frame coding transmission of present frame;
Step 502, from the temporal envelope information that subsequent frame coding transmits, extract the subsequent frame part in the temporal envelope information of present frame;
When the quantity of temporal envelope information is fixed value, can be from the temporal envelope information that the subsequent frame coding transmits according to this fixed value, the subsequent frame part in the temporal envelope information of extraction present frame;
During by the input signal characteristics determined, can from the temporal envelope information that the subsequent frame coding transmits, extract the quantity of the temporal envelope information of present frame in the quantity of temporal envelope information; And then, use subsequent frame partly to obtain the temporal envelope information of present frame according to the quantity of the temporal envelope information of present frame;
Step 503, use subsequent frame partly obtain the temporal envelope information of present frame.
The correct transmission because present frame is failed, therefore the subsequent frame in the temporal envelope information of needs use present frame partly recovers the temporal envelope information of present frame, because subsequent frame partly is the part of temporal envelope information, therefore in order to make this partial information can reduce whole temporal envelope information as much as possible, can partly carry out interpolation to subsequent frame and handle the temporal envelope information that obtains present frame;
From the above, when present embodiment fails correctly to transmit at present frame, can use the temporal envelope information of subsequent frame partial reduction present frame of the temporal envelope information of present frame, thereby can use the remainder of the temporal envelope information that belongs to present frame to decode, because the decoding that the temporal envelope information that belongs to present frame that is to use is carried out, therefore the continuity of the holding signal that can try one's best makes auditory perception comfortable.
Fig. 6 has described the flow process of the coding/decoding method embodiment five that the embodiment of the invention provides, comprising:
Step 601, decoding obtain the temporal envelope information of the subsequent frame coding transmission of present frame;
Step 602, from the temporal envelope information that subsequent frame coding transmits, extract the subsequent frame part in the temporal envelope information of present frame;
Step 603, use subsequent frame partly obtain the temporal envelope information of present frame;
The M signal of step 604, the previous frame that obtains during with the reconstructed frequency domain signal of the previous frame of the present frame preserved and inverse transformation reconstruct carries out inverse transformation reconstruct, obtains the reconstruct time-domain signal of present frame;
Further, obtain the M signal that the reconstruct time-domain signal can further be preserved the present frame that produces when carrying out inverse transformation reconstruct well;
Wherein, preserve the M signal of the present frame that produces when carrying out inverse transformation reconstruct, can when next frame carries out inverse transformation reconstruct, use;
The mode of inverse transformation reconstruct is corresponding with the mode of the transition coding that when coding adopts, and what for example adopt when coding is 50% MDCT conversion, then when inverse transformation reconstruct, adopt corresponding anti-MDCT conversion and 50% overlap-add;
Temporal envelope information after step 606, the use is smoothly carried out anti-normalization operation to the reconstruct time-domain signal, obtains the present frame information of reconstruct.
From the above, when present embodiment fails correctly to transmit at present frame, can use the temporal envelope information of subsequent frame partial reduction present frame of the temporal envelope information of present frame, thereby can use the remainder of the temporal envelope information that belongs to present frame to decode, because the decoding that the temporal envelope information that belongs to present frame that is to use is carried out, therefore the continuity of the holding signal that can try one's best makes auditory perception comfortable.
The embodiment of the invention also provides the embodiment six of coding/decoding method, description also be situation during with temporal envelope information separated into two parts, flow process is as follows:
After the current k LOF, can't obtain the frequency domain envelope and the frequency domain spectra parameter of k frame, the temporal envelope information { A that can from code stream, decode and obtain the k+1 frame
0, k+1 q, A
2, k+1 q, A
4, k+1 q, A
6, k+1 q, A
8, k+1 q, A
1, k q, A
3, k q, A
5, k q, A
7, k q, A
9, k q, the subsequent frame part of the temporal envelope information that belongs to the k frame is wherein arranged.
To be retained in previous frame reconstructed frequency domain signal { y_swb in the codec register
K-1(0), y_swb
K-1(1), L L, y_swb
K-1And one group of signal { imdct that is retained in the codec register when anti-MDCT conversion, producing of previous frame signal (319) }
K-1(0), imdct
K-1(1), L L, imdct
K-1(319) } carry out inverse transformation reconstruct together, obtain reconstruct time-domain signal { y_hi
k' (0), y_hi
k' (1), L L, y_hi
k' (319) }, and the M signal { imdct that the k frame signal is produced when the inverse transformation reconstruct
k(0), imdct
k(1), L L, imdct
k(319) } be retained in the decoded state register, use when carrying out inverse transformation reconstruct for the next frame signal.
Temporal envelope information { A from the k+1 frame
0, k+1 q, A
2, k+1 q, A
4, k+1 q, A
6, k+1 q, A
8, k+1 q, A
1, k q, A
3, k q, A
5, k q, A
7, k q, A
9, k qIn extract the temporal envelope information { A of current k frame
1, k q, A
3, k q, A
5, k q, A
7, k q, A
9, k q, it is carried out interpolation, obtain the comparatively coarse temporal envelope information { A of a portrayal of current k frame
0, k q, A
1, k q, A
2, k q, A
3, k q, A
4, k q, A
5, k q, A
6, k q, A
7, k q, A
8, k q, A
9, k q, wherein
i＝2，4，6，8。
To { A
0, k q, A
1, k q, A
2, k q, A
3, k q, A
4, k q, A
5, k q, A
6, k q, A
7, k q, A
8, k q, A
9, k qCarry out the inverse quantization identical, smooth operation with coding side, obtain the temporal envelope information after level and smooth
Temporal envelope after utilizing smoothly is to the time-domain signal { y_hi of reconstruct
k' (0), y_hi
k' (1), L L, y_hi
k' (319) } carry out anti-normalization operation, obtain the ultra-broadband signal { y_hi of reconstruct
k(0), y_hi
k(1), L L, y_hi
k(319) }.Further, after carrying out anti-normalization operation, can carry out some aftertreatments, as back Filtering Processing, back filtering enhancement process etc.
Wherein, need to prove that the above-mentioned flow process of the embodiment of the invention only is the implementation that the embodiment of the invention provides, is not to limit each step only can carry out according to above-mentioned flow process.
The embodiment of the invention also provides code device, and Fig. 7 has described the structure of the code device embodiment that the embodiment of the invention provides, comprising:
Temporal envelope information acquisition unit 701 is used to obtain the quantified time domain enveloping information of present frame;
Coding transmitting element 704 is used to encode and sends assembled unit 703 and makes up the temporal envelope information that the present frame that obtains need be encoded and be transmitted.
From the above, present embodiment will be worked as each frame and all will be divided into two parts at least, comprise present frame part and subsequent frame part, and present frame part and subsequent frame partly are placed in the coding of at least two frames and divide transmission, have only present frame partly to belong to present frame in the temporal envelope information that coding sends like this, and remainder is placed in the temporal envelope information of subsequent frame and sends, if thereby decoding end can not be correctly decoded present frame, also can from subsequent frame, extract the temporal envelope information that belongs to present frame, and then can use the temporal envelope information that belongs to present frame to decode, because the decoding that the temporal envelope information that belongs to present frame that is to use is carried out, therefore the continuity of the holding signal that can try one's best makes auditory perception comfortable.
Further, the code device that provides of the embodiment of the invention can also comprise:
The frequency-region signal acquiring unit is used for after temporal envelope information acquisition unit 701 is obtained quantified time domain enveloping information, uses the frequency-region signal of quantified time domain enveloping information acquisition present frame;
Frequency domain envelope information extraction unit is used to extract the frequency domain envelope information of frequency-region signal, chooses the frequency domain spectra parameter;
Quantifying unit is used to quantize frequency domain envelope information extraction unit frequency domain envelope information of extracting and the frequency domain spectra parameter of choosing, the frequency domain spectra parameter after frequency domain envelope information after obtaining quantizing and the quantification;
At this moment, coding transmitting element 704, also be used to encode and send the frequency domain envelope information after the quantification and quantize after the frequency domain spectra parameter.
Wherein, above-mentioned frequency-region signal acquiring unit specifically can comprise: smooth unit, be used for after quantifying unit obtains quantified time domain enveloping information, and quantified time domain enveloping information is carried out level and smooth interpolation handle, obtain the temporal envelope after level and smooth; The normalization unit is used to use the temporal envelope after level and smooth that time-domain signal is carried out normalized, obtains the time-domain signal after the normalization; The transition coding unit is used to use the time-domain signal after the normalization of time-domain signal after the normalization and previous frame data to constitute array; Array is carried out the frequency-region signal that transition coding obtains present frame.
The embodiment of the invention also provides decoding device, and the structure that Fig. 8 has described the decoding device embodiment one that the embodiment of the invention provides comprises:
Decoding unit 801, being used to decode obtains the temporal envelope information that present frame coding transmits;
Acquiring unit 803 is used for obtaining the subsequent frame part of the temporal envelope information of present frame;
From the above, have only present frame partly to belong to present frame in the temporal envelope information of the present frame that present embodiment receives, remainder is transmitted by other frames, even therefore present frame can not be correctly decoded or lose, also can from other frames, obtain the remainder of the temporal envelope information of present frame, and then can use the remainder of the temporal envelope information that belongs to present frame to decode, because the decoding that the temporal envelope information that belongs to present frame that is to use is carried out, therefore the continuity of the holding signal that can try one's best makes auditory perception comfortable; Simultaneously, when the present frame normal transmission, also can the information of present frame be correctly decoded, can not influence the continuity of signal.
Further, above-mentioned decoding unit 801, also being used to decodes obtains the frequency domain envelope information and the frequency domain spectra parameter information of present frame; At this moment, the above-mentioned decoding device that provides of the embodiment of the invention can also comprise:
Reconfiguration unit is used to use frequency domain envelope information and frequency domain spectra parameter information, obtains the reconstructed frequency domain signal of present frame;
The inverse transformation reconfiguration unit is used for the M signal to the previous frame of the present frame of reconstruct frequency-region signal and preservation, carries out inverse transformation reconstruct and obtains the reconstruct time-domain signal;
Level and smooth interpolation unit is used for the temporal envelope information of present frame is carried out level and smooth interpolation operation, obtains the temporal envelope information after level and smooth;
Anti-normalization unit, be used to use the reconstruct time-domain signal and level and smooth after temporal envelope information, carry out the present frame information that anti-normalization operation obtains reconstruct.
Further, the decoding device embodiment one that the embodiment of the invention provides can also comprise the preservation unit, is used for preserving the reconstructed frequency domain signal of present frame after reconfiguration unit obtains the reconstructed frequency domain signal of present frame; Also be used for after the inverse transformation reconfiguration unit obtains the reconstruct time-domain signal M signal of the present frame that obtains when preserving inverse transformation reconstruct;
Fig. 9 has described the structure of the decoding device embodiment two that the embodiment of the invention provides, comprising:
Decoding unit 901, the temporal envelope information that the subsequent frame coding that obtains present frame of being used to decode transmits;
Temporal envelope information acquisition unit 903 is used to use subsequent frame partly to obtain the temporal envelope information of present frame.
From the above, when present embodiment fails correctly to transmit at present frame, can use the temporal envelope information of subsequent frame partial reduction present frame of the temporal envelope information of present frame, thereby can use the remainder of the temporal envelope information that belongs to present frame to decode, because the decoding that the temporal envelope information that belongs to present frame that is to use is carried out, therefore the continuity of the holding signal that can try one's best makes auditory perception comfortable.
Further, the decoding device embodiment two that provides of the embodiment of the invention can also comprise:
Reconstruct time-domain signal acquiring unit, the M signal of the previous frame that obtains when being used for the reconstructed frequency domain signal of previous frame of the present frame that will preserve and inverse transformation reconstruct carries out inverse transformation reconstruct, obtains the reconstruct time-domain signal of present frame;
Smooth unit is used for the temporal envelope information of present frame is carried out the operation of level and smooth interpolation, obtains the temporal envelope information behind present frame level and smooth;
Anti-normalization unit is used to use the temporal envelope information after level and smooth that the reconstruct time-domain signal is carried out anti-normalization operation, obtains the present frame information of reconstruct.
Further, the decoding device embodiment two that the embodiment of the invention provides can also comprise the preservation unit, is used for preserving the M signal of the present frame that produces when carrying out inverse transformation reconstruct after reconstruct time-domain signal acquiring unit obtains the reconstruct time-domain signal of present frame.
One of ordinary skill in the art will appreciate that all or part of step that realizes in the foregoing description method is to instruct relevant hardware to finish by program, described program can be stored in a kind of computer-readable recording medium, this program comprises the steps: when carrying out
Obtain the quantified time domain enveloping information of present frame;
Described quantified time domain enveloping information is divided into two parts at least, comprises present frame part and subsequent frame part;
Make up the described present frame part and the subsequent frame part of frame data before, obtain the temporal envelope information that present frame need be encoded and be transmitted;
Coding also sends the temporal envelope information that described present frame need be encoded and be transmitted.
And/or comprise the steps:
Decoding obtains the temporal envelope information that the present frame coding transmits;
From the temporal envelope information that described present frame coding transmits, extract the present frame part in the temporal envelope information of described present frame;
Obtain the subsequent frame part in the temporal envelope information of described present frame;
Make up the temporal envelope information that described present frame part and described subsequent frame partly obtain described present frame.
And/or comprise the steps:
Decoding obtains the temporal envelope information of the subsequent frame coding transmission of described present frame;
From the temporal envelope information that described subsequent frame coding transmits, extract the subsequent frame part in the temporal envelope information of described present frame;
Use described subsequent frame partly to obtain the temporal envelope information of described present frame.
The above-mentioned storage medium of mentioning can be a ROM (read-only memory), disk or CD etc.
More than coding method and device, coding/decoding method and device that the embodiment of the invention provided are described in detail, the explanation of above embodiment just is used for help understanding method of the present invention and thought thereof; Simultaneously, for one of ordinary skill in the art, according to thought of the present invention, the part that all can change in specific embodiments and applications, in sum, this description should not be construed as limitation of the present invention.
Claims (26)
1. a voice coding method is characterized in that, comprising:
Obtain the time-domain signal of present frame, extract the temporal envelope information of described time-domain signal, the temporal envelope information of extracting is quantized to obtain quantified time domain enveloping information;
Described quantified time domain enveloping information is divided into two parts at least, comprises present frame part and subsequent frame part;
Make up the described present frame part and the subsequent frame part of frame before, obtain the temporal envelope information that present frame need be encoded and be transmitted;
Coding also sends the temporal envelope information that described present frame need be encoded and be transmitted.
2. coding method as claimed in claim 1 is characterized in that, obtains further comprising after the quantified time domain enveloping information:
Use described quantified time domain enveloping information to obtain the frequency-region signal of described present frame;
Extract the frequency domain envelope information of described frequency-region signal, choose the frequency domain spectra parameter;
Quantize described frequency domain envelope information and described frequency domain spectra parameter, the frequency domain spectra parameter after frequency domain envelope information after obtaining quantizing and the quantification;
Described coding also sends the described present frame temporal envelope information that transmits of need encoding and comprises:
Coding also sends temporal envelope information, the frequency domain envelope information after the quantification and the frequency domain spectra parameter after the quantification that described present frame need be encoded and be transmitted.
3. coding method as claimed in claim 2 is characterized in that, the frequency-region signal that the described quantified time domain enveloping information of described use obtains described present frame is specially:
Described quantified time domain enveloping information is carried out level and smooth interpolation handle, obtain the temporal envelope after level and smooth;
Use described temporal envelope after level and smooth that described time-domain signal is carried out normalized, obtain the time-domain signal after the normalization;
Use the time-domain signal after the normalization of time-domain signal after the described normalization and previous frame data to constitute array;
Described array is carried out the frequency-region signal that transition coding obtains described present frame.
4. coding method as claimed in claim 3 is characterized in that, described array is carried out the frequency-region signal that transition coding obtains described present frame be specially:
Described array is carried out the discrete cosine transform of 50% overlapping correction.
5. as the arbitrary described coding method of claim 1 to 4, it is characterized in that being divided into described quantified time domain enveloping information at least, two parts are specially:
With described quantified time domain enveloping information separated into two parts; Described present frame partly is the odd number part of described quantified time domain enveloping information;
Or
Described present frame partly is the even number part of described quantified time domain enveloping information.
6. coding method as claimed in claim 5 is characterized in that, the quantity of the temporal envelope information of the described time-domain signal of extraction is determined by the input signal feature of described time-domain signal.
7. coding method as claimed in claim 6 is characterized in that, makes up the described present frame part and the subsequent frame part of frame data before, obtains need the encode step of the temporal envelope information that transmits of present frame and comprises:
Make up described present frame part, before the temporal envelope information of subsequent frame part, the described present frame that extracts of frame data quantity and described before the quantity of temporal envelope information of frame, obtain the temporal envelope information that described present frame need be encoded and be transmitted.
8. coding method as claimed in claim 5, it is characterized in that, make up the described present frame part and the subsequent frame part of frame data before, when obtaining present frame and need encode the temporal envelope information that transmits, if described present frame is first frame, the subsequent frame of described frame data before is the subsequent frame part of frame data before for setting in advance partly.
9. coding method as claimed in claim 5 is characterized in that, the quantity of the temporal envelope information of the described time-domain signal of extraction is preset quantity.
10. a tone decoding method is characterized in that, comprising:
Decoding obtains the temporal envelope information that the present frame coding transmits;
From the temporal envelope information that described present frame coding transmits, extract the present frame part in the temporal envelope information of described present frame;
Decoding obtains the temporal envelope information of the subsequent frame coding transmission of described present frame;
From the temporal envelope information that described subsequent frame coding transmits, extract the subsequent frame part in the temporal envelope information of described present frame;
Make up the temporal envelope information that described present frame part and described subsequent frame partly obtain described present frame.
11. coding/decoding method as claimed in claim 10 is characterized in that, from the temporal envelope information that described present frame coding transmits, the step of extracting the present frame part in the temporal envelope information of described present frame comprises:
From the temporal envelope information that described present frame coding transmits, extract the quantity of the temporal envelope information of described present frame;
According to the quantity of the temporal envelope information of described present frame, from the temporal envelope information that described present frame coding transmits, extract the present frame part in the temporal envelope information of described present frame.
12. coding/decoding method as claimed in claim 10 is characterized in that, described present frame is a last frame, and the subsequent frame in the described temporal envelope information of obtaining described present frame partly is specially:
Use preset mode to obtain subsequent frame part in the temporal envelope information of described present frame.
13. as claim 10 or 12 described coding/decoding methods, it is characterized in that, also comprise:
Decoding obtains the frequency domain envelope information and the frequency domain spectra parameter information of described present frame;
Use described frequency domain envelope information and described frequency domain spectra parameter information, obtain the reconstructed frequency domain signal of described present frame;
The M signal that obtains when the previous frame of the described present frame of described reconstructed frequency domain signal and preservation is carried out inverse transformation reconstruct carries out inverse transformation reconstruct and obtains the reconstruct time-domain signal;
The temporal envelope information of described present frame is carried out the operation of level and smooth interpolation, obtain the temporal envelope information after level and smooth;
Use described reconstruct time-domain signal and described temporal envelope information after level and smooth, carry out the present frame information that anti-normalization operation obtains reconstruct.
14. coding/decoding method as claimed in claim 13 is characterized in that, the reconstructed frequency domain signal that described frequency domain envelope information of described use and described frequency domain spectra parameter information obtain described present frame is specially:
Generation random noise spectrum uses described frequency domain envelope information that described random noise spectrum is carried out the energy adjustment, calculates the spectrum envelope information through energy adjustment random noise spectrum;
Described frequency domain spectra parameter information is substituted described spectrum envelope information, obtain the reconstructed frequency domain signal of described present frame.
15. coding/decoding method as claimed in claim 13 is characterized in that, described inverse transformation reconstruct is specially:
The discrete cosine transform of uncorrecting and 50% overlap-add.
16. a tone decoding method is characterized in that, comprising:
Decoding obtains the temporal envelope information of the subsequent frame coding transmission of present frame;
From the temporal envelope information that described subsequent frame coding transmits, extract the subsequent frame part in the temporal envelope information of described present frame;
Use subsequent frame in the temporal envelope information of described present frame partly to obtain the temporal envelope information of described present frame.
17. coding/decoding method as claimed in claim 16 is characterized in that, the step of using described subsequent frame partly to obtain the temporal envelope information of described present frame comprises:
From the temporal envelope information that described subsequent frame coding transmits, extract the quantity of the temporal envelope information of described present frame;
According to the quantity of the temporal envelope information of described present frame, use described subsequent frame partly to obtain the temporal envelope information of described present frame.
18. as claim 16 or 17 described coding/decoding methods, it is characterized in that, also comprise:
The M signal of the reconstructed frequency domain signal of the previous frame of the described present frame preserved, the described previous frame that obtains when carrying out inverse transformation reconstruct with described previous frame is carried out inverse transformation reconstruct, obtain the reconstruct time-domain signal of described present frame;
The temporal envelope information of described present frame is carried out the operation of level and smooth interpolation, obtain the temporal envelope information behind described present frame level and smooth;
Use described temporal envelope information after level and smooth that described reconstruct time-domain signal is carried out anti-normalization operation, obtain the present frame information of reconstruct.
19. coding/decoding method as claimed in claim 18 is characterized in that, described inverse transformation reconstruct is specially: the discrete cosine transform of uncorrecting and 50% overlap-add.
20. a sound encoding device is characterized in that, comprising:
The temporal envelope information acquisition unit is used to obtain the time-domain signal of present frame, extracts the temporal envelope information of described time-domain signal, and the temporal envelope information of extracting is quantized to obtain quantified time domain enveloping information;
Split cells is used for the quantified time domain enveloping information that described temporal envelope information acquisition unit is obtained is divided into two parts at least, comprises present frame part and subsequent frame part;
Assembled unit is used to make up the described present frame part and the subsequent frame part of frame data before, obtains the temporal envelope information that present frame need be encoded and be transmitted;
The coding transmitting element is used to encode and sends described assembled unit and makes up the temporal envelope information that the described present frame that obtains need be encoded and be transmitted.
21. code device as claimed in claim 20 is characterized in that, also comprises:
The frequency-region signal acquiring unit is used for using described quantified time domain enveloping information to obtain the frequency-region signal of described present frame after described temporal envelope information acquisition unit is obtained quantified time domain enveloping information;
Frequency domain envelope information extraction unit is used to extract the frequency domain envelope information of described frequency-region signal, chooses the frequency domain spectra parameter;
Quantifying unit is used to quantize described frequency domain envelope information extraction unit described frequency domain envelope information of extracting and the described frequency domain spectra parameter of choosing, the frequency domain spectra parameter after frequency domain envelope information after obtaining quantizing and the quantification;
Described coding transmitting element, also be used to encode and send the frequency domain envelope information after the described quantification and quantize after the frequency domain spectra parameter.
22. code device as claimed in claim 21 is characterized in that, described frequency-region signal acquiring unit comprises:
Smooth unit is used for after described temporal envelope information acquisition unit obtains quantified time domain enveloping information, described quantified time domain enveloping information is carried out level and smooth interpolation handle, and obtains the temporal envelope after level and smooth;
The normalization unit is used to use described temporal envelope after level and smooth that described time-domain signal is carried out normalized, obtains the time-domain signal after the normalization;
The transition coding unit is used to use the time-domain signal after the normalization of time-domain signal after the described normalization and previous frame data to constitute array; Described array is carried out the frequency-region signal that transition coding obtains described present frame.
23. an audio decoding apparatus is characterized in that, comprising:
Decoding unit, being used to decode obtains the temporal envelope information that present frame coding transmits;
Extraction unit is used for the temporal envelope information that transmits from described present frame coding, extracts the present frame part in the temporal envelope information of described present frame;
Acquiring unit, the temporal envelope information that the subsequent frame coding that obtains described present frame of being used to decode transmits from the temporal envelope information that described subsequent frame coding transmits, is extracted the subsequent frame part in the temporal envelope information of described present frame;
Assembled unit is used to make up the temporal envelope information that described present frame part and described subsequent frame partly obtain described present frame.
24. decoding device as claimed in claim 23 is characterized in that, described decoding unit, and also being used to decodes obtains the frequency domain envelope information and the frequency domain spectra parameter information of described present frame;
Described decoding device also comprises:
Reconfiguration unit is used to use described frequency domain envelope information and described frequency domain spectra parameter information, obtains the reconstructed frequency domain signal of described present frame;
The inverse transformation reconfiguration unit is used for the previous frame of described reconstructed frequency domain signal and the described present frame preserved is carried out the M signal that inverse transformation reconstruct obtains, and carries out inverse transformation reconstruct and obtains the reconstruct time-domain signal;
Level and smooth interpolation unit is used for the temporal envelope information of described present frame is carried out level and smooth interpolation operation, obtains the temporal envelope information after level and smooth;
Anti-normalization unit is used to use described reconstruct time-domain signal and described temporal envelope information after level and smooth, carries out the present frame information that anti-normalization operation obtains reconstruct.
25. an audio decoding apparatus is characterized in that, comprising:
Decoding unit, the temporal envelope information that the subsequent frame coding that obtains present frame of being used to decode transmits;
Extraction unit is used for the temporal envelope information that transmits from described subsequent frame coding, extracts the subsequent frame part in the temporal envelope information of described present frame;
The temporal envelope information acquisition unit is used for using the subsequent frame of the temporal envelope information of described present frame partly to obtain the temporal envelope information of described present frame.
26. decoding device as claimed in claim 25 is characterized in that, also comprises:
Reconstruct time-domain signal acquiring unit, the M signal that is used for obtaining when the reconstructed frequency domain signal of previous frame of the described present frame that will preserve and described previous frame carry out inverse transformation reconstruct carries out inverse transformation reconstruct, obtains the reconstruct time-domain signal of described present frame;
Smooth unit is used for the temporal envelope information of described present frame is carried out the operation of level and smooth interpolation, obtains the temporal envelope information behind described present frame level and smooth;
Anti-normalization unit is used to use described temporal envelope information after level and smooth that described reconstruct time-domain signal is carried out anti-normalization operation, obtains the present frame information of reconstruct.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2008100891481A CN101552008B (en) | 2008-04-01 | 2008-04-01 | Voice coding method, coding device, decoding method and decoding device |
| PCT/CN2009/071022 WO2009121280A1 (en) | 2008-04-01 | 2009-03-26 | An encoding method and apparatus, a decoding method and apparatus |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2008100891481A CN101552008B (en) | 2008-04-01 | 2008-04-01 | Voice coding method, coding device, decoding method and decoding device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101552008A CN101552008A (en) | 2009-10-07 |
| CN101552008B true CN101552008B (en) | 2011-11-16 |
Family
ID=41134837
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2008100891481A Expired - Fee Related CN101552008B (en) | 2008-04-01 | 2008-04-01 | Voice coding method, coding device, decoding method and decoding device |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN101552008B (en) |
| WO (1) | WO2009121280A1 (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103280222B (en) * | 2013-06-03 | 2014-08-06 | 腾讯科技（深圳）有限公司 | Audio encoding and decoding method and system thereof |
| CN113782039A (en) * | 2017-08-10 | 2021-12-10 | 华为技术有限公司 | Time domain stereo coding and decoding method and related products |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1441949A (en) * | 2000-05-11 | 2003-09-10 | 艾利森电话股份有限公司 | Forward error correction in speech coding |
| CN1732512A (en) * | 2002-12-31 | 2006-02-08 | 诺基亚有限公司 | Method and device for compressed-domain packet loss concealment |
| CN101071568A (en) * | 2005-11-23 | 2007-11-14 | 美国博通公司 | Method and system of audio decoder |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP4464488B2 (en) * | 1999-06-30 | 2010-05-19 | パナソニック株式会社 | Speech decoding apparatus, code error compensation method, speech decoding method |
| EP1414176B1 (en) * | 2002-10-24 | 2010-03-31 | Panasonic Corporation | Communication device and communication method immune to burst error, program for executing the method, and computer-readable storage medium storing the program |
| US20050049853A1 (en) * | 2003-09-01 | 2005-03-03 | Mi-Suk Lee | Frame loss concealment method and device for VoIP system |
| US8280728B2 (en) * | 2006-08-11 | 2012-10-02 | Broadcom Corporation | Packet loss concealment for a sub-band predictive coder based on extrapolation of excitation waveform |
-
2008
- 2008-04-01 CN CN2008100891481A patent/CN101552008B/en not_active Expired - Fee Related
-
2009
- 2009-03-26 WO PCT/CN2009/071022 patent/WO2009121280A1/en active Application Filing
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1441949A (en) * | 2000-05-11 | 2003-09-10 | 艾利森电话股份有限公司 | Forward error correction in speech coding |
| CN1732512A (en) * | 2002-12-31 | 2006-02-08 | 诺基亚有限公司 | Method and device for compressed-domain packet loss concealment |
| CN101071568A (en) * | 2005-11-23 | 2007-11-14 | 美国博通公司 | Method and system of audio decoder |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101552008A (en) | 2009-10-07 |
| WO2009121280A1 (en) | 2009-10-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101809657B (en) | Method and device for noise filling | |
| CN101939782B (en) | Adaptive transition frequency between noise fill and bandwidth extension | |
| CN101521014B (en) | Audio bandwidth expansion coding and decoding devices | |
| JP6321072B2 (en) | Audio / voice encoding method and audio / voice decoding method | |
| CN101512639B (en) | Method and equipment for voice/audio transmitter and receiver | |
| KR101221918B1 (en) | A method and an apparatus for processing a signal | |
| KR101693280B1 (en) | Method, apparatus, and system for processing audio data | |
| JP5096582B2 (en) | Noise generating apparatus and method | |
| CN103310793B (en) | A kind of water sound real-time digital speech communication method | |
| CN106847297B (en) | Prediction method of high-frequency band signal, encoding/decoding device | |
| CN1312658C (en) | Perceptually improved encoding of acoustic signals | |
| CN113140225A (en) | Voice signal processing method and device, electronic equipment and storage medium | |
| CN103280222A (en) | Audio encoding and decoding method and system thereof | |
| CN101325059B (en) | Method and apparatus for transmitting and receiving encoding-decoding speech | |
| CN101552008B (en) | Voice coding method, coding device, decoding method and decoding device | |
| EP3642839B1 (en) | Audio signal encoding and decoding | |
| Madane et al. | Speech compression using Linear predictive coding | |
| CN101740030B (en) | Method and device for transmitting and receiving speech signals | |
| US20120123788A1 (en) | Coding method, decoding method, and device and program using the methods | |
| CN101170590B (en) | A method, system and device for transmitting encoding stream under background noise | |
| Balam et al. | Multiple descriptions and path diversity for voice communications over wireless mesh networks | |
| CN114863942A (en) | Model training method for voice quality conversion, method and device for improving voice quality | |
| WO2009146734A1 (en) | Multi-channel audio coding | |
| CN105632504A (en) | ADPCM codec and method of packet loss concealment in ADPCM codec | |
| CN117476024A (en) | Audio encoding method, audio decoding method, apparatus, and readable storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20111116 Termination date: 20190401 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |