CN114724567A - Dynamic switching method for lossy or lossless compression of fixed bandwidth audio-visual data - Google Patents

Abstract

The application discloses a dynamic switching method for lossy or lossless compression of fixed bandwidth audio-visual data, which comprises the following steps: calculating a predetermined maximum energy value that each frame of the plurality of frames can be compressed losslessly according to the set bandwidth; reading original voice and information data of each frame; calculating the energy value of the original voice and information data of each frame; and judging whether the energy value of the original audio-visual data of each frame is larger than a preset maximum energy value, if so, performing lossy compression on the original audio-visual data of the frame, and if not, performing lossless compression on the original audio-visual data of the frame.

Description

Dynamic switching method of lossy or lossless compression of fixed bandwidth audio data

technical field

The present application relates to a method for compressing audio information data streams, in particular to a dynamic switching method for lossy or lossless compression of fixed bandwidth audio information data.

Background technique

The digital audio data is usually encoded in a frame-by-frame (frame) manner and then transmitted to the receiving end via a wired or wireless network for decoding and playback. There are two types of coding methods: lossy compression and lossless compression. Lossless compression can obtain audio data that is exactly the same as before compression after decompression, but it is relatively lossy compression. The compression efficiency of lossless compression is usually poor, and lossy compression removes the subtle differences in sound that the human ear usually cannot discern. The average user cares about the convenience and stability of use, while the music lovers who pursue high sound quality hope that the degree of audio distortion is as small as possible, and it is best to reproduce the original sound.

In order to solve the aforementioned problems, MPEG-4SLS proposes a method of mixing lossy and lossless compression to meet the different needs of the two users. It adds the difference from the original audio data to the base of the lossy compression. According to the requirement, only part or all of the data of lossy compression is transmitted, such as Advanced Audio Coding (AAC) lossy compression base additional difference part to complete lossless compression. However, compared with AAC, the bandwidth of MPEG-4SLS is greatly increased by 128Kbps, and it is no longer a fixed bandwidth. Instead, it increases the complexity and instability of network transmission and processing.

SUMMARY OF THE INVENTION

Based on this, the technical problem to be solved by the present application is to provide a dynamic switching method for lossy or lossless compression of audio information data with a fixed bandwidth in view of the deficiencies of the prior art, which includes the following steps: calculating a plurality of frames according to the set bandwidth The predetermined maximum energy value that each frame in can be lossless compression; read the original audio data of each frame; calculate the energy value of the original audio data of each frame; and judge whether the energy value of the original audio data of each frame is greater than the predetermined maximum energy value If yes, perform lossy compression on the original audio data of the frame to generate audio lossy compressed data, if not, perform lossless compression on the original audio data of the frame to generate audio lossless compressed data.

According to an example embodiment, the dynamic switching method for lossy or lossless compression of fixed-bandwidth audio signal data further comprises the steps of: calculating the square value of the amplitude of each waveform in the plurality of waveforms of the audio signal signal of the original audio signal data of each frame; summing the respective squared values of the plurality of waveforms to generate a total energy value; and dividing the total energy value by the number of the plurality of waveforms to obtain an average energy value, and taking the logarithm of the average energy value to calculate The energy of the original audio data of each frame is expressed by the following equation: E=log((x ₁ ² +x ₂ ² +...+x _n ² )/n), where E represents the energy of the original audio data of each frame, x ₁ to x _n represent the respective amplitudes of the plurality of waveforms, and n represents the number of the plurality of waveforms.

According to an exemplary embodiment, the dynamic switching method for lossy or lossless compression of fixed-bandwidth audio data further includes the following steps: judging whether the energy value of the original audio data of each frame is greater than a predetermined maximum energy value, and if so, converting the original audio data of the frame Perform lossy compression on the data to generate audio lossy compressed data, if not, perform the next step; determine whether the previous frame of the current frame is lossless compression, if so, perform lossless compression on the original audio data of the current frame to generate audio lossless compression Data, do not perform the subsequent steps, if not, perform the subsequent steps in sequence; apply the window function of the Modified Discrete Cosine Transform (MDCT) to the original audio data of the previous frame, according to the time domain mixing. Time-Domain Aliasing Cancellation (TDAC) is used to calculate the function audio data; and the function audio data and the original audio data of the current frame are subjected to lossless compression to generate audio lossless compressed data.

According to an exemplary embodiment, the dynamic switching method for lossy or lossless compression of fixed-bandwidth audio information data further includes the steps of: calculating a predetermined maximum length allowed after compression of the original audio information data of each frame according to the set bandwidth; Whether the data length of the audio information lossless compressed data of the frame is greater than the predetermined maximum length, if not, keep the audio information lossless compressed data, if so, perform lossy compression on the original audio information data of the frame to generate audio information lossy compressed data, replace the data with a data length greater than the predetermined length. Lossless compressed data of the maximum length of the message.

According to an exemplary embodiment, the dynamic switching method for lossy or lossless compression of fixed-bandwidth audio information data further includes the following steps: performing lossy compression on the original audio information data of the previous frame; deciding to perform lossy compression on the original audio information data of the current frame Compression; the original audio data of the previous frame and the original audio data of the current frame are subjected to improved discrete cosine transform after applying the window function to calculate the functional audio data; and lossy compression of the functional audio data to generate audio lossy Compressed data.

According to an exemplary embodiment, the dynamic switching method for lossy or lossless compression of fixed-bandwidth audio information data further includes the steps of: performing lossy compression on the original audio information data of the previous frame; deciding to perform lossless compression on the original audio information data of the current frame ; Apply the window function of the improved discrete cosine transform to the original audio data of the previous frame, and calculate the functional audio data according to the time-domain aliasing elimination method; and perform the function audio data with the complete original audio data of the current frame. Lossless compression to produce audio lossless compressed data.

According to an example embodiment, the method for dynamically switching between lossy or lossless compression of fixed bandwidth audio data further comprises the steps of: performing lossless compression on the original audio data of the previous frame to generate lossless compressed audio data of the previous frame; and It is decided to perform lossless compression on the original audio data of the current frame to generate lossless compressed audio data of the current frame.

According to an exemplary embodiment, the method for dynamically switching between lossy or lossless compression of fixed-bandwidth audio information data further includes the steps of: compressing the original audio information data of the previous frame losslessly to generate audio information lossless compressed data; The original audio data of the current frame and the original audio data of the current frame are applied with a window function, and then an improved discrete cosine transform is performed to calculate a function audio data; and the function audio data is subjected to lossy compression.

Description of drawings

FIG. 1 is a flow chart of compression steps of a dynamic switching method for lossy or lossless compression of fixed-bandwidth audio information data according to an embodiment of the present invention.

FIG. 2 is a flow chart of decompression steps of a dynamic switching method for lossy or lossless compression of fixed-bandwidth audio information data according to an embodiment of the present invention.

3 is a schematic diagram of overlapping original audio data of multiple frames of a dynamic switching method for lossy or lossless compression of fixed bandwidth audio data according to an embodiment of the present invention.

FIG. 4 is a schematic diagram of decompression of a previous frame and a current frame of lossy compression according to a dynamic switching method for lossy or lossless compression of fixed-bandwidth audio information data according to an embodiment of the present invention.

5 is a schematic diagram of decompression of a lossy compressed previous frame and a lossless compressed current frame according to a dynamic switching method for lossy or lossless compression of fixed-bandwidth audio information data according to an embodiment of the present invention.

FIG. 6 is a schematic diagram of a method for dynamically switching between lossy or lossless compression of fixed-bandwidth audio information data according to an embodiment of the present invention, and a schematic diagram of decompression of a lossless compressed previous frame and a current frame.

FIG. 7 is a schematic diagram of a method for dynamically switching between lossy or lossless compression of fixed-bandwidth audio information data according to an embodiment of the present invention, and a schematic diagram of decompression of the previous frame of lossless compression and the current frame of lossy compression.

Detailed ways

Please refer to FIG. 1 , which is a flow chart of compression steps of a dynamic switching method for lossy or lossless compression of fixed-bandwidth audio information data according to an embodiment of the present invention. As shown in FIG. 1 , the dynamic switching method for lossy or lossless compression of fixed-bandwidth audio information data according to an embodiment of the present invention may include steps S101 to S115 , and the execution order, content, and times of these steps can be adjusted according to actual needs. These steps S101 to S115 are performed for each frame of the plurality of frames, so as to individually determine whether each frame adopts lossy compression or lossless compression under a preset bandwidth/fixed bandwidth. Different frames may adopt different compression methods, for example, but not limited to, the previous frame adopts lossy compression, and the current frame is dynamically switched to adopt lossless compression. For the convenience of description, hereinafter, the currently processed frame is simply referred to as the current frame.

In step S101, the energy value of the original audio data of the current frame is calculated. For example, the square value of the amplitude of each of the multiple waveforms of the audio signal of the original audio data of each frame is calculated. Next, the respective squared values of the plurality of waveforms are summed to generate an energy total value. Divide the total energy by the number of waveforms to obtain an average energy value. Next, the logarithm of the average energy value is taken to calculate the energy of the original audio data of each frame. It is expressed by the following equation: E=log((x ₁ ² +x ₂ ² +...+x _n ² )/n), where E represents the energy of the original audio data of each frame, and x ₁ to x _n respectively represent a plurality of waveforms The respective multiple amplitudes, n represents the number of multiple waveforms/number of multiple acquired multiple amplitudes.

In step S103, according to a preset bandwidth, a predetermined maximum energy value that can be losslessly compressed is calculated, and it is determined whether the energy value of the original audio data of the current frame is greater than the predetermined maximum energy value. If the energy value of the original audio data of the current frame is greater than the predetermined maximum energy value, that is, E>Emax (where E represents the energy value of the original audio data of the current frame, and Emax represents the predetermined maximum energy value), the lossy compression method is decided to be selected. , and then step S105 is executed. Conversely, if the energy value of the original audio data of the current frame is not greater than (ie equal to or smaller than) the predetermined maximum energy value, it is decided to select the lossless compression mode, and then step S107 is executed.

In step S105, lossy compression is performed on the original audio data of the current frame with an energy value greater than a predetermined maximum energy value, so as to generate lossy compressed audio data. In this way, when the energy value of the original audio data is too large, the lossless compression is still attempted, resulting in a waste of computing time. If necessary, then step S115 can be executed.

In step S107, it is determined whether the original audio information data of the previous frame of the current frame is compressed in a lossless compression manner.

If the original audio data of the previous frame adopts lossy compression, in order to restore the complete original audio data of the previous frame smoothly during decompression, it is necessary to obtain the original audio data of the previous frame and the current frame of the previous frame. Therefore, if the original audio data of the previous frame adopts lossy compression, when the lossless compression of the original audio data of the current frame is performed, step S109 needs to be performed, and then step S111 is performed.

In step S109, the original audio information data of the previous frame (for example, the previous frame B) is applied to the window function (window function) of the Modified Discrete Cosine Transform (MDCT), according to the time domain aliasing elimination method ( Time-Domain Aliasing Cancellation, TDAC) to calculate a function of audio data.

If the original audio data of the previous frame is compressed using a lossy compression method, the lossless compression range of the current frame needs to be extended forward to the overlap portion of the current frame and the previous frame. Therefore, after step S109 is performed, step S111 is then performed to perform lossless compression on the audio data of the function and the complete original audio data of the current frame to generate a lossless compressed audio data.

Conversely, if it is determined that the original audio data of the previous frame and the current frame are all compressed by lossless, then when the original audio data of the current frame is compressed losslessly, it is not necessary to extend the compression range of the current frame to the previous frame ( That is, it is not necessary to perform step S109), but step S111 is directly performed to directly perform lossless compression on the original audio data of the current frame.

In step S113, according to the set bandwidth, a predetermined maximum length allowed after lossless compression of each frame is calculated, and it is judged whether the data length of the audio signal lossless compressed data of the current frame is greater than the predetermined maximum length.

If the data length of the audio signal lossless compressed data of the current frame is not greater than the predetermined maximum length, that is, when L≤Lmax (where L represents the data length of the audio signal lossless compressed data of the current frame, and Lmax represents the predetermined maximum length), then step S115 can be executed. . Conversely, if the data length of the audio signal lossless compressed data of the current frame is greater than the predetermined maximum length, step S105 is executed to perform lossy compression on the original audio data of the current frame, and then step S115 is executed.

In step S115, the compressed audio data, ie audio lossy compressed data or audio lossless compressed data, is transmitted.

Please refer to FIG. 2 , which is a flowchart of the decompression steps of the dynamic switching method for lossy or lossless compression of fixed-bandwidth audio information data according to an embodiment of the present invention.

As shown in FIG. 2 , the dynamic switching method for lossy or lossless compression of fixed-bandwidth audio information data in this embodiment may include steps S201 to S213.

In step S201, the header of the compressed audio data of the current frame is read.

In step S203, according to the read content, it is determined whether the current frame adopts lossless compression (ie, whether it is the above-mentioned audio signal lossless compression data). If yes, then step S205 is executed. If not, then step S207 is executed.

In step S205, lossless decompression is performed on the original audio data of the current frame.

In step S207, lossy decompression and improved inverse discrete cosine transform are performed on the original audio data of the current frame.

In step S209, it is judged whether the compressed audio data of the previous frame of the current frame adopts lossless compression. If the previous frame adopts lossless compression, it is not necessary to remove the overlapping part of the current frame and the previous frame through time domain aliasing cancellation (TDAC) to restore the previous frame, so step S213 is executed next. Conversely, if the previous frame is lossy compression, then steps S211 and S213 are performed in sequence.

In step S211, according to the Time Domain Aliasing Elimination (TDAC) method, the first half of the decompressed data of the current frame is superimposed with the second half of the improved inverse discrete cosine transform result of the previous frame to restore the previous frame. A frame of raw audio data.

In step S213, the decompressed original audio data of the previous frame is output.

After repeatedly performing the above steps of the embodiments of the present invention, under a fixed bandwidth, lossy compression or lossless compression can be individually selected according to the characteristics of each frame, so that the compression methods used for consecutive frames may also be the same or different. The main purpose of the present invention is that the compression methods used in consecutive frames can be different. For example, the previous frame adopts lossless compression, but the next frame (ie, the current frame) of the previous frame may be dynamically switched to lossy compression. The next frame may be dynamically switched to lossless compression, which is only illustrated here, and the present invention is not limited to this. Read the description in more detail below.

Please refer to FIG. 4 , which is a schematic diagram of decompression of the previous frame and the current frame of lossy compression according to the dynamic switching method of lossy or lossless compression of fixed bandwidth audio information data according to an embodiment of the present invention.

The steps S101 to S115 and S201 to S213 included in the dynamic switching method for lossy or lossless compression of fixed-bandwidth audio information data according to the embodiment of the present invention are applicable to the previous frame B, the previous frame B, the above frame B, The previous frame A of a frame, and the current frame C.

First, as shown in Figure 4, the second half of the improved inverse discrete cosine transform result of the previous frame B and the first half of the improved inverse discrete cosine transform result of the current frame C are combined according to the time domain aliasing cancellation method (TDAC). , superimposed to restore the original audio data of the previous frame.

As shown in Figure 3, the previous frame B applies the general MDCT window function. According to the time domain aliasing elimination method, the operation is as follows: The second half of IMDCT(MDCT(WA, W _R B)) is W _R (W _R B+(W _R B) _R )=W _R ² B+WW _R B _R , and the first half of IMDCT(MDCT(WB,WR C)) is W(WB-W _{R B R )} =W ² B-WW _R B _R , add the two halves (WR ² B+WW _R B _R )+(W ² B-WW _{R B R )} =(WR ² +W ² ) _B = _B , where (WR ² +W ² )=1, where B represents the original audio data of the previous frame (ie, the previous frame), R represents the inverse, and B _R represents the inverse waveform of the B waveform.

As mentioned above, when the audio data of each frame of lossy compression is used, when restoring, it is necessary to perform TDAC operation with the original audio data of the continuous previous frame and the next frame, so that the lossy compression can be restored. of each frame.

Please refer to FIG. 5 , which is a schematic diagram of decompression of a lossy compressed previous frame and a lossless compressed current frame according to a dynamic switching method for lossy or lossless compression of fixed-bandwidth audio information data according to an embodiment of the present invention.

Steps S101 to S115 and S201 to S213 included in the dynamic switching method for lossy or lossless compression of fixed-bandwidth audio information data according to the embodiment of the present invention are applicable to the previous frame B using lossy compression as shown in FIG. Lossless compressed current frame C.

First, as shown in FIG. 5 , the previous frame B is subjected to lossy decompression and improved inverse discrete cosine transform, wherein the second half of the transform result is WR ² B+WW _{R B R .}

The current frame C is subjected to lossless compression to obtain the original audio information data of the current frame C and previously in step S109, the window function of the improved discrete cosine transform is applied to the original audio information data of the previous frame B, and converted to obtain W ² B-WW _R B _R.

According to the time-domain aliasing elimination method, after adding W _R ² B+WW _R B _R and W ² B-WW _R B _R , the complete original audio data of the previous frame B is restored.

Please refer to FIG. 6 , which is a schematic diagram illustrating the decompression of the lossless compressed previous frame and the current frame according to a dynamic switching method for lossy or lossless compression of fixed-bandwidth audio information data according to an embodiment of the present invention.

The steps S101 to S115 and S201 to S213 included in the dynamic switching method for lossy or lossless compression of fixed-bandwidth audio information data according to the embodiment of the present invention are applicable to the previous frame B shown in FIG. frame C.

First, as shown in Figure 6, if the audio data of the previous frame B is lossless compressed, and the audio data of the current frame C is also lossless compressed, then the audio data of the previous frame is decompressed to restore the lossless compression data. The original audio data of the previous frame B.

As mentioned above, by adopting the original audio data of each frame of lossless compression, the original audio data of the frame can be directly restored after decompression, and it is not necessary to perform the MDCT operation with the original audio data of the continuous previous frame and the next frame. .

Please refer to FIG. 7 , which is a schematic diagram of decompression of a lossless compressed previous frame and a lossy compressed next frame according to a dynamic switching method for lossy or lossless compression of fixed-bandwidth audio information data according to an embodiment of the present invention.

Steps S101 to S115 and S201 to S213 included in the dynamic switching method for lossy or lossless compression of fixed-bandwidth audio information data according to the embodiment of the present invention are applicable to the previous frame B using lossless compression as shown in FIG. Compressed current frame C.

If the audio data of the previous frame B is lossless compressed, and the audio data of the current frame C is lossy compression, then the audio data of the previous frame is decompressed, and the original audio data of the previous frame B can be restored. .

As mentioned above, using the original audio data of each frame of lossless compression, the complete original audio data can be directly restored after decompression, without the need to perform continuous processing with the original audio data of the previous and next frames of this frame. Operation of TDAC. However, when restoring the original audio data of a frame with lossy compression, it is necessary to perform TDAC operation with the original audio data of the previous frame and the original audio data of the next frame, so that the lossy compression can be restored. of each frame.

In summary, the present invention provides a dynamic switching method for lossy or lossless compression of fixed bandwidth audio signal data, which has the following main features:

Dynamically switch lossy and lossless compression methods under fixed bandwidth;

Support for Modified Discrete Cosine Transform (MDCT);

Filter out the frames that cannot be compressed losslessly within a fixed bandwidth according to the amount of energy, and do not need to try lossless compression for each frame, reducing the time-consuming of compression operations;

The present invention directly compresses the original audio data of one frame without loss, unlike most of the lossless compression methods on the market, which calculate the difference between the original audio data and the lossy compression restoration.

For a further understanding of the features and technical content of the present invention, please refer to the following detailed descriptions and drawings of the present invention. However, the drawings provided are only for reference and description, and are not intended to limit the present invention.

The contents disclosed above are only preferred feasible embodiments of the present invention, and are not intended to limit the claims of the present invention. Therefore, any equivalent technical changes made by using the contents of the description and drawings of the present invention are included in the rights of the present invention. in the request.

Claims (8)

1. a method for dynamic switching of lossy or lossless compression of fixed bandwidth audio information data, is characterized in that, described method comprises the following steps: According to the set bandwidth, calculating a predetermined maximum energy value that can be losslessly compressed for each of the multiple frames; Read the original message data of each described frame; Calculate the energy value of the original audio data of each of the frames; and determine whether the energy value of the original audio data of each of the frames is greater than the predetermined maximum energy value, if so, use the frame's The original audio data is subjected to lossy compression to generate audio lossy compressed data, and if not, the original audio data of the frame is subjected to lossless compression to generate audio lossless compressed data. 2. method according to claim 1, is characterized in that, described method also comprises the following steps: calculating the square value of the amplitude of each of the waveforms in the multiple waveforms of the audio signal of the original audio data of each of the frames; summing the respective squared values of the plurality of waveforms to generate a total energy value; and dividing the total energy value by the number of the plurality of waveforms to obtain an average energy value, taking the The logarithm of the average energy value to calculate the energy of the raw audio data for each of the frames, expressed by the following equation: E=log((x ₁ ² +x ₂ ² +...+x _n ² )/n) , where E represents the energy of the original audio data of each frame, x ₁ to x _n represent the respective amplitudes of the multiple waveforms, and n represents the number of the multiple waveforms. 3. method according to claim 1, is characterized in that, described method also comprises the following steps: Determine whether the energy value of the original audio data of each of the frames is greater than the predetermined maximum energy value, and if so, perform lossy compression on the original audio data of the frame to generate the audio lossy compressed data , if not, go to the next step; Determine whether the previous frame of the current frame is lossless compression, if so, perform lossless compression on the original audio data of the current frame to generate the audio lossless compressed data, and do not perform subsequent steps, if not, according to perform subsequent steps in sequence; Applying the window function (window function) of the improved discrete cosine transform (Modified DiscreteCosine Transform, MDCT) to the original audio information of the above-mentioned frame, according to the time-domain aliasing elimination method (Time-Domain Aliasing Cancellation, TDAC) , to calculate functional audio data; and perform lossless compression on the functional audio data and the original audio data of the current frame to generate the audio lossless compressed data. 4. method according to claim 1, is characterized in that, described method also comprises the following steps: According to the set bandwidth, calculate the predetermined maximum length allowed after the original audio data of each of the frames is compressed; Determine whether the data length of the audio signal lossless compressed data of each frame is greater than the predetermined maximum length, if not, keep the audio signal lossless compressed data, if so, perform lossy compression on the original audio data of the frame To generate the audio lossy compressed data, the audio lossless compressed data having a data length greater than the predetermined maximum length is replaced. 5. method according to claim 1, is characterized in that, described method also comprises the following steps: Lossy compression is performed on the original audio data of the above-mentioned frame; Deciding to perform lossy compression on the original audio data of the current frame; The original information data of the previous frame and the original information data of the current frame are subjected to improved discrete cosine transform after applying the window function, and the function information data are calculated; lossy compression to generate the audio lossy compressed data. 6. The method according to claim 1, wherein the method further comprises the steps of: Lossy compression is performed on the original audio data of the above-mentioned frame; Deciding to perform lossless compression on the original audio data of the current frame; Applying the window function of the improved discrete cosine transform to the original audio data of the above-mentioned frame, according to the time-domain aliasing elimination method, to calculate the functional audio data; and combining the functional audio data with the current frame Lossless compression is performed together with the complete original audio data to generate the audio lossless compressed data. 7. The method according to claim 1, wherein the method further comprises the steps of: Perform lossless compression on the original audio data of the previous frame to generate the audio lossless compressed data of the previous frame; and decide to perform lossless compression on the original audio data of the current frame to generate The audio of the current frame is losslessly compressed data. 8. The method according to claim 1, wherein the method further comprises the steps of: Lossless compression is performed on the original audio data of the previous frame to generate the audio lossless compressed data; Deciding to perform lossy compression on the original audio data of the current frame; The original audio information data of the above-mentioned frame and the original audio information data of the current frame are subjected to improved discrete cosine transform after applying the window function to calculate the functional audio information data; Lossy compression.

Images