CN103258552B - How to adjust playback speed - Google Patents

Abstract

The invention provides a method for adjusting playing speed, which utilizes related data of frequency of audio data analyzed in an auditory perception decoding process to judge whether to discard or copy partial audio data so as to change the playing speed in the decoding process. Therefore, the invention does not need a large number of registers to register the audio data.

Description

How to adjust playback speed

technical field

The present invention relates to a media processing method and its device, in particular to a method and its device for adjusting media playback speed.

Background technique

When users listen to audio compression files such as MP3/WMA/AAC (MPEG-1AudioLayer3/WindowsMediaAudio/AdvancedAudioCoding) on a multimedia playback platform, they may speed up the playback speed to find the segment they want to listen to, or slow down the playback speed to listen carefully to the details of a segment (expansion). In order not to greatly distort the playback quality due to the change of the playback speed, the TimeScale Modification (TSM) method is widely used in the industry. Traditional time-domain duration adjustment methods, such as Overlap Add (OLA) or Synchronized Overlap Add (SynchronizedOLA), mainly divide the input audio signal into many segment signals, and overlap two segment signals adjacent in time. And the overlapped area is weighted by fading out and fading in. However, such a duration adjustment method requires a large number of registers to store segment signals.

In addition, the existing duration adjustment method also utilizes the Short-Time Discrete Fourier Transform (ST-DFT) to convert the input audio signal from the time domain to the frequency domain for analysis, but when it is converted back to the time domain after analysis, it will Having problems with phase distortion.

U.S. Patent Publication No. 20050010397 discloses a time length adjustment method using short-time Fourier transform, which mainly selects specific spectral bands (SpectralBand) of audio data based on changes in the frequency response of human auditory perception. Metric, used for phase locking. Each spectrum band is marked with a spectrum peak (SpectralPeak). Spectral peaks and spectral lines close to or far from the spectral peaks undergo different phase processing. Therefore, when the subsequent audio data must be converted back to the time domain for signal window reconstruction (Reconstruction), it is easy to cause phase distortion and affect the playback quality.

Contents of the invention

Therefore, the present invention mainly provides a method and device for adjusting playback speed that do not require a large number of registers.

The present invention discloses a method for adjusting playback speed, including: an auditory perception decoding device receives an audio data; the auditory perception decoding device performs frequency analysis of a first audio frame of the audio data; obtains a frequency analysis related information First frequency domain analysis data; receive a speed adjustment signal; when the speed adjustment signal indicates to speed up the playback speed of the audio data, according to the first frequency domain analysis data, determine whether to discard the first audio frame; in the speed adjustment When the signal indicates to slow down the playing speed of the audio data, according to the first frequency domain analysis data, it is judged whether to copy the first audio frame; when the first audio frame is judged to be discardable, the auditory perception decoding device discards the At least a part of the data of the first audio frame; and when the first audio frame is determined to be copyable, the auditory perception decoding device copies at least a part of the data of the first audio frame.

The present invention also discloses a method for adjusting playback speed, which includes: an auditory perception decoding device receives audio data, the audio data includes a plurality of audio frames; the auditory perception decoding device performs frequency analysis on the plurality of audio frames; receiving A speed adjustment signal; when the speed adjustment signal indicates that the playback speed of the audio data is accelerated to (N/(N-M)) times, each audio frame of the N consecutive audio frames in the plurality of audio frames is used to perform An adjustment judgment program for judging whether the processed audio frame can be discarded, wherein N and M are positive integers; when it is judged through the adjustment judgment program that there are M audio frames among the N consecutive audio frames that can be discarded, the auditory perception The decoding device discards at least a part of the data of the M audio frames; when the speed adjustment signal indicates to slow down the playback speed of the audio data to (N/(N+M)) times, the N audio frames in the plurality of audio frames Each audio frame of the continuous audio frame executes an adjustment judging program for judging whether the processed audio frame can be copied; and through the adjustment judging program, it is judged that there are M audio frames in the N consecutive audio frames that can be copied , the auditory perception decoding device copies at least a part of data of the M audio frames. Wherein, the adjustment judging procedure includes: obtaining a first frequency domain analysis data corresponding to a processed first audio frame and related to the frequency analysis; when the speed adjustment signal indicates to increase the playback speed of the audio data, according to the According to the first frequency domain analysis data, it is judged whether to discard at least a part of the data of the first audio frame; and when the speed adjustment signal indicates to slow down the playback speed of the audio data, according to the first frequency domain analysis data, it is judged whether to copy the audio data At least a portion of data for the first audio frame.

The present invention also discloses a method for accelerating playback speed, which includes an auditory perception decoding device receiving an audio data; the auditory perception decoding device performs a frequency analysis of a first audio frame of the audio data; and obtains a first frequency analysis related to the frequency analysis. A frequency domain analysis data; receiving an acceleration adjustment signal; judging whether to discard the first audio frame according to the first frequency domain analysis data; and when the first audio frame is judged to be discardable, the auditory perception decoding device according to The acceleration adjusts a playback speed indicated by the signal, discarding at least a part of data of the first audio frame.

The present invention also discloses a method for slowing down the playback speed, which includes an auditory perception decoding device receiving an audio data; the auditory perception decoding device performs frequency analysis of a first audio frame of the audio data; obtains a frequency analysis related information First frequency domain analysis data; receiving a slowing speed adjustment signal; judging whether to copy the first audio frame according to the first frequency domain analysis data; and when the first audio frame is judged to be copyable, the auditory perception The decoding device copies at least a part of data of the first audio frame according to a playback speed indicated by the slowing down adjustment signal.

The method and device for adjusting the playback speed provided by the present invention do not require a large number of registers.

Description of drawings

FIG. 1 is a flow chart of a process in Embodiment 1 of the present invention.

FIG. 2 is a flow chart of a process in Embodiment 1 of the present invention.

FIG. 3 is a flow chart of a process in Embodiment 1 of the present invention.

FIG. 4A and FIG. 4B are flowcharts of a process according to Embodiment 1 of the present invention.

FIG. 5 is a flow chart of a process in Embodiment 1 of the present invention.

FIG. 6 is a flow chart of a forced copy/discard process according to Embodiment 1 of the present invention.

FIG. 7 is a schematic block diagram of a speed adjusting device according to an embodiment of the present invention.

Figure number:

10, 20, 30, 40, 60 processes

50 speed adjustment device

500 audio reading device

510 processor unit

520 storage units

530 input unit

540 output unit

522 program code

100, 102, 104, 106, 108, 110, 112, 114, 116, 118, 200, 202, 204, 206, 208, 210, 212, 300, 302, 304, 306, 308, 310, 312, 314, 316, 318, 320, 322, 324, 400, 402, 404, 406, 408, 410, 412, 414, 416, 418, S910, S920, S930, S940, S950, 602, 604, 606, 608, 610, 612, 614, 616, 618, 620, 622, 624, 626 steps

Detailed ways

Fig. 1 is a flow chart of speed adjustment for changing the speed of sound playback in an embodiment of the present invention. Please refer to Fig. 1, this embodiment is applicable to the multimedia player such as television set, set-top box, digital video disc player, MP3 player, in order to determine the audio frequency of the sound effect data read when playing according to the available buffer capacity of the player. The number of frames, and according to the distribution of the sum of the differences of these audio frames, determine the content of the audio data to be played, and provide a better playback effect.

Firstly, audio data segmented into multiple audio frames is acquired (step S910). Wherein, the audio data includes audio data of TV programs or multimedia files, and each audio frame in the audio data includes multiple frequency components.

After the audio data of the audio frame is obtained, frequency domain analysis of the audio frame can be performed (step S920 ). Wherein, a frequency component is calculated when performing frequency domain analysis on the audio frame, and the calculation method of the frequency component may be to use Fast Fourier Transform (FastFourierTransform, FFT) to obtain frequency domain complex values of each frequency point, so as to convert a The audio frame is divided into multiple FFT frequency points, and then the energy values of these FFT frequency points are respectively calculated as the energy of each frequency component. Another way is to use a filter bank (FilterBank) to divide an audio frame into multiple sub-bands (Sub-band), and calculate the energy value of each sub-band as the energy of each frequency component.

After the playback device receives the instruction to change the playback speed input by the user, it will judge whether the user wants to play fast or slow (step S930), and dynamically adjust the playback speed according to the multiple A of the playback speed and the above-mentioned multiple audio frames. The ratio of the audio frame of the audio data to be played, and the audio data is played, where A is a positive number.

Among them, when the user wants to perform fast playback, the playback device will delete the audio frames that meet the discarding conditions according to an adjustment judgment process (the principle will be described in detail below) to achieve fast playback (step S950) (for example, the audio frame 1, 2, 3, wherein 2 is deleted, then play 1, 3); On the contrary, when the user wants to perform slow playback, the playback device will copy the audio frames that meet the copying conditions according to the adjustment and judgment process to achieve Play slowly (step S940) (for example, audio frames 1, 2, 3, where 2 is repeated, then play 1, 2, 2, 3). In practical applications, the multiple A of the playback speed may be a decimal number, such as 1.75 times or 0.75 times.

For example, when the playback device executes fast playback at 2x speed, it can discard the B/2 audio frames that meet the discarding conditions in the B audio frames according to the judgment process, so as to play out the audio data that changes greatly The content of the audio frame, so that the user can still hear important information in the audio data during fast playback. On the other hand, when the playback device executes slow playback at a speed of 0.66 times, the B/2 audio frames that meet the copying conditions can be repeated once among the B audio frames according to the judgment process, so as to repeatedly play out the audio data The content of the audio frame changes less in the middle, so that the user can hear the extended audio content without changing the pitch during the slow playback process. Through the above method, the playback device can use the original available buffer capacity to copy and discard the audio frame data without affecting the normal playback of the audio data. In other words, the playback device can save the use of buffer registers and can Most of the detailed characteristics of the sound are maintained, providing users with a listening effect of quick browsing and focused playback.

It is worth mentioning that the above method of distinguishing frequency components and calculating their energy is only an embodiment of the present invention. Those skilled in the art can change the FFT length or the number of filter bank subbands according to actual needs, or use Wavelet transform, discrete cosine transform (DiscreteCosine Transform, DCT) or other technologies are used to distinguish frequency components and calculate their energy, and this embodiment does not limit the scope thereof.

For compressed audio data, such as: MPEG, AC3, DTS, WMA, AAC, etc., it has been divided into audio frames during compression, and each frequency component in each audio frame is Compression is performed after calculation. Therefore, when playing the compressed audio data of the above specifications, the playback device only needs to decompress the received compressed audio data to obtain the audio data that has been divided into multiple audio frames and all the frequency components in each audio frame, which can be directly Calculate the energy of these frequency components.

Please refer to FIG. 2 , FIG. 3 , FIG. 4A and FIG. 4B . FIG. 2 is a flow chart of speed adjustment in one embodiment of the present invention, and FIG. 3 , FIG. 4A and FIG. 4B are flow charts of adjustment judgment in another embodiment of the present invention. The speed adjustment process 10 can be implemented on a perceptual audio decoding device, so as to cooperate with the adjustment judgment process 20 to adjust the audio playback speed under the perceptual audio decoding program. The speed adjustment process 10 includes the following steps:

Step 100: start.

Step 102: Receive an audio frame of an audio data.

Step 104: Perform entropy decoding (Entropy Decoding) of the audio frame.

Step 106: Perform inverse quantization (InverseQuantization) of the audio frame.

Step 108 : Perform frequency analysis of the audio frame according to an auditory perception module, and execute the adjustment judgment process 20 .

Step 110: According to a judgment result output by the adjustment judgment process 20, judge whether to discard the audio frame? If yes, go to step 118; if not, go to step 112.

Step 112: Perform an inverse Modified Discrete Cosine Transform (IMDCT) of the audio frame according to the window type of the audio frame.

Step 114: According to the determination result of the adjustment determination process 20, determine whether to copy the audio frame? If yes, go to step 116; if not, go to step 118.

Step 116: copy the audio frame and preset the judgment result of the next time as "no copy", and execute step 112.

Step 118: When there is a next audio frame, receive the next audio frame of the audio frame, and proceed to step 104.

As can be seen from the above, the speed adjustment process 10 performs auditory perception decoding on each audio frame of the audio data one by one, and the audio data may be audio data in compressed formats such as MP3/WMA/AAC. First, each audio frame is subjected to entropy decoding, such as Huffman decoding. Next, the audio frame is dequantized, which may include decoding the scale factor (ScaleFactor) used for quantization at the encoder. After the dequantization is completed, the adjustment judgment process 20 judges whether the audio frame needs to be copied, discarded or not, according to the frequency domain analysis data (hereinafter referred to as the frequency domain analysis data) and a speed adjustment signal indication, and generates related critical result. When the speed adjustment signal indicates to speed up the playback speed of the audio data, the adjustment judgment process 20 judges whether to discard the audio frame according to the frequency domain analysis data; when the speed adjustment signal indicates to slow down the playback speed of the audio data, the adjustment judgment process 20 is based on Analyze the data in the frequency domain to determine whether to copy the audio frame. The speed adjustment signal indication can be generated according to the user changing the playback speed by using the audio playback system. The detailed operation principle of the adjustment judgment process 20 will be described later.

The auditory perception decoding device first judges whether the audio frame needs to be discarded according to the judgment result, and if it needs to be discarded, then as described in step 118, it decodes the next audio frame instead, so that during the playback of the audio data, the audio frame The data will not be played, in order to achieve the purpose of speeding up the playback. On the contrary, if the audio frame does not need to be discarded, the auditory perception decoding device performs the inverse improved discrete cosine transform and synthesis of the audio frame according to the window type of the audio frame, which is a kind of inverse time-frequency conversion, which can be long window or The unit of the short window is to convert the frequency domain data of the audio frame (which can be included in the frequency domain analysis data) into time domain data. After an inverse improved discrete cosine transform is completed, the auditory perception decoding device will judge whether the audio frame needs to be copied. If it needs to be copied, the default judgment result of the next round is "No", that is, no copying is required in the next judgment. In addition, the copied audio frame is subjected to an inverse improved discrete cosine transform, so that during the playback of the audio data, the audio frame data will be played twice to achieve the purpose of slowing down the playback speed. Since the judgment result is set such that this audio frame does not need to be copied when it is judged next time, the speed adjustment process 10 proceeds to decode the next audio frame. Therefore, according to the determination result of the adjustment determination process 20, the speed adjustment process 10 can discard/copy each audio frame of the audio data to speed up/slow down the playback speed.

The undiscarded short-window data will still go through the inverse improved DCT and windowing in step 112, and will be played after the auditory perception decoding is completed.

Please note that the audio frame in the present invention is the smallest unit of data discarding and duplication, and each audio format will contain different relative ratios of long and short windows; for example: in A format, the length of a long window is regarded as an audio frame , and the length of a long window may be a combination of four short windows or several short windows, that is, four short windows or several short windows will be regarded as an audio frame; another example, in the B format, an audio frame It depends on the matching of its long and short windows. In auditory perceptual coding, the audio frame data composed of long windows represents a relatively stable signal range, while the audio frame data composed of short windows represents a relatively sharply changing signal range. Therefore, in adjusting the playback speed, only copying or discarding the data belonging to the long window will not affect the playback quality.

Therefore, the aforementioned frequency-domain analysis data may include a window type indicator, which is used to indicate that the window type used for the inverse modified DCT of the audio frame is a long window or a short window. In this case, the adjustment judgment process 20 in FIG. 3 includes the following steps:

Step 200: Receive a speed adjustment signal indication.

Step 202: Obtain frequency-domain analysis data of a window-type indicator including the audio frame.

Step 204: Determine whether the window type indicator indicates that the audio box belongs to the long window type? If yes, go to step 208 ; if not, go to step 206 .

Step 206: Generate a judgment result of "do not discard/do not copy".

Step 208: Determine whether the speed adjustment signal indicates to speed up the playback speed of the audio data? If yes, go to step 210 ; if not, go to step 212 .

Step 210: Generate a judgment result of "discard".

Step 212: Generate a judgment result of "copy".

The adjustment judgment process 20 in FIG. 3 mainly uses the window type of the audio frame as a criterion for whether the audio frame needs to be discarded/duplicated. As can be seen from the above, when the speed adjustment signal indicates to increase the playback speed of the audio data, and the window type indicator indicates that the audio frame belongs to a long window, the determination of the adjustment and determination process 20 can be discarded. When the speed adjustment signal indicates to slow down the playback speed of the audio data, and the window type indicator indicates a long window, the adjustment determination process 20 determines that copying is possible. In other words, when the window type indicator indicates that the audio frame belongs to other window types (such as short window, long-to-short window, etc.), the adjustment judgment process 20 indicates the speed adjustment process 10 and the audio frame does not need to be discarded or copied.

In addition to the window index, the aforementioned frequency domain analysis data may further include a spectral line (SpectralLine) data of the audio frame. Figure 4A and Figure 4B adjust the judgment process 20 while using the window type and spectral line data of the audio frame as the criteria for whether the audio frame needs to be discarded/duplicated, which includes the following steps:

Step 300: Receive a speed adjustment signal indication.

Step 302: Obtain frequency domain analysis data of the audio frame, which includes a window index and a spectral line data.

Step 304: Determine whether the window type indicator indicates that the audio box belongs to the long window type? If yes, go to step 308 ; if not, go to step 306 .

Step 306: Generate a judgment result of "do not discard/do not copy".

Step 308: Divide the spectral line data into multiple frequency band units, and calculate an energy sum Pcurr of the multiple frequency band units.

Step 310: Obtain an energy sum Pprev corresponding to the plurality of frequency band units of the previous audio frame of the audio frame.

Step 312: Calculate a total energy difference Pdiff=Pprev-Pcurr.

Step 314: Determine |Pdiff|<THa? If yes, go to step 316 ; if not, go to step 306 .

Step 316: Determine if Pdiff>THb? If yes, go to step 318 ; if not, go to step 306 .

Step 318: Determine if Pprev<THc and Pcurr<THc? If yes, go to step 320 ; if not, go to step 306 .

Step 320: Determine whether the speed adjustment signal indicates to speed up the playback speed of the audio data? If yes, go to step 322 ; if not, go to step 324 .

Step 322: Generate a judgment result of "discard".

Step 324: Generate a judgment result of "copy".

It can be seen from the above that for other audio frame window types other than the long window, the adjustment and judgment process 20 in FIG. 4A and FIG. 4B also indicates the speed adjustment process 10 that the audio frame does not need to be discarded or copied. In FIG. 4A and FIG. 4B adjustment judgment process 20, the frequency band unit division of spectral line data can be different according to system requirements, for example, spectral line data can be directly divided into frequency band units that are continuous and occupy the frequency range of all audio frames, In this way, the energy sums Pcurr and Pprev are calculated to be the total energy of the audio frame and the previous audio frame respectively. Alternatively, the spectral line data can be divided into frequency band units classified as tone-like or noise-like according to signal flatness. The division of the frequency band unit and its energy calculation can refer to the method of the frequency component, and the detailed operation will not be repeated here. In addition, the adjustment judgment process 20 in Fig. 4A and Fig. 4B defines the threshold values THa, THb and THc, which are respectively the values of the system based on the single-frequency signal, the post-masking effect in the auditory perception and the silence (Silence) signal. The characteristics of the threshold given by the characteristics should be well known to those skilled in the art, so details are not described here. Therefore, the adjustment judgment process 20 in FIG. 4A and FIG. 4B will only indicate that the audio frame needs to be discarded or copied when the following conditions are met. The threshold THa of the total difference; (ii) The total energy difference Pdiff is greater than the threshold THb related to the back shadowing effect; (iii) The total energy Pprev is less than the threshold THc related to the silence signal and the total energy Pcurr is also less than Threshold THc. If any of the above conditions is not met, the adjustment judgment process 20 in FIG. 4A and FIG. 4B indicates the speed adjustment process 10 that the audio frame does not need to be discarded or copied. When all the conditions (i), (ii) and (iii) above are met, the adjustment judgment process 20 instructs the speed adjustment process 10 that the audio frame needs to be discarded/duplicated according to the speed up/slow down indication of the speed adjustment signal.

In the adjustment and judgment process 20 shown in FIG. 4A and FIG. 4B , those skilled in the art can use any combination of the length window and the conditions (i) to (iii) or any combination of the four as a criterion for judging whether discarding/copying is required, It is not limited to fully meeting these four conditions. For example, an audio frame may be discarded or duplicated when it is determined that the window is long. The aforementioned auditory perception decoding process, such as entropy decoding, inverse quantization, frequency analysis of audio frames and inverse improved discrete cosine transform, should be well known to those skilled in the art. The present invention mainly utilizes the existing frequency analysis in auditory perception decoding Information, as a reference for copying or discarding audio frames, so it will not encounter the problem of phase distortion in the subsequent signal reconstruction (Reconstruction) of auditory perception decoding. In addition, the present invention can judge the audio frame to be copied or discarded in real time. Therefore, in the process of adjusting the playback speed, there is no need for a large number of registers to store the data of the audio frame before and after, thereby saving the production cost.

Please refer to FIG. 5 , which is a flow chart of a speed adjustment process 40 according to an embodiment of the present invention. The speed adjustment process 40 can be implemented on the auditory perception decoding device, using the speed adjustment process 10 to discard, copy or process each audio frame normally, and then adjust the playback speed of the audio data to the playback speed expected by the user, which includes the following steps:

Step 400: Receive an audio data, and the audio data is input in continuous audio frames.

Step 402: Receive and determine a speed adjustment signal. When the speed adjustment signal indicates that the playback speed of the audio data is accelerated to (N/(N-M)) times, step 404 is executed; when the speed adjustment signal indicates that the playback speed of the audio data is slowed down to (N/(N+M) )) times, execute step 410.

Step 404: Through the speed adjustment process 10, determine whether each audio frame of the N consecutive audio frames in the plurality of audio frames can be discarded.

Step 406: Determine whether there are M audio frames among the N consecutive audio frames that can be discarded? If yes, go to step 408 ; if not, go to step 420 .

Step 408: The auditory perception decoding device discards at least a part of data of the M audio frames.

Step 410: Obtain the next group of N consecutive audio frames of the audio data, and proceed to step 404.

Step 412: Through the speed adjustment process 10, determine whether each audio frame of the N consecutive audio frames in the plurality of audio frames can be copied.

Step 414: Determine whether there are M audio frames among the N consecutive audio frames that can be copied? If yes, go to step 416 ; if not, go to step 424 .

Step 416: The auditory perception decoding apparatus copies at least a part of data of the M audio frames.

Step 418: Obtain the next group of N consecutive audio frames of the audio data, and proceed to step 412.

Step 420: Determine whether the K groups of N consecutive audio frames have been processed, and whether there are less than K×M audio frames in the K group of N consecutive audio frames that can be discarded? If yes, go to step 422 ; if not, go to step 410 .

Step 422: Discard all or part of the data of the subsequent audio frames.

Step 424: Determine whether the K groups of N consecutive audio frames have been processed, and whether there are less than K×M audio frames in the K group of N consecutive audio frames that can be copied? If yes, go to step 426 ; if not, go to step 418 .

Step 426: All or part of the data of the audio frame after copying.

According to the speed adjustment process 40, when the speed adjustment signal indicates that the playback speed is increased to (N/(N-M)) times, each audio frame of each group of N consecutive audio frames will be judged by the speed adjustment process 10 whether it can be discarded. When M audio frames among the N consecutive audio frames are determined to be discarded, the auditory perception decoding device may discard at least a part of the data of the M audio frames, for example discard all of the M audio frames or their long window type data. Similarly, when the speed adjustment signal indicates that the playback speed is slowed down to (N/(N+M)) times, the execution of each audio frame will be judged by the speed adjustment process 10 whether it can be copied. When it is determined that M audio frames among the N consecutive audio frames can be copied, the auditory perception decoding device may copy at least a part of data of the M audio frames. The speed adjustment process 40 copies or discards M audio frame data (or long window data) under N audio frames, so as to obtain N/(N±M) times the playback speed desired by the user.

In addition, in the speed adjustment process 40 , there may be no M audio frames in consecutive groups of N audio frames that can be discarded or copied. In this case, as shown in steps 420-422 and 424-426, this embodiment can set a maximum limit value K, when there are no K×M audio frames in the continuous K groups of N audio frames, it can be discarded or copied , the auditory perception decoding device starts to forcibly discard or copy all or part of the data of the subsequent audio frame, so that the playback speed can reach the speed expected by the user. For example, when only (K×M-L) audio frames in K consecutive groups of N audio frames meet the conditions for discarding or copying, the auditory perception decoding device discards or copies all received L audio frames Or a part of the data to maintain the playback speed at (N/(N-M)) times or (N/(N+M)) times.

In the embodiment of the present invention, the process of judging whether there are K×M audio frames in the continuous K groups of N audio frames that can be discarded or copied does not necessarily only require M audio frames for each group of N audio frames, as long as the total It only needs to have K×M audio frames. For example, when K is set to 2, if the first group of N audio frames has (M-1) audio frames that can be discarded or copied, then the second group of N audio frames needs to have (M+1) audio frame, so that the playback speed can reach the speed expected by the user.

Please refer to FIG. 6 , which is a schematic block diagram of a process 60 according to an embodiment of the present invention. The process 60 is used to realize the above-mentioned concept of forcibly discarding or duplicating audio frames, which includes the following steps:

Step 602: Receive and determine a speed adjustment signal. When the speed adjustment signal indicates that the playback speed of the audio data is accelerated to (N/(N-M)) times, step 604 is executed; when the speed adjustment signal indicates that the playback speed of the audio data is slowed down to (N/(N+M) )) times, execute step 616.

Step 604: Set a parameter i=1.

Step 606: Determine the number of discardable audio frames among the newly received N audio frames.

Step 608: Determine whether there are i×M audio frames Ndiscard that can be discarded in total? If not, go to step 610; if yes, go to step 614.

Step 610: i=i+1.

Step 612: Determine whether i is greater than a threshold K? If not, go to step 606; if yes, go to step 614.

Step 614: Execute a reset process, which includes setting i=1, discarding Ndiscard audio frames and forcibly discarding (K×M-Ndiscard) audio frames in subsequent newly received audio frames, or only discarding Ndiscard audio frames , but do not forcibly discard (K×M-Ndiscard) audio frames in subsequent newly received audio frames.

Step 616: Set i=1.

Step 618: Determine the number of audio frames that can be copied among the newly received N audio frames.

Step 620: Determine whether the total number Ncopy of audio frames that can be copied is i×M? If not, go to step 622; if yes, go to step 626.

Step 622: i=i+1.

Step 624: Determine whether i is greater than the threshold value K? If not, go to step 618; if yes, go to step 626.

Step 626: Execute a reset process, which includes setting i=1, copying Ncopy audio frames and forcing (K×M-Ncopy) audio frames to be copied in subsequent newly received audio frames, or only copying Ncopy audio frames , but not forcibly copying (K×M-Ncopy) audio frames in subsequent newly received audio frames.

According to the process 60, in order not to excessively reduce the quality of audio playback, the embodiment of the present invention does not forcefully copy or discard less than K×M audio frames in subsequent newly received audio frames, that is, (K×M-Ndiscard) or (K×M-Ncopy) audio frames. The detailed operating principles of the other process 60 have been disclosed above, so they will not be repeated here.

Please refer to FIG. 7 , which is a schematic block diagram of a speed adjusting device 50 according to an embodiment of the present invention. The speed adjusting device 50 includes an audio reading device 500 , a processor unit 510 , a storage unit 520 , an input unit 530 and an output unit 540 . The audio reading device 500 can be a CD/DVD player or a network card device, used to obtain an audio data AU_DATA and send it to the processor unit 510 for processing through the storage unit 520 . The input unit 530 can be a keyboard, a mouse, a voice input or other devices through which the user can interact with the speed adjustment device 50 , and is used to generate a speed adjustment signal PLR_ADJ to the processor unit 510 according to the user input signal. The storage unit 520 can be a non-volatile memory for storing the program code 522, which can be processed by the processor unit 510 to implement any one of the aforementioned processes (such as the adjustment judgment process 20, the speed adjustment process 40, etc.) or a combination thereof. The output unit 540 can be a speaker for playing the audio data processed by the processor unit 510 . For example, when the user speeds up the playback speed through the input unit 530, the processor unit 510 can use the speed adjustment process 40 to discard the audio frames of the played audio data according to the corresponding speed adjustment signal PLR_ADJ, and the audio frames will be discarded. The discarded audio data is sent to the playback unit 540 for playback, so that the user can perceive audio acceleration. Since the speed adjusting device 50 is mainly used to realize any one of the aforementioned processes (such as the adjustment judgment process 20 , the speed adjustment process 40 , etc.) or a combination thereof, please refer to the foregoing for the main operating principles.

In summary, the embodiment of the present invention utilizes the frequency-related data (such as window type/spectrum line data) of the audio data analyzed in the auditory perception decoding process to determine whether to discard or copy part of the audio data, so that the decoding process can also Achieving a change in playback speed. In this way, the present invention does not require a large number of registers to store audio data.

The above descriptions are only preferred embodiments of the present invention, and all equivalent changes and modifications made according to the claims of the present invention shall fall within the scope of the present invention.

Claims (17)

1. adjust a method for broadcasting speed, it is characterized in that, the method for described adjustment broadcasting speed includes:

One Auditory Perception decoding device receives a voice data;

Described Auditory Perception decoding device carries out the frequency analysis of one first audio frequency frame of described voice data;

Obtain the one first frequency-domain analysis data about described frequency analysis;

Receive a speed adjustment signal, judge to accelerate broadcasting speed or to slow down broadcasting speed to carry out adjustment broadcasting speed;

It is characterized in that comprising:

When described speed adjustment signal designation accelerates the broadcasting speed of described voice data, according to described first frequency-domain analysis data, judge whether to give up described first audio frequency frame;

In described first audio frequency frame be judged as can give up time, described Auditory Perception decoding device gives up described first audio frequency frame; And

When described speed adjustment signal designation slows down the broadcasting speed of described voice data, according to described first frequency-domain analysis data, judge whether to copy described first audio frequency frame;

In described first audio frequency frame be judged as can copy time, described Auditory Perception decoding device copies described first audio frequency frame;

Described first frequency-domain analysis data comprise the window type index being used to refer to the window type that described first audio frequency frame is changed for a frequency domain to the time domain used in described Auditory Perception decoding device;

When described speed adjustment signal designation accelerates the broadcasting speed of described voice data, judge whether that giving up described first audio frequency frame comprises according to described first frequency-domain analysis data: the broadcasting speed accelerating described voice data in described speed adjustment signal designation, and described window type index instruction described first audio frequency frame is when belonging to long window type, judgement can give up described first audio frequency frame; And judge whether that copying described first audio frequency frame comprises according to described first frequency-domain analysis data when described speed adjustment signal designation slows down the broadcasting speed of described voice data: the broadcasting speed slowing down described voice data in described speed adjustment signal designation, and described window type index instruction described first audio frequency frame is when belonging to long window type, judgement can copy described first audio frequency frame.

2. the method for claim 1, is characterized in that, described first frequency-domain analysis data separately comprise the spectrum line data of described first audio frequency frame.

3. method as claimed in claim 2, is characterized in that, when described speed adjustment signal designation accelerates the broadcasting speed of described voice data, judge whether that giving up described first audio frequency frame comprises according to described first frequency-domain analysis data:

When the described first audio frequency frame of described window type index instruction belongs to long window type, described spectrum line Data Placement is gone out multiple band unit;

Calculate one first energy sum total of described multiple band unit;

The one second audio frequency frame obtaining described voice data corresponds to one second energy sum total of described multiple band unit, and described second audio frequency frame is the previous audio frequency frame by the process of described Auditory Perception decoding device of described first audio frequency frame;

Calculate an energy sum total poor, poor=described second energy sum total of described energy sum total-described first energy sum total;

Absolute value in described energy sum total difference is less than one first threshold value being relevant to class simple signal energy sum total difference, described energy sum total difference be greater than be relevant to Auditory Perception after one second threshold value, the described second energy sum total of covering be less than one the 3rd threshold value that is relevant to quiet audio number and at least one condition that described first energy sum total is less than above three conditions of described 3rd threshold value meets time, judge to give up the data belonging to long window type in described first audio frequency frame or described first audio frequency frame; And

When described speed adjustment signal designation slows down the broadcasting speed of described voice data, judge whether that copying described first audio frequency frame comprises according to described first frequency-domain analysis data:

When the described first audio frequency frame of described window type index instruction belongs to long window type, described spectrum line Data Placement is gone out described multiple band unit;

Calculate described first energy sum total;

Obtain described second energy sum total;

Calculate described energy sum total poor;

Absolute value in described energy sum total difference is less than one first threshold value being relevant to class simple signal energy sum total difference, described energy sum total difference be greater than be relevant to Auditory Perception after one second threshold value, the described second energy sum total of covering be less than one the 3rd threshold value that is relevant to quiet audio number and at least one condition that described first energy sum total is less than above three conditions of described 3rd threshold value meets time, judge to copy the data belonging to long window type in described first audio frequency frame or described first audio frequency frame.

4. method as claimed in claim 3, it is characterized in that, when the described first audio frequency frame of described window type index instruction belongs to long window type, described spectrum line Data Placement is gone out described multiple band unit to comprise, when the described first audio frequency frame of described window type index instruction belongs to long window type, according to the flatness of spectrum line data, described spectrum line Data Placement is gone out to classify as class simple signal or the described multiple band unit for noise like.

5. adjust a method for broadcasting speed, it is characterized in that, the method for described adjustment broadcasting speed includes:

One Auditory Perception decoding device receives a voice data, and described voice data comprises multiple audio frequency frame;

Described Auditory Perception decoding device carries out the frequency analysis of described multiple audio frequency frame;

Receive a speed adjustment signal;

The feature of described method is to comprise:

During in the broadcasting speed that described speed adjustment signal designation accelerates described voice data to (N/ (N-M)) times, to the adjustment determining program whether each audio frequency frame execution audio frequency frame be used for handled by judgement of the N number of continuous audio frequency frame in described multiple audio frequency frame can be given up, wherein N, M are positive integer;

In by described adjustment determining program, when judging have M audio frequency frame to give up in described N number of continuous audio frequency frame, described Auditory Perception decoding device gives up the data at least partially of described M audio frequency frame;

During in the broadcasting speed that described speed adjustment signal designation slows down described voice data to (N/ (N+M)) times, the adjustment determining program whether the audio frequency frame handled by being used for judging can copy is performed to each audio frequency frame of the N number of continuous audio frequency frame in described multiple audio frequency frame; And

In by described adjustment determining program, when judging have M audio frequency frame to copy in described N number of continuous audio frequency frame, described Auditory Perception decoding device copies the data at least partially of described M audio frequency frame;

Described adjustment determining program comprises:

Obtain corresponding to one first handled audio frequency frame, one first frequency-domain analysis data about described frequency analysis;

When described speed adjustment signal designation accelerates the broadcasting speed of described voice data, according to described first frequency-domain analysis data, judge whether the data at least partially giving up described first audio frequency frame; And

When described speed adjustment signal designation slows down the broadcasting speed of described voice data, according to described first frequency-domain analysis data, judge whether the data at least partially copying described first audio frequency frame;

Described first frequency-domain analysis data comprise the window type index being used to refer to the window type that described first audio frequency frame is changed for the frequency domain in described Auditory Perception decoding device to time domain;

When described speed adjustment signal designation accelerates the broadcasting speed of described voice data, judge whether that giving up described first audio frequency frame comprises according to described first frequency-domain analysis data: the broadcasting speed accelerating described voice data in described speed adjustment signal designation, and described window type index instruction described first audio frequency frame is when belonging to long window type, judge to give up the data belonging to long window type in described first audio frequency frame or described first audio frequency frame; And judge whether that copying described first audio frequency frame comprises according to described first frequency-domain analysis data when described speed adjustment signal designation slows down the broadcasting speed of described voice data: the broadcasting speed slowing down described voice data in described speed adjustment signal designation, and described window type index instruction described first audio frequency frame is when belonging to long window type, judge to copy the data belonging to long window type in described first audio frequency frame or described first audio frequency frame.

6. method as claimed in claim 5, it is characterized in that, described first frequency-domain analysis data separately comprise the spectrum line data of described first audio frequency frame.

7. method as claimed in claim 6, is characterized in that, when described speed adjustment signal designation accelerates the broadcasting speed of described voice data, judge whether that giving up described first audio frequency frame comprises according to described first frequency-domain analysis data:

When the described first audio frequency frame of described window type index instruction belongs to long window type, described spectrum line Data Placement is gone out multiple band unit;

Calculate one first energy sum total of described multiple band unit;

The one second audio frequency frame obtaining described voice data corresponds to one second energy sum total of described multiple band unit, and described second audio frequency frame is the previous audio frequency frame by the process of described Auditory Perception decoding device of described first audio frequency frame;

Calculate an energy sum total poor, poor=described first energy sum total of described energy sum total-described second energy sum total;

Absolute value in described energy sum total difference is less than one first threshold value being relevant to class simple signal energy sum total difference, described energy sum total difference be greater than be relevant to Auditory Perception after one second threshold value, the described second energy sum total of covering be less than one the 3rd threshold value that is relevant to quiet audio number and at least one condition that described first energy sum total is less than three conditions of described 3rd threshold value meets time, judge to give up the data belonging to long window type in described first audio frequency frame or described first audio frequency frame; And

When described speed adjustment signal designation slows down the broadcasting speed of described voice data, judge whether that copying described first audio frequency frame comprises according to described first frequency-domain analysis data:

When the described first audio frequency frame of described window type index instruction belongs to long window type, described spectrum line Data Placement is gone out described multiple band unit;

Calculate described first energy sum total;

Obtain described second energy sum total;

Calculate described energy sum total poor;

Absolute value in described energy sum total difference is less than one first threshold value being relevant to class simple signal energy sum total difference, described energy sum total difference be greater than be relevant to Auditory Perception after one second threshold value, the described second energy sum total of covering be less than one the 3rd threshold value that is relevant to quiet audio number and at least one condition that described first energy sum total is less than above three conditions of described 3rd threshold value meets time, judge to copy the data belonging to long window type in described first audio frequency frame or described first audio frequency frame.

8. method as claimed in claim 7, it is characterized in that, when the described first audio frequency frame of described window type index instruction belongs to long window type, described spectrum line Data Placement is gone out described multiple band unit to comprise, when the described first audio frequency frame of described window type index instruction belongs to long window type, according to the flatness of spectrum line data, described spectrum line Data Placement is gone out described multiple band unit, and each band unit classifies as a class simple signal classification or a noise like classification.

9. method as claimed in claim 5, it is characterized in that, the method for described adjustment broadcasting speed separately comprises:

In by described adjustment determining program, when judging do not have K × M audio frequency frame to give up in the N number of continuous audio frequency frame of K group, described Auditory Perception decoding device gives up the data at least partially of at least one audio frequency frame after the N number of continuous audio frequency frame of described K group, and wherein K is positive integer; And

In by described adjustment determining program, when judging do not have K × M audio frequency frame to copy in the N number of continuous audio frequency frame of K group, described Auditory Perception decoding device is replicated in the data at least partially of at least one audio frequency frame after the N number of continuous audio frequency frame of described K group.

10. accelerate a method for broadcasting speed, it is characterized in that, the method for described acceleration broadcasting speed includes:

One Auditory Perception decoding device receives a voice data;

Described Auditory Perception decoding device carries out the frequency analysis of one first audio frequency frame of described voice data;

Obtain the one first frequency-domain analysis data about described frequency analysis;

Receive one and accelerate adjustment signal;

The feature of described method is to comprise:

According to described first frequency-domain analysis data, judge whether to give up described first audio frequency frame; And

In described first audio frequency frame be judged as can give up time, described Auditory Perception decoding device according to described accelerate adjustment signal indicated by a broadcasting speed, give up the data at least partially of described first audio frequency frame;

Described first frequency-domain analysis data comprise the window type index being used to refer to the window type that described first audio frequency frame is changed to time domain for the frequent territory used in described Auditory Perception decoding device;

Judge whether that giving up described first audio frequency frame comprises according to described first frequency-domain analysis data: when the described first audio frequency frame of described window type index instruction belongs to long window type, judgement can give up described first audio frequency frame.

11. methods as claimed in claim 10, is characterized in that, described first frequency-domain analysis data separately comprise the spectrum line data of described first audio frequency frame.

12. methods as claimed in claim 11, is characterized in that, judge whether that giving up described first audio frequency frame comprises according to described first frequency-domain analysis data:

When the described first audio frequency frame of described window type index instruction belongs to long window type, described spectrum line Data Placement is gone out multiple band unit;

Calculate one first energy sum total of described multiple band unit;

The one second audio frequency frame obtaining described voice data corresponds to one second energy sum total of described multiple band unit, and described second audio frequency frame is the previous audio frequency frame by the process of described Auditory Perception decoding device of described first audio frequency frame;

Calculate an energy sum total poor, poor=described second energy sum total of described energy sum total-described first energy sum total; And

Absolute value in described energy sum total difference is less than one first threshold value being relevant to class simple signal energy sum total difference, described energy sum total difference be greater than be relevant to Auditory Perception after one second threshold value that covers and described second energy sum total be less than one the 3rd threshold value that is relevant to quiet audio number and at least one condition that described first energy sum total is less than above three conditions of described 3rd threshold value meets time, judge to give up the data belonging to long window type in described first audio frequency frame or described first audio frequency frame.

13. methods as claimed in claim 12, it is characterized in that, when the described first audio frequency frame of described window type index instruction belongs to long window type, described spectrum line Data Placement is gone out described multiple band unit to comprise, when the described first audio frequency frame of described window type index instruction belongs to long window type, according to the flatness of spectrum line data, described spectrum line Data Placement is gone out to classify as class simple signal or the described multiple band unit for noise like.

14. 1 kinds of methods slowing down broadcasting speed, is characterized in that, described in slow down broadcasting speed method include:

One Auditory Perception decoding device receives a voice data;

Described Auditory Perception decoding device carries out the frequency analysis of one first audio frequency frame of described voice data;

Obtain the one first frequency-domain analysis data about described frequency analysis;

Receive one and slowly adjust signal;

The feature of described method is to comprise:

According to described first frequency-domain analysis data, judge whether to copy described first audio frequency frame; And

In described first audio frequency frame be judged as can copy time, described Auditory Perception decoding device according to described slowly adjustment signal make a broadcasting speed of instruction, copy the data at least partially of described first audio frequency frame;

Described first frequency-domain analysis data comprise the window type index being used to refer to the window type that described first audio frequency frame is changed for a frequency domain to the time domain used in described Auditory Perception decoding device;

Judge whether that copying described first audio frequency frame comprises according to described first frequency-domain analysis data: when the described first audio frequency frame of described window type index instruction belongs to long window type, judgement can copy described first audio frequency frame.

15. methods as claimed in claim 14, is characterized in that, described first frequency-domain analysis data separately comprise the spectrum line data of described first audio frequency frame.

16. methods as claimed in claim 15, is characterized in that, judge whether that giving up described first audio frequency frame comprises according to described first frequency-domain analysis data:

When the described first audio frequency frame of described window type index instruction belongs to long window type, described spectrum line Data Placement is gone out multiple band unit;

Calculate one first energy sum total of described multiple band unit;

The one second audio frequency frame obtaining described voice data corresponds to one second energy sum total of described multiple band unit, and described second audio frequency frame is the previous audio frequency frame by the process of described Auditory Perception decoding device of described first audio frequency frame;

Calculate an energy sum total poor, poor=described second energy sum total of described energy sum total-described first energy sum total; And

Absolute value in described energy sum total difference is less than one first threshold value being relevant to class simple signal energy sum total difference, described energy sum total difference be greater than be relevant to Auditory Perception after one second threshold value, the described second energy sum total of covering be less than one the 3rd threshold value that is relevant to quiet audio number and at least one condition that described first energy sum total is less than above three conditions of described 3rd threshold value meets time, judge to copy the data belonging to long window type in described first audio frequency frame or described first audio frequency frame.

17. methods as claimed in claim 16, it is characterized in that, when the described first audio frequency frame of described window type index instruction belongs to long window type, described spectrum line Data Placement is gone out described multiple band unit to comprise, when the described first audio frequency frame of described window type index instruction belongs to long window type, according to the flatness of spectrum line data, described spectrum line Data Placement is gone out to classify as class simple signal or the described multiple band unit for noise like.