Transient speech signal encoding method and device, decoding method and device, processing system and computer-readable storage medium

FIELD OF THE INVENTION

0001The present invention relates to the field of communication technologies, and in particular, to a transient signal encoding method and device, decoding method and device, processing system and computer-readable storage medium.

BACKGROUND OF THE INVENTION

0002Due to the limitation of the number of bit rates for encoding speech signals and the influence of human auditory characteristics, low-frequency band information in a speech signal is always encoded first in an audio encoding algorithm. Compared with narrow-band speech, wide-band speech has the characteristics of being fuller and more natural; therefore, the acoustic quality can be improved by increasing the bandwidth for transmitting speech signals. When the number of bit rates for encoding speech signals is small, a bandwidth extension technology can be adopted to extend the bandwidth range of the speech signals and improve the quality of the speech signals.

0003In recent years, the bandwidth extension technology has developed significantly, and has found commercial applications in several fields, including acoustic enhancement of bass loudspeakers and high frequency enhancement of coded voice and audio.

0004Among current bandwidth extension methods, the encoding technology of low-frequency band information adopts existing encoding and decoding algorithms; and during the process of encoding and decoding high-frequency band information, a small number of bits are generally adopted to encode the high-frequency band information, and the high-frequency band information is recovered at a decoding end by using the correlation between the high-frequency and low-frequency bands.

0005US publication application <patcit id="pcit0001" dnum="US2004196913A1"><text>US2004/196913A1</text></patcit> provides a computationally efficient technique for compression encoding of an audio signal, and further provides a technique to enhance the sound quality of the encoded audio signal. This is accomplished by including more accurate attack detection and a computationally efficient quantization technique. The improved audio coder converts the input audio signal to a digital audio signal. The audio coder then divides the digital audio signal into larger frames having a long-block frame length and partitions each of the frames into multiple short-blocks. The audio coder then computes short-block audio signal characteristics for each of the partitioned short-blocks based on changes in the input audio signal. The audio coder further compares the computed short-block characteristics to a set of threshold values to detect presence of an attack in each of the short-blocks and changes the long-block frame length of one or more short-blocks upon detecting the attack in the respective one or more short-blocks.

0006The article "Pre-echo reduction in the ITU-T G.729.1 coder" presents a new method to address the pre-echo artifact of transform coding at low rates. The input signal is assumed to be coded in two stages: in time domain first and then in transform domain. This is for instance the case in CELP + transform embedded coding. The first stage reconstructs a signal that is usually free of pre-echo. Therefore transform coding can exploit this reconstructed signal at side information for pre-echo detection and reduction. The proposed method is implemented as an adaptive limiter at the decoder side and does not need transmission of auxiliary data. It is part of the recently standardized ITU-T G.729.1 coder, in which it is used in two separated subbands. Experiment test results show that this method has a significant impact on quality in G.729.1 with very small complexity.

0007However, it has become apparent that a transient signal has the following characteristics different from those of a non-transient signal: in the time domain, the signal energy of the transient signal has a large instant change; while in the frequency domain, the frequency spectrum of the transient signal is smooth. In the prior art, the time envelope of the transient signal is not modified, and due to the influence of the processing in the signal encoding process, such as process by frame by frame, time-frequency transform, and frequency envelope, the transient signal is likely to generate a pre-echo; therefore, the prior art has the disadvantage that the effect of the transient signal recovered at the decoding end is not satisfactory.

SUMMARY OF THE INVENTION

0008The present invention is directed to a transient signal encoding method and device, decoding method and device, and processing system for speech signals or audio signals, which are configured to improve the quality of recovery of transient signals.

0009The invention is defined in the independent claims; further embodiments are defined in the dependent claims. According to the transient signal encoding method and device, decoding method and device, and processing system for speech signals or audio signals of the present invention, the time envelope is modified according to characteristics of the transient signal, such that the difference between the amplitude value of the time envelope having the maximal amplitude value and the amplitude values of the time envelopes of the other sub-frames before the sub-frame corresponding to the time envelope having the maximal amplitude value is more distinct, thereby improving the effect of recovery of the transient signal.

BRIEF DESCRIPTION OF THE DRAWINGS

0010To make the technical solutions under the present invention clearer, the accompanying drawings for illustrating embodiments or the prior art are outlined in the following. Apparently, the accompanying drawings are for the exemplary purpose only. <ul id="ul0001" list-style="none" compact="compact"><li><figref idref="f0001">FIG. 1</figref> is a flow chart of a transient signal encoding method according to a first embodiment;</li><li><figref idref="f0002">FIG. 2</figref> is a flow chart of a transient signal encoding method according to a second embodiment;</li><li><figref idref="f0003">FIG. 3</figref> is a block diagram of an embodiment of an encoding end modifying a time envelope of a transient signal according to the second embodiment;</li><li><figref idref="f0004">FIG. 4</figref> is a flow chart of a transient signal decoding method according to a third embodiment;</li><li><figref idref="f0005">FIG. 5</figref> is a flow chart of a transient signal decoding method according to a fourth embodiment;</li><li><figref idref="f0006">FIG. 6</figref> is a block diagram of an embodiment of a decoding end modifying a time envelope of a transient signal according to the fourth embodiment;</li><li><figref idref="f0007">FIG. 7</figref> is a schematic structural view of a transient signal encoding device according to a fifth embodiment;</li><li><figref idref="f0008">FIG. 8</figref> is a schematic structural view of a transient signal encoding device according to a sixth embodiment;</li><li><figref idref="f0009">FIG. 9</figref> is a schematic structural view of a transient signal decoding device according to a seventh embodiment;</li><li><figref idref="f0009">FIG 10</figref> is a schematic structural view of a transient signal decoding device according to an eighth embodiment; and</li><li><figref idref="f0010">FIG 11</figref> is a schematic structural view of a transient signal processing system according to a ninth embodiment.</li></ul>

DETAILED DESCRIPTION OF THE EMBODIMENTS

0011The technical solutions of the present invention are further described in detail with reference to the accompanying drawings and embodiments in the following. The technical solutions according to the embodiments will be clearly and completely described in the following with reference to the accompanying drawings. It is distinct that the embodiments to be described are only a part rather than all of the embodiments of the present invention.

0012<figref idref="f0001">FIG 1</figref> is a flow chart of a transient signal encoding method according to a first embodiment. As shown in <figref idref="f0001">FIG 1</figref>, the method includes the following steps.

0013In Step 11, a sub-frame where a time envelope having a maximal amplitude value (that is, a maximal time envelope) is located is obtained from time envelopes of all sub-frames of an input transient signal, in which the sub-frame is the reference sub-frame described in the embodiments.

0014When the number of bits for encoding an input signal is insufficient, a small number of bits are generally adopted to encode important information of high-frequency band During the process of encoding the high-frequency band information, when the number of available bits is constant, in order to better recover the high-frequency band information, the input signals may be classified, for example, the input signals may be classified into transient signals and non-transient signal, so as to adopt different encoding technologies for different types of signals. This embodiment mainly relates to processing of the transient signals.

0015A method for obtaining a time envelope according to an embodiment includes: dividing an input signal into one or more sub-frames; obtaining energy information of each sub-frame, for example, the energy of each sub-frame and the square root of energy information of each sub-frame, to obtain the energy information; and schematically expressing waveform characteristics or amplitude trends of the input time-domain signal by using the obtained energy information.

0016After the time envelope of the transient signal is obtained, the time envelope may be modified according to the characteristics of the transient signal, such that in the modified time envelope, the difference between the amplitude values of time envelopes of the sub-frames included in the transient signal is more distinct, which is specifically represented in that the difference between the amplitude value of the time envelope having the maximal amplitude value and the amplitude values of other time envelopes is more distinct, so as to highlight the characteristics of the transient signal.

0017In Step 13, an amplitude value of the time envelope of each sub-frame before the reference sub-frame is adjusted in such a way that a first difference is greater than a preset first threshold, in which the first difference is a difference between the amplitude value of the time envelope of each sub-frame before the reference sub-frame and the amplitude value of the maximal time envelope.

0018The first threshold may be determined by the following method: decreasing the amplitude value of the time envelope of each sub-frame before the reference sub-frame to 1/8 to 1/2 of the original amplitude value, obtaining a difference between the adjusted amplitude values of the time envelopes of the sub-frames and the amplitude value of the time envelope of the reference sub-frame, and using the difference as the first threshold.

0019In Step 15, the adjusted time envelope is written into encoding bitstream.

0020Based on the technical solution, the adjustment of the time envelope may further include: calculating an average amplitude value of the time envelopes of each sub-frame after the reference sub-frame; and adjusting the amplitude value of the time envelope of each sub-frame after the reference sub-frame in such a way that a second difference is greater than a preset second threshold when the average amplitude value is lower than or equal to a preset reference value, in which the second difference is a difference between the amplitude value of the time envelope of each sub-frame after the reference sub-frame and the amplitude value of the maximal time envelope.

0021The preset reference value may be selected to be 1/3 to 3/5 of the amplitude value of the time envelope of the reference sub-frame; and the second threshold may be determined by the following method: decreasing the amplitude value of the time envelope of each sub-frame after the reference sub-frame to 1/8 to 1/2 of the original amplitude value, obtaining a difference between the adjusted amplitude values of the time envelopes of the sub-frames and the amplitude value of the time envelope of the reference sub-frame, and using the difference as the second threshold.

0022After the amplitude value of the time envelope of each sub-frame other than the reference sub-frame is adjusted, the adjustment of the time-domain signal in the technical solution may further include: <ul id="ul0002" list-style="none" compact="compact"><li>adjusting the amplitude value of the time envelope of the reference sub-frame in such a way that an average energy of the adjusted time envelope of each sub-frame of the transient signal is greater than a preset third threshold.</li></ul>

0023The third threshold may be selected from the range satisfying the following condition: the average energy of the adjusted time envelope of each sub-frame of the transient signal is equivalent to the average energy of the adjusted time envelope of each sub-frame, for example, the former is 0.8 to 1.2 times the latter.

0024In order to improve the quality of the transient signal recovered at the decoding end, the time envelope corresponding to the transient signal needs to be encoded more finely In this embodiment, during the process of finely encoding the time envelope corresponding to the transient signal, the time envelope of the transient signal can be modified according to the characteristics of the transient signal distinguished from the non-transient signal, such that the difference between the amplitude values of the time envelopes of the sub-frames included by the transient signal is more distinct, thereby improving the quality of the transient signal recovered at the decoding end.

0025In this embodiment, the time envelope of the transient signal is modified according to the characteristics of the transient signal, the difference between the amplitude values of the time envelopes of the sub-frames of the transient signal is enlarged, and the modified time envelope information is sent to the decoding end; and therefore, the position information of the transient signal is encoded and the encoded position information is sent to the decoding end without consuming any number of bits, that is, the technical effect of improving the quality of the transient signal recovered at the decoding end can be realized without increasing the number of bits required by the encoding end.

0026<figref idref="f0002">FIG 2</figref> is a flow chart of a transient signal encoding method according to a second embodiment. As shown in <figref idref="f0002">FIG. 2</figref>, the method includes the following steps.

0027In Step 21, an input signal is decomposed into a low-frequency band signal and a high-frequency band signal; and as for the low-frequency band signal, Step 23 is performed, and as for the high-frequency band signal, Step 25 is performed.

0028In Step 23, parameters of the low-frequency band signal in the input signal are input into a bitstream; and Step 217 is performed.

0029In this embodiment, the parameters of the low-frequency band signal (broadband signal) are input into the bitstream through an encoder.

0030In Step 25, a signal type of the input signal (the high-frequency signal) is determined, and signal type information is input into the bitstream, in which the signal type information is configured to indicate whether the input signal (that is, the signal being currently encoded) is a transient signal or a non-transient signal.

0031In order to more accurately determine the signal type of the input signal, Step 25 may include Steps 2501 to 2509 (not shown).

0032In Step 2501, a long frame is formed with a preset number of consecutive frames in the high-frequency band signal, and an average energy of the long frame is calculated.

0033In a mode for forming the long frame, three consecutive frames including the current frame and two frames before the current frame are combined to form a long frame; or, three consecutive frames including the current frame, a frame before the current frame, and a frame after the current frame are combined to form a long frame; or, several other consecutive frames are combined to form a long frame. The average energy of the long frame may be calculated according to Formula (1): <maths id="math0001" num="(1),"><math display="block"><mrow><mi mathvariant="italic">gain</mi><mo>=</mo><mfrac><mrow><mn>1</mn></mrow><mrow><mn>2</mn></mrow></mfrac><mo>⁢</mo><mi>log</mi><mo>⁢</mo><mn>10</mn><mfenced><mi mathvariant="italic">ener</mi><mo>/</mo><mi>N</mi></mfenced></mrow></math><img file="EP2352145B1_D0001.tif" /></maths> where, <maths id="math0002"><math display="inline"><mrow><mi mathvariant="italic">ener</mi><mo>=</mo><mrow><mstyle displaystyle="true"><mrow><munderover><mo>∑</mo><mrow><mi>i</mi><mo>=</mo><mn>0</mn></mrow><mrow><mi>N</mi></mrow></munderover></mrow></mstyle><mrow><mi>x</mi><mfenced open="[" close="]"><mi>i</mi></mfenced><mo>*</mo><mi>x</mi><mfenced open="[" close="]"><mi>i</mi></mfenced></mrow></mrow><mn>.</mn></mrow></math><img file="EP2352145B1_D0002.tif" /></maths>

0034In Formula (1), <i>gain</i> is the average energy of the long frame; <i>x</i>[<i>i</i>] is a signal value of an ith sampling point of the time-domain signal; and <i>N</i> is the total number of sampling points of the whole long frame.

0035In Step 2503, the long frame is divided into several sub-frames, and an average energy of each sub-frame is calculated.

0036In this embodiment, it is assumed that each frame has a frame length of 5 ms, then the frame length of a long frame is 15 ms; the frame length of a long frame includes 480 sampling points, and if a long frame is divided into 12 sub-frames, the frame length of each sub-frame is 40 sampling points. An average energy <i>sub_gain</i>[<i>i</i>] of each sub-frame is calculated.

0037In Step 2505, a third difference and a fourth difference are calculated respectively, in which the third difference is a maximal difference between the average energy of each sub-frame and the average energy of the long frame, and the third difference is calculated according to Formula (2); and the fourth difference is a maximal difference between average energies of two consecutive sub-frames, and the fourth difference is calculated according to Formula (3). <maths id="math0003" num="(2)"><math display="block"><mrow><mi>max_</mi><mrow><mi mathvariant="italic">deviation</mi></mrow><mo>=</mo><mi>max</mi><mfenced><mi mathvariant="italic">sub_gain</mi><mfenced open="[" close="]"><mi>i</mi></mfenced><mo>,</mo><mi mathvariant="italic">gain</mi></mfenced></mrow></math><img file="EP2352145B1_D0003.tif" /></maths>

0038In Formula (2), <i>sub_gain</i>[<i>i</i>] represents the average energy of each sub-frame, gain represents the average energy of the long frame, and max_<i>deviation</i> represents a maximal difference between the average energy of each sub-frame and the average energy of the long frame, that is, the third difference in the preferred embodiments of the present invention. <maths id="math0004" num="(3)"><math display="block"><mrow><mi>max_</mi><mrow><mi mathvariant="italic">rise</mi></mrow><mo>=</mo><mi>max</mi><mo>⁢</mo><mfenced><mi mathvariant="italic">sub_gain</mi><mfenced open="[" close="]"><mi>i</mi></mfenced><mo>,</mo><mi mathvariant="italic">sub_gain</mi><mo>⁢</mo><mfenced open="[" close="]"><mi>i</mi><mo>+</mo><mn>1</mn></mfenced></mfenced></mrow></math><img file="EP2352145B1_D0004.tif" /></maths>

0039In Formula (3), <i>Sub_gain</i>[<i>i</i>] and <i>sub_gain</i>[<i>i</i> + 1] represent the average energies of two consecutive sub-frames respectively, and max_<i>rise</i> represents a maximal difference between the average energies of two consecutive sub-frames in a long frame, that is, the fourth difference in the preferred embodiments of the present invention.

0040In Step 2507, the average energy of the long frame is compared with a fourth threshold, the third difference is compared with a fifth threshold, and the fourth difference is compared with a sixth threshold, and if the average energy of the long frame is greater than the fourth threshold, the third difference is greater than the fifth threshold, and the fourth difference is greater than the sixth threshold (that is, Formula (4) is satisfied), it is determined that the high-frequency band signal is a transient signal; otherwise, it is determined that the high-frequency band signal is a non-transient signal. <maths id="math0005" num="(4)"><math display="block"><mrow><mi mathvariant="italic">gain</mi><mo>></mo><mi>α</mi><mo>⁢</mo><mn>1</mn><mo>,</mo><mi>max_</mi><mrow><mi mathvariant="italic">deviation</mi></mrow><mo>></mo><mi>α</mi><mo>⁢</mo><mn>2</mn><mo>,</mo><mi>max_</mi><mrow><mi mathvariant="italic">rise</mi></mrow><mo>></mo><mi>α</mi><mo>⁢</mo><mn>3</mn></mrow></math><img file="EP2352145B1_D0005.tif" /></maths>

0041In Formula (4), α1 represents the fourth threshold, α2 represents the fifth threshold; and α3 represents the sixth threshold. The values of α1, α2, and α3 are correlated to the amplitude of the input transient signal, and when the overall amplitude of the transient signal is large, the values of α1, α2, and α3 are large; and when the overall amplitude of the transient signal is small, the values of α1, α2, and α3 are small. For example, when the reference power level of the time envelope of the transient signal is -26 dB, the values of α1, α2, and α3 are in the ranges of 5 &lt; α1 &lt; 10, 2 &lt; α2 &lt; 5, 1 &lt; α3 &lt; 3.

0042In Step 2509, the obtained category information is input into a bitstream, and the category information includes transient signal information and non-transient signal information; and Step 217 is performed. As for a transient signal, Step 27 is performed; and as for a non-transient signal, the time envelope and the frequency-domain envelope of the non-transient signal can be obtained by using a method in the prior art, which will not be repeated herein.

0043It should be noted that, the method for classifying the input signal may be used in combination with the modification of the time envelope according to the present invention; moreover, when the time envelope of each sub-frame of the transient signal is not modified, the method for classifying the input signal may be used in combination with the method for encoding the transient signal in the prior art, and at this time, the accuracy of the identification of the transient signal can also be improved, thereby improving the effect of recovery of the transient signal at the decoding end.

0044In Step 26, the time envelope of each sub-frame of the input signal is calculated respectively, and if the signal type of the input signal is a transient signal, Step 27 is performed; and if the signal type of the input signal is a non-transient signal, Step 29 is performed.

0045In Step 27, the time envelope of the transient signal is modified.

0046In order to highlight the characteristics of the transient signal, the time envelope of the transient signal is modified, for example, modification of increasing the amplitude value is performed on the time envelope having the maximal amplitude value, and/or, modification of decreasing the amplitude value is performed on other time envelopes. Specifically, Step 27 may include Step 2701 to Step 2719. <figref idref="f0003">FIG 3</figref> is a block diagram of an embodiment of an encoding end modifying a time envelope of a transient signal according to the second embodiment. As shown in <figref idref="f0003">FIG 3</figref>, the modification performed on the time envelope of the transient signal includes Step 2701 to Step 2719.

0047In Step 2701, the time envelope of each sub-frame of the transient signal is calculated, so as to obtain the time envelope <i>tEnv</i>[<i>i</i>] of each sub-frame

0048In Step 2703, by searching in the time envelopes of the sub-frames obtained in Step 2701, a sub-frame where the maximal time envelope is located and position information corresponding to the sub-frame are obtained, in which the sub-frame is the reference sub-frame in the embodiments of the present invention, and for the convenience of illustration, the position information of the reference sub-frame is represented as pos in the following.

0049In Step 2705, the position information (i) of the current sub-frame is compared with the position information (pos) of the reference sub-frame, and if the current sub-frame is before the reference sub-frame (that is, i&lt;pos), Step 2707 is performed; otherwise, Step 2709 is performed.

0050In Step 2707, modification of decreasing the amplitude value is performed on the time envelope of the current sub-frame, so as to obtain a first modified envelope, and Step 2719 is performed. The proportion by which the amplitude value is decreased may be determined according to the difference between the amplitude values of the time envelopes corresponding to the sub-frames and the amplitude value of the time envelope corresponding to the reference sub-frame, and if the difference is large, a small proportion by which the amplitude value is decreased may be selected; otherwise, a large proportion by which the amplitude value is decreased may be selected. <figref idref="f0003">FIG 3</figref> shows modification of dividing the value of the time envelope of the current sub-frame by 2, such that the modified time envelope (that is, the first modified envelope) of the current sub-frame is decreased to be 1/2 of the time envelope before modification, that is: <maths id="math0006"><math display="block"><mrow><mrow><mrow><mi mathvariant="italic">tEnv</mi><mfenced open="[" close="]"><mi>i</mi></mfenced><mo>⁢</mo><mi>ʹ</mi></mrow></mrow><mo>=</mo><mfrac><mrow><mn>1</mn></mrow><mrow><mn>2</mn></mrow></mfrac><mo>⁢</mo><mi mathvariant="italic">tEnv</mi><mfenced open="[" close="]"><mi>i</mi></mfenced><mo>,</mo></mrow></math><img file="EP2352145B1_D0006.tif" /></maths> where, <i>tEnv</i>[<i>i</i>]' represents the modified time envelope of the current sub-frame, and <i>tEnv</i>[<i>i</i>] represents the time envelope of the current sub-frame before modification.

0051In Step 2709, an average value <maths id="math0007"><math display="inline"><mrow><msubsup><mrow><mi mathvariant="italic">avrg</mi></mrow><mrow><mi mathvariant="italic">pos</mi><mo>+</mo><mn>1</mn></mrow><mrow><mi>N</mi></mrow></msubsup></mrow></math><img file="EP2352145B1_D0007.tif" /></maths> of the time envelope of each sub-frame after the reference sub-frame is calculated: <maths id="math0008"><math display="block"><mrow><msubsup><mrow><mi mathvariant="italic">avrg</mi></mrow><mrow><mi mathvariant="italic">pos</mi><mo>+</mo><mn>1</mn></mrow><mrow><mi>N</mi></mrow></msubsup><mo>=</mo><mrow><mstyle displaystyle="true"><mrow><munderover><mrow><mo>∑</mo></mrow><mrow><mi>i</mi><mo>=</mo><mi mathvariant="italic">pos</mi><mo>+</mo><mn>1</mn></mrow><mrow><mi>N</mi></mrow></munderover></mrow></mstyle><mrow><mi mathvariant="italic">tEnv</mi><mfenced open="[" close="]"><mi>i</mi></mfenced></mrow></mrow><mo>/</mo><mfenced><mi>N</mi><mo>-</mo><mi mathvariant="italic">pos</mi><mo>-</mo><mn>1</mn></mfenced><mn>.</mn></mrow></math><img file="EP2352145B1_D0008.tif" /></maths>

0052In Step 2711, the average value <maths id="math0009"><math display="inline"><mrow><msubsup><mrow><mi mathvariant="italic">avrg</mi></mrow><mrow><mi mathvariant="italic">pos</mi><mo>+</mo><mn>1</mn></mrow><mrow><mi>N</mi></mrow></msubsup></mrow></math><img file="EP2352145B1_D0009.tif" /></maths> of the time envelope of each sub-frame after the reference sub-frame is compared with a preset reference value, in which the preset reference value in this embodiment is 1/2 of the time envelope corresponding to the reference sub-frame, that is, <maths id="math0010"><math display="inline"><mrow><mfrac><mrow><mn>1</mn></mrow><mrow><mn>2</mn></mrow></mfrac><mo>⁢</mo><mi mathvariant="italic">tEnv</mi><mfenced open="[" close="]"><mi mathvariant="italic">pos</mi></mfenced><mo>,</mo></mrow></math><img file="EP2352145B1_D0010.tif" /></maths> and if <maths id="math0011"><math display="inline"><mrow><msubsup><mrow><mi mathvariant="italic">avrg</mi></mrow><mrow><mi mathvariant="italic">pos</mi><mo>+</mo><mn>1</mn></mrow><mrow><mi>N</mi></mrow></msubsup><mo>≤</mo><mfrac><mrow><mn>1</mn></mrow><mrow><mn>2</mn></mrow></mfrac><mo>⁢</mo><mi mathvariant="italic">tEnv</mi><mfenced open="[" close="]"><mi mathvariant="italic">pos</mi></mfenced><mo>,</mo></mrow></math><img file="EP2352145B1_D0011.tif" /></maths> Step 2713 is performed; otherwise, the time envelope of the current sub-frame is not modified, and Step 2719 is performed. If the difference between the average value of the time envelope of each sub-frame after the reference sub-frame and the preset reference value is large, it indicates that the reference sub-frame corresponding to the maximal time envelope of the original signal is abruptly changed with respect to the sub-frame thereafter; and in order to enable the recovered signal to meet the original characteristics, the sub-frames may be modified. If the difference between the average value of the time envelope of each sub-frame after the reference sub-frame and the preset reference value is small, it indicates that the reference sub-frame corresponding to the maximal time envelope of the original signal is not abruptly changed with respect to the sub-frame thereafter, and at this time, the sub-frames may not be modified. Preferably, the preset reference value is 1/3 to 3/5 of the maximal time envelope of the transient signal.

0053In Step 2713, the position information of the current sub-frame is compared with the position information of the reference sub-frame, so as to determine whether the current sub-frame is the reference sub-frame, and if yes, Step 2715 is performed; otherwise, Step 2717 is performed.

0054In Step 2715, modification of increasing the amplitude value is performed on the time envelope corresponding to the reference sub-frame, so as to obtain a second modified envelope; and Step 2719 is performed.

0055<figref idref="f0003">FIG. 3</figref> shows modification of multiplying the value of the time envelope of the current sub-frame by <maths id="math0012"><math display="inline"><mrow><msqrt><mrow><mn>2</mn></mrow></msqrt><mo>,</mo></mrow></math><img file="EP2352145B1_D0012.tif" /></maths> such that the modified time envelope (that is, the second modified envelope) of the reference sub-frame is increased to be <maths id="math0013"><math display="inline"><mrow><msqrt><mrow><mn>2</mn></mrow></msqrt></mrow></math><img file="EP2352145B1_D0013.tif" /></maths> times the time envelope before modification, that is:<maths id="math0014"><math display="block"><mrow><mi mathvariant="italic">tEnv</mi><mfenced open="[" close="]"><mi mathvariant="italic">pos</mi></mfenced><mo>⁢</mo><mi>ʹ</mi><mo>=</mo><msqrt><mrow><mn>2</mn></mrow></msqrt><mo>⁢</mo><mi mathvariant="italic">tEnv</mi><mfenced open="[" close="]"><mi mathvariant="italic">pos</mi></mfenced><mo>,</mo></mrow></math><img file="EP2352145B1_D0014.tif" /></maths> where, <i>tEnv</i>[<i>pos</i>]' represents the modified time envelope of the reference sub-frame, and <i>tEnv</i>[<i>pos</i>] represents the time envelope of the reference sub-frame before modification.

0056In Step 2717, modification of decreasing the amplitude value is performed on the time envelope of the current sub-frame, so as to obtain a third modified envelope, and Step 2719 is performed. The proportion by which the amplitude value is decreased may be determined according to the difference between the amplitude values of the time envelopes corresponding to the sub-frames and the amplitude value of the time envelope corresponding to the reference sub-frame, and if the difference is large, a small proportion by which the amplitude value is decreased may be selected; otherwise, a large proportion by which the amplitude value is decreased may be selected.

0057<figref idref="f0003">FIG 3</figref> shows modification of dividing the value of the time envelope of the current sub-frame by 2, such that the modified time envelope (that is, the third modified envelope) of the current sub-frame is decreased to be 1/2 of the time envelope before modification, that is: <maths id="math0015"><math display="block"><mrow><mrow><mrow><mi mathvariant="italic">tEnv</mi><mfenced open="[" close="]"><mi>i</mi></mfenced><mo>⁢</mo><mi>ʹ</mi></mrow></mrow><mo>=</mo><mfrac><mrow><mn>1</mn></mrow><mrow><mn>2</mn></mrow></mfrac><mo>⁢</mo><mi mathvariant="italic">tEnv</mi><mfenced open="[" close="]"><mi>i</mi></mfenced><mo>,</mo></mrow></math><img file="EP2352145B1_D0015.tif" /></maths> where, <i>tEnv</i>[<i>i</i>]' represents the modified time envelope of the current sub-frame, and <i>tEnv</i>[<i>i</i>] represents the time envelope of the current sub-frame before modification.

0058In Step 2719, the first modified envelope obtained in Step 2707, the second modified envelope obtained in Step 2715, and the third modified envelope obtained in Step 2717 are combined, to obtain the modified time envelope of the transient signal.

0059Through Step 2701 to Step 2719, the modification of the time envelope of the transient signal is completed, and the modified time envelope of the transient signal is obtained.

0060In Step 211, time-frequency transform is performed on the high-frequency band signal in the input signal, so as to obtain a frequency-domain signal of the high-frequency band signal.

0061When the frequency-domain envelope of the transient signal is obtained, the time-domain signal corresponding to the transient signal is transformed to the frequency domain through a transform method such as fast Fourier transform (FFT) and modified discrete cosine transform (MDCT), so as to obtain the frequency-domain signal corresponding to the transient signal in the frequency domain.

0062No limitation is imposed on the time sequence of Step 211 and Step 25.

0063In Step 213, the frequency-domain envelope of each sub-band of the frequency-domain signal is calculated, so as to obtain the frequency-domain envelope of the high-frequency band signal.

0064The frequency-domain envelope refers to: dividing the frequency-domain signal into one or more sub-bands, obtaining energy information of each sub-band or obtaining the square root of the energy information of each sub-band, and schematically expressing spectral waveform characteristics or amplitude trends of the frequency-domain signal by using the obtained energy information or the obtained square root of the energy information. Therefore, the frequency-domain signal is divided into one or more sub-bands, and the energy information of each sub-band or the square root of the energy information of each sub-band is obtained, and the frequency-domain envelope of each sub-band of the frequency-domain signal is obtained by using the obtained energy information or the obtained square root of the energy information.

0065In Step 215, the obtained frequency-domain envelope of the high-frequency band signal is quantified, and then is added in the bitstream; and Step 217 is performed.

0066In Step 217, the bitstream added with the parameters of the low-frequency band signal, the signal type information of the high-frequency band signal, the frequency-domain envelope and the modified time envelope are sent to the decoding end, in which the signal type information is configured to indicate whether the signal being currently encoded is a transient signal or a non-transient signal, such that the decoding end can determine the type of the decoded current signal according to the signal type information.

0067In this embodiment, when the signal type of the current high-frequency band signal is determined, identification of the transient signal is performed by combining information of several consecutive frames in the high-frequency band signal, and therefore, the accuracy of the identification of the transient signal is improved, and the transient signal can be separated from the input high-frequency band signal more accurately; moreover, in this embodiment, the time envelope corresponding to the separated transient signal is modified, such that the difference between the amplitude values of the time envelopes of the sub-frame of the transient signal is more distinct, thereby improving the quality of the transient signal recovered at the decoding end.

0068<figref idref="f0004">FIG. 4</figref> is a flow chart of a transient signal decoding method according to a third embodiment. As shown in <figref idref="f0004">FIG 4</figref>, the method includes the following steps.

0069In Step 41, a sub-frame where a time envelope having a maximal amplitude value (that is, a maximal time envelope) is located is obtained from time envelopes of all sub-frames of a pre-obtained signal having a signal type of a transient signal, in which the sub-frame is the reference sub-frame described in the embodiments of the present invention.

0070The modification of the time envelope of the transient signal may be performed at the encoding end or the decoding end. In this embodiment, the time envelope is modified according to the characteristics of the transient signal at the decoding end, such that in the modified time envelope, the difference between the amplitude value of the time envelope having the maximal amplitude value of the sub-frames of the transient signal and the amplitude values of other time envelopes is more distinct, so as to highlight the characteristics of the transient signal.

0071In Step 43, an amplitude value of the time envelope of each sub-frame before the reference sub-frame is adjusted in such a way that a first difference is greater than a preset first threshold, in which the first difference is a difference between the amplitude value of the time envelope of each sub-frame before the reference sub-frame and the amplitude value of the maximal time envelope.

0072The first threshold may be determined by the following method: decreasing the amplitude value of the time envelope of each sub-frame before the reference sub-frame to 1/8 to 1/2 of the original amplitude value, obtaining a difference between the adjusted amplitude values of the time envelopes of the sub-frames and the amplitude value of the time envelope of the reference sub-frame, and using the difference as the first threshold.

0073Based on the technical solution, the adjustment of the time envelope may further include: calculating an average amplitude value of the time envelopes of each sub-frame after the reference sub-frame; and adjusting the amplitude value of the time envelope of each sub-frame after the reference sub-frame in such a way that a second difference is greater than a preset second threshold when the average amplitude value is lower than or equal to a preset reference value, in which the second difference is a difference between the amplitude value of the time envelope of each sub-frame after the reference sub-frame and the amplitude value of the maximal time envelope.

0074The preset reference value may be selected to be 1/3 to 3/5 of the amplitude value of the time envelope of the reference sub-frame; and the second threshold may be determined by the following method: decreasing the amplitude value of the time envelope of each sub-frame after the reference sub-frame to 1/8 to 1/2 of the original amplitude value, obtaining a difference between the adjusted amplitude values of the time envelopes of the sub-frames and the amplitude value of the time envelope of the reference sub-frame, and using the difference as the second threshold.

0075After the amplitude value of the time envelope of each sub-frame other than the reference sub-frame is adjusted, the adjustment of the time-domain signal in the technical solution may further include: <ul id="ul0003" list-style="none" compact="compact"><li>adjusting the amplitude value of the time envelope of the reference sub-frame in such a way that an average energy of the adjusted time envelope of each sub-frame of the transient signal is greater than a preset third threshold.</li></ul>

0076The third threshold may be selected from the range satisfying the following condition: the average energy of the adjusted time envelope of each sub-frame of the transient signal is equivalent to the average energy of the adjusted time envelope of each sub-frame, for example, the former is 0.8 to 1.2 times the latter.

0077In Step 45, a pre-obtained time-domain signal is modified according to the adjusted time envelope, so as to obtain a recovered transient signal.

0078The bitstream from the encoding end is decoded, to obtain the frequency-domain envelope of each sub-band of the signal having a signal type of a transient signal. A frequency-domain excitation signal is obtained from normalized low-frequency-band frequency-domain signals or random noises, a frequency-domain signal is generated according to the frequency-domain excitation signal and the frequency-domain envelope, and frequency-time transform is performed on the frequency-domain signal to obtain the time-domain signal. Then, the time-domain signal is modified according to the modified time envelope, such that the transient signal is recovered at the decoding end.

0079In this embodiment, the time envelope of the transient signal is modified at the decoding end, such that in the modified time envelope, the difference between the amplitude value of the time envelope having the maximal amplitude value and the amplitude values of other time envelopes is more distinct, so as to highlight the characteristics of the transient signal, thereby improving the quality of the transient signal recovered at the decoding end.

0080<figref idref="f0005">FIG 5</figref> is a flow chart of a transient signal decoding method according to a fourth embodiment. As shown in <figref idref="f0005">FIG 5</figref>, the method includes the following steps.

0081In Step 51, a bitstream from an encoding end is decoded, to obtain a time envelope and signal type information of a high-frequency band signal, and if the signal type is a transient signal, Step 52 is performed; and if the signal type is a non-transient signal, Step 518 is performed.

0082In Step 52, when the obtained signal type information indicates that the signal type is a transient signal, the time envelope is modified, so as to obtain a modified time envelope; and Step 518 is performed.

0083In order to highlight the characteristics of the transient signal when the current signal type is a transient signal, the time envelope is modified: modification of increasing the amplitude value is performed on the time envelope having the maximal amplitude value, and/or, modification of decreasing the amplitude value is performed on other time envelopes. Step 52 may include Step 5201-Step 5219. <figref idref="f0006">FIG 6</figref> is a block diagram of an embodiment of a decoding end modifying a time envelope of a transient signal according to the fourth embodiment. As shown in <figref idref="f0006">FIG 6</figref>, when the current signal type is a transient signal, the modification performed on the time envelope by the decoding end includes Step 5201 to Step 5219.

0084In Step 5201, the bitstream from the encoding end is decoded, to obtain a time envelope of each sub-frame of the high-frequency band signal and signal type information. If the signal type information indicates that the type of the current signal in the bitstream is a transient signal, Step 5203 is performed, to modify the time envelope; and if the signal type information indicates that the type of the current signal in the bitstream is a non-transient signal, the signal is decoded by using a decoding method in the prior art to recover the non-transient signal, which will not be repeated herein.

0085In Step 5203, by searching in the time envelopes of the sub-frames obtained in Step 5201, a sub-frame where the maximal time envelope is located and position information corresponding to the sub-frame are obtained, in which the sub-frame is the reference sub-frame in the embodiments of the present invention, and for the convenience of illustration, the position information of the reference sub-frame is represented as pos in the following.

0086In Step 5205, the position information (i) of the current sub-frame is compared with the position information (pos) of the reference sub-frame, and if the current sub-frame is before the reference sub-frame (that is, i&lt;pos), Step 5207 is performed; otherwise, Step 5209 is performed.

0087In Step 5207, modification of decreasing the amplitude value is performed on the time envelope of the current sub-frame, so as to obtain a first modified envelope, and Step 5219 is performed.

0088For example, modification of dividing the value of the time envelope of the current sub-frame by 4 is adopted, such that the modified time envelope (that is, the first modified envelope in the embodiments of the present invention) of the current sub-frame is decreased to be 1/4 of the time envelope before modification, that is: <maths id="math0016"><math display="block"><mrow><mrow><mrow><mi mathvariant="italic">tEnv</mi><mfenced open="[" close="]"><mi>i</mi></mfenced><mo>⁢</mo><mi>ʹ</mi></mrow></mrow><mo>=</mo><mfrac><mrow><mn>1</mn></mrow><mrow><mn>4</mn></mrow></mfrac><mo>⁢</mo><mi mathvariant="italic">tEnv</mi><mfenced open="[" close="]"><mi>i</mi></mfenced><mo>,</mo></mrow></math><img file="EP2352145B1_D0016.tif" /></maths> where, <i>tEnv</i>[<i>i</i>]' represents the modified time envelope of the current sub-frame, and <i>tEnv</i>[<i>i</i>] represents the time envelope of the current sub-frame before modification.

0089In Step 5209, an average value <maths id="math0017"><math display="inline"><mrow><msubsup><mrow><mi mathvariant="italic">avrg</mi></mrow><mrow><mi mathvariant="italic">pos</mi><mo>+</mo><mn>1</mn></mrow><mrow><mi>N</mi></mrow></msubsup></mrow></math><img file="EP2352145B1_D0017.tif" /></maths> of the time envelope of each sub-frame after the reference sub-frame is calculated: <maths id="math0018"><math display="block"><mrow><msubsup><mrow><mi mathvariant="italic">avrg</mi></mrow><mrow><mi mathvariant="italic">pos</mi><mo>+</mo><mn>1</mn></mrow><mrow><mi>N</mi></mrow></msubsup><mo>=</mo><mrow><mstyle displaystyle="true"><mrow><munderover><mo>∑</mo><mrow><mi>i</mi><mo>=</mo><mi mathvariant="italic">pos</mi><mo>+</mo><mn>1</mn></mrow><mrow><mi>N</mi></mrow></munderover></mrow></mstyle><mrow><mi mathvariant="italic">tEnv</mi><mfenced open="[" close="]"><mi>i</mi></mfenced></mrow></mrow><mo>/</mo><mfenced><mi>N</mi><mo>-</mo><mi mathvariant="italic">pos</mi><mo>-</mo><mn>1</mn></mfenced><mn>.</mn></mrow></math><img file="EP2352145B1_D0018.tif" /></maths>

0090In Step 5211, the average value <maths id="math0019"><math display="inline"><mrow><msubsup><mrow><mi mathvariant="italic">avrg</mi></mrow><mrow><mi mathvariant="italic">pos</mi><mo>+</mo><mn>1</mn></mrow><mrow><mi>N</mi></mrow></msubsup></mrow></math><img file="EP2352145B1_D0019.tif" /></maths> of the time envelope of each sub-frame after the reference sub-frame is compared with a preset reference value, in which the preset reference value in this embodiment is 1/4 of the time envelope corresponding to the reference sub-frame, that is, <maths id="math0020"><math display="inline"><mrow><mfrac><mrow><mn>1</mn></mrow><mrow><mn>4</mn></mrow></mfrac><mo>⁢</mo><mi mathvariant="italic">tEnv</mi><mfenced open="[" close="]"><mi mathvariant="italic">pos</mi></mfenced><mo>,</mo></mrow></math><img file="EP2352145B1_D0020.tif" /></maths> and if <maths id="math0021"><math display="inline"><mrow><msubsup><mrow><mi mathvariant="italic">avrg</mi></mrow><mrow><mi mathvariant="italic">pos</mi><mo>+</mo><mn>1</mn></mrow><mrow><mi>N</mi></mrow></msubsup><mo>≤</mo><mfrac><mrow><mn>3</mn></mrow><mrow><mn>5</mn></mrow></mfrac><mo>⁢</mo><mi mathvariant="italic">tEnv</mi><mfenced open="[" close="]"><mi mathvariant="italic">pos</mi></mfenced><mo>,</mo></mrow></math><img file="EP2352145B1_D0021.tif" /></maths> Step 5213 is performed; otherwise, the time envelope of the current sub-frame is not modified, and Step 5219 is performed.

0091In Step 5213, the position information of the current sub-frame is compared with the position information of the reference sub-frame, so as to determine whether the current sub-frame is the reference sub-frame, and if yes, Step 5215 is performed; otherwise, Step 5217 is performed.

0092In Step 5215, modification of increasing the amplitude value is performed on the time envelope corresponding to the reference sub-frame, so as to obtain a second modified envelope; and Step 5219 is performed.

0093For example, modification of multiplying the value of the time envelope of the current sub-frame by <maths id="math0022"><math display="inline"><mrow><msqrt><mrow><mn>3</mn></mrow></msqrt></mrow></math><img file="EP2352145B1_D0022.tif" /></maths> is performed, such that the modified time envelope (that is, the second modified envelope) of the reference sub-frame is increased to be <maths id="math0023"><math display="inline"><mrow><msqrt><mrow><mn>3</mn></mrow></msqrt></mrow></math><img file="EP2352145B1_D0023.tif" /></maths> times the time envelope before modification, that is: <maths id="math0024"><math display="block"><mrow><mrow><mrow><mi mathvariant="italic">tEnv</mi><mfenced open="[" close="]"><mi mathvariant="italic">pos</mi></mfenced><mo>⁢</mo><mi>ʹ</mi></mrow></mrow><mo>=</mo><msqrt><mrow><mn>3</mn></mrow></msqrt><mo>⁢</mo><mi mathvariant="italic">tEnv</mi><mfenced open="[" close="]"><mi mathvariant="italic">pos</mi></mfenced><mo>,</mo></mrow></math><img file="EP2352145B1_D0024.tif" /></maths> where, <i>tEnv</i>[<i>pos</i>]' represents the modified time envelope of the reference sub-frame, and <i>tEnv</i>[<i>pos</i>] represents the time envelope of the reference sub-frame before modification.

0094In Step 5217, modification of decreasing the amplitude value is performed on the time envelope of the current sub-frame, so as to obtain a third modified envelope, and Step 5219 is performed.

0095For example, modification of dividing the value of the time envelope of the current sub-frame by 4 is adopted, such that the modified time envelope (that is, the third modified envelope) of the current sub-frame is decreased to be 1/4 of the time envelope before modification, that is: <maths id="math0025"><math display="block"><mrow><mrow><mrow><mi mathvariant="italic">tEnv</mi><mfenced open="[" close="]"><mi>i</mi></mfenced><mo>⁢</mo><mi>ʹ</mi></mrow></mrow><mo>=</mo><mfrac><mrow><mn>1</mn></mrow><mrow><mn>4</mn></mrow></mfrac><mo>⁢</mo><mi mathvariant="italic">tEnv</mi><mfenced open="[" close="]"><mi>i</mi></mfenced><mo>,</mo></mrow></math><img file="EP2352145B1_D0025.tif" /></maths> where, <i>tEnv</i>[<i>i</i>]' represents the modified time envelope of the current sub-frame, and <i>tEnv[i]</i> represents the time envelope of the current sub-frame before modification.

0096In Step 5219, the first modified envelope obtained in Step 5207, the second modified envelope obtained in Step 5215, the third modified envelope obtained in Step 5217, and the time envelope that does not meet the modification conditions in Step 5211 and is not subjected to time-domain modification are combined, to obtain the modified time envelope of the transient signal.

0097Through Step 5201 to Step 5219, the modification of the time envelope of the transient signal is completed, and the modified time envelope of the transient signal is obtained.

0098In Step 53, the bitstream from the encoding end is decoded, to obtain the low-frequency band signal; and Step 519 is performed.

0099In this embodiment, the low-frequency band signal in the bitstream is decoded by a decoder.

0100No limitation is imposed on the time sequence of Step 51 and Step 53.

0101In Step 55, a frequency-domain excitation signal of the high-frequency band signal is generated.

0102The frequency-domain excitation signal of the high-frequency band signal is obtained from normalized low-frequency-band frequency-domain signals or random noises.

0103In Step 57, the bitstream from the encoding end is decoded, to obtain the frequency-domain envelope of each sub-band of the high-frequency band signal.

0104No limitation is imposed on the time sequence of Step 55 and Step 57.

0105In Step 59, the frequency-domain excitation signal is modified by using the frequency-domain envelope of each sub-band of the high-frequency band signal.

0106The objective of the modification is to enable the energy of the recovered frequency spectrum to be equivalent to the energy of the real high-frequency band spectrum.

0107In Step 511, a high-frequency-band frequency-domain signal is generated according to the modified frequency-domain excitation signal; and the high-frequency-band frequency-domain signal may be calculated according to Formula (5): <maths id="math0026" num="(5)."><math display="block"><mrow><mi mathvariant="italic">spectrum</mi><mfenced open="[" close="]"><mi>i</mi></mfenced><mo>=</mo><mi mathvariant="italic">exc</mi><mfenced open="[" close="]"><mi>i</mi></mfenced><mo>*</mo><mi mathvariant="italic">fEnv</mi><mfenced open="[" close="]"><mi>j</mi></mfenced></mrow></math><img file="EP2352145B1_D0026.tif" /></maths>

0108In Formula (5), <i>exc</i>[<i>i</i>] represents the frequency-domain excitation signal; <i>fEnv</i>[<i>j</i>] represents the frequency-domain envelope; and <i>spectrum</i>[<i>i</i>] represents the high-frequency-band frequency-domain signal.

0109In Step 513, frequency-time transform is performed on the generated high-frequency-band frequency-domain signal.

0110In Step 515, the time-domain signal is generated. If the type of the high-frequency band signal is a transient signal, Step 516 is performed, and if the high-frequency band signal is a non-transient signal, Step 517 is performed.

0111In Step 516, the time-domain signal having a signal type of a transient signal is adjusted, to obtain the adjusted time-domain signal <i>signal</i>'[<i>i</i>]<i>.</i>

0112A preset number of sampling points in the reference sub-frame are selected; and signal amplitude of each of the selected sampling points is adjusted in such a way that a fifth difference is greater than a seventh threshold, in which the fifth difference is a difference between the signal amplitude value of each of the selected sampling points and a maximal amplitude value of the reference sub-frame.

0113The seventh threshold may be selected from the following range: decreasing the amplitudes of the selected sampling points to be 1/2 of the original amplitudes, and obtaining the differences between the adjusted amplitudes of the sampling points and the maximal amplitude among the amplitudes of the sampling points included in the reference sub-frame.

0114A preset number of sampling points included in the sub-frame where the time envelope having the maximal amplitude value is located are selected, and the signal amplitudes of the sampling points are decreased, so as to adjust the time-domain signal. The specific method for adjustment of the time-domain signal and the preset number of sampling points required to be adjusted are mainly dependent upon the characteristics of the original input signal.

0115For example, during the adjustment of the time-domain signal, a preset number of sampling points included in the sub-frame where the time envelope having the maximal amplitude value is located are selected sequentially, for example, the sampling points in the first 1/4 sub-frame length included in the time-domain signal corresponding to the reference sub-frame where the time envelope having the maximal amplitude value is located are selected, and the amplitude values of the selected sampling points are divided by 2. If a number of bits can be used for transmitting flag information at the encoding end and the decoding end, these bit positions can be used to carry the flag information to the decoding end, for example, when the encoding end has a bit for transmitting the flag information, the decoding end can determine whether to adjust the preset number of sampling points according to the flag bit; when the encoding end has multiple bit positions for carrying the flag information, the decoding end can determine which sampling points need to be adjusted according to the received flag bits; and when the encoding end has sufficient bit positions for carrying the flag information, the decoding end can determine whether each sampling point needs to be adjusted according to the received flag information.

0116It should be noted that, the method for adjusting the time-domain signal may be used in combination with the modification of the time envelope according to the present invention; moreover, when the time envelope of each sub-frame of the transient signal is not modified, the method for adjusting the time-domain signal may be used in combination with the method for encoding the transient signal in the prior art, and at this time, the characteristics of the transient signal can also be highlighted, thereby improving the effect of recovery of the transient signal.

0117In Step 517, the obtained time-domain signal <i>signal'</i>[<i>i</i>] is normalized.

0118In Step 518, by using the modified time envelope obtained in Step 52, the normalized time-domain signal having a signal type of a transient signal is modified, so as to obtain a recovered transient signal; and by using the time envelope signal having a signal type of non-transient signal obtained in Step 51, the corresponding time-domain signal is modified, so as to obtain a recovered non-transient signal.

0119The normalized time-domain signal having a signal type of a transient signal may be modified according to Formula (6): <maths id="math0027" num="(6)."><math display="block"><mrow><mi mathvariant="italic">signal</mi><mfenced open="[" close="]"><mi>i</mi></mfenced><mo>=</mo><mi mathvariant="italic">signalʹ</mi><mfenced open="[" close="]"><mi>i</mi></mfenced><mo>*</mo><mi mathvariant="italic">tEnv</mi><mfenced open="[" close="]"><mi>j</mi></mfenced><mo>/</mo><mi mathvariant="italic">tEnv</mi><mfenced open="[" close="]"><mi>j</mi></mfenced><mo>⁢</mo><mi>ʹ</mi></mrow></math><img file="EP2352145B1_D0027.tif" /></maths>

0120In Formula (6), <i>signal'</i>[<i>i</i>] represents the modified time-domain signal; <i>tEnv</i>[<i>j</i>] represents the modified time envelope; <i>tEnv</i>[<i>j</i>]' represents the time envelope of the modified time-domain signal (<i>signal</i>'[<i>i</i>]); and <i>signal</i>[<i>i</i>] represents the time-domain signal of the high-frequency band signal.

0121In Step 519, the recovered low-frequency band signal and high-frequency band signal are combined, to obtain the output wide-frequency band signal, in which the recovered high-frequency band signal includes the recovered transient signal and the recovered non-transient signal.

0122In this embodiment, no limitation is imposed on the time sequence of Step 51, Step 57, and Step 53.

0123In this embodiment, the time envelope corresponding to the transient signal in the high-frequency band signal obtained through decoding at the decoding end is modified, such that the difference between the amplitude values of the time envelopes of all sub-frames corresponding to the transient signal is more distinct, thereby improving the quality of the transient signal recovered at the decoding end; moreover, in this embodiment, before the time-domain signal is modified by using the time envelope, the amplitudes of the sampling points before the time-domain signal of the sub-frame having the maximal time envelope are decreased, so as to highlight the characteristics of the transient signal, thereby significantly improving the output effect of the transient signal in the output signal.

0124<figref idref="f0007">FIG. 7</figref> is a schematic structural view of a transient signal encoding device according to a fifth embodiment. As shown in <figref idref="f0007">FIG. 7</figref>, the transient signal encoding device of this embodiment includes: a reference sub-frame obtaining module 71, a first amplitude value adjusting module 72, and a bitstream writing module 73.

0125The reference sub-frame obtaining module 71 is configured to obtain a reference sub-frame where a time envelope having a maximal amplitude value (that is, a maximal time envelope) is located from time envelopes of all sub-frames of an input transient signal.

0126The first amplitude value adjusting module 72 is configured to adjust an amplitude value of the time envelope of each sub-frame before the reference sub-frame in such a way that a first difference is greater than a preset first threshold, in which the first difference is a difference between the amplitude value of the time envelope of each sub-frame before the reference sub-frame and the amplitude value of the maximal time envelope. The first threshold may be determined by the following method: decreasing the amplitude value of the time envelope of each sub-frame before the reference sub-frame to 1/8 to 1/2 of the original amplitude value, obtaining a difference between the adjusted amplitude values of the time envelopes of the sub-frames and the amplitude value of the time envelope of the reference sub-frame, and using the difference as the first threshold.

0127The bitstream writing module 73 is configured to write the adjusted time envelope into bitstream.

0128Based on the technical solution, the transient signal encoding device of this embodiment further includes: an average amplitude value calculation module 74, a second amplitude value adjusting module 75, and a third amplitude value adjusting module 76.

0129The average amplitude value calculation module 74 is configured to calculate an average amplitude value of the time envelopes of each sub-frame after the reference sub-frame.

0130The second amplitude value adjusting module 75 is configured to adjust the amplitude value of the time envelope of each sub-frame after the reference sub-frame in such a way that a second difference is greater than a preset second threshold when the average amplitude value is lower than or equal to a preset reference value, in which the second difference is a difference between the amplitude value of the time envelope of each sub-frame after the reference sub-frame and the amplitude value of the maximal time envelope. The preset reference value may be selected to be 1/3 to 3/5 of the amplitude value of the time envelope of the reference sub-frame; and the second threshold may be determined by the following method: decreasing the amplitude value of the time envelope of each sub-frame after the reference sub-frame to 1/8 to 1/2 of the original amplitude value, obtaining a difference between the adjusted amplitude values of the time envelopes of the sub-frames and the amplitude value of the time envelope of the reference sub-frame, and using the difference as the second threshold.

0131The third amplitude value adjusting module 76 is configured to adjust an amplitude value of the time envelope of the reference sub-frame in such a way that an average energy of the adjusted time envelope of each sub-frame of the transient signal is greater than a preset third threshold, after the amplitude value of the time envelope of each sub-frame other than the reference sub-frame is adjusted. The third threshold may be selected from the range satisfying the following condition: the average energy of the adjusted time envelope of each sub-frame of the transient signal is equivalent to the average energy of the adjusted time envelope of each sub-frame, for example, the former is 0.8 to 1.2 times the latter.

0132In this embodiment, during the process of finely encoding the time envelope corresponding to the transient signal, the first amplitude value adjusting module can modify the time envelope of the transient signal according to the characteristics of the transient signal, such that the difference between the amplitude values of the time envelopes of the sub-frames included by the transient signal is more distinct, thereby improving the quality of the transient signal recovered at the decoding end.

0133<figref idref="f0008">FIG. 8</figref> is a schematic structural view of a transient signal encoding device according to a sixth embodiment. Different from the embodiment in <figref idref="f0007">FIG 7</figref>, the transient signal encoding device of this embodiment further includes a signal type determination module 77.

0134The signal type determination module 77 is configured to determine a signal type of the input signal, and write signal type information in the encoding bitstream, in which the signal type includes a transient signal or a non-transient signal.

0135The signal type determination module 77 may include a long frame average energy calculation unit 771, a sub-frame average energy calculation unit 772, a difference calculation unit 773, and a signal type determination unit 774.

0136The long frame average energy calculation unit 771 is configured to form a long frame with a preset number of consecutive frames in the input signal and calculate an average energy of the long frame.

0137The sub-frame average energy calculation unit 772 is configured to divide the long frame into multiple sub-frames and calculate an average energy of each sub-frame.

0138The difference calculation unit 773 is configured to calculate a third difference and a fourth difference respectively, in which the third difference is a maximal difference between the average energy of each sub-frame and the average energy of the long frame, and the fourth difference is a maximal difference between average energies of two consecutive sub-frames.

0139The signal type determination unit 774 is configured to determine that the input signal is a transient signal when the average energy of the long frame is greater than a fourth threshold, the third difference is greater than a fifth threshold, and the fourth difference is greater than a sixth threshold; otherwise, determine that the input signal is a non-transient signal.

0140In this embodiment, when the signal type of the current high-frequency band signal is determined, identification of the transient signal is performed by combining information of several consecutive frames in the high-frequency band signal, and therefore, the accuracy of the identification of the transient signal is improved, and the transient signal can be separated from the input high-frequency band signal more accurately; moreover, in this embodiment, the time envelope corresponding to the separated transient signal is modified, such that the difference between the amplitude values of the time envelopes of the sub-frame of the transient signal is more distinct, thereby improving the quality of the transient signal recovered at the decoding end.

0141<figref idref="f0009">FIG. 9</figref> is a schematic structural view of a transient signal decoding device according to a seventh embodiment. As shown in <figref idref="f0009">FIG. 9</figref>, the transient signal encoding device of this embodiment includes: a reference sub-frame obtaining module 91, a first amplitude value adjusting module 92, and a time-domain signal modification module 93.

0142The reference sub-frame obtaining module 91 is configured to obtain a reference sub-frame where a time envelope having a maximal amplitude value (that is, a maximal time envelope) is located from time envelopes of all sub-frames of a pre-obtained signal having a signal type of a transient signal.

0143The first amplitude value adjusting module 92 is configured to adjust an amplitude value of the time envelope of each sub-frame before the reference sub-frame in such a way that a first difference is greater than a preset first threshold, in which the first difference is a difference between the amplitude value of the time envelope of each sub-frame before the reference sub-frame and the amplitude value of the maximal time envelope.

0144The time-domain signal modification module 93 is configured to modify a pre-obtained time-domain signal according to the adjusted time envelope, so as to obtain a recovered transient signal.

0145In this embodiment, the time envelope of the transient signal is modified by the time envelope modification module at the decoding end, such that in the modified time envelope, the difference between the amplitude value of the time envelope having the maximal amplitude value and the amplitude values of other time envelopes is more distinct, so as to highlight the characteristics of the transient signal, thereby improving the quality of the transient signal recovered at the decoding end.

0146<figref idref="f0009">FIG. 10</figref> is a schematic structural view of a transient signal decoding device according to an eighth embodiment. Different from the embodiment in <figref idref="f0009">FIG. 9</figref>, the transient signal decoding device of this embodiment further includes: an average amplitude value calculation module 94, a second amplitude value adjusting module 95, and a third amplitude value adjusting module 96.

0147The average amplitude value calculation module 94 is configured to calculate an average amplitude value of the time envelopes of each sub-frame after the reference sub-frame.

0148The second amplitude value adjusting module 95 is configured to adjust the amplitude value of the time envelope of each sub-frame after the reference sub-frame in such a way that a second difference is greater than a preset second threshold when the average amplitude value is lower than or equal to a preset reference value, in which the second difference is a difference between the amplitude value of the time envelope of each sub-frame after the reference sub-frame and the amplitude value of the maximal time envelope.

0149The third amplitude value adjusting module 96 is configured to adjust an amplitude value of the time envelope of the reference sub-frame in such a way that an average energy of the adjusted time envelope of each sub-frame of the transient signal is greater than a preset third threshold, after the amplitude value of the time envelope of each sub-frame other than the reference sub-frame is adjusted.

0150Based on the technical solution, the transient signal decoding device of this embodiment may further include a time-domain signal adjusting module 97.

0151The time-domain signal adjusting module 97 is configured to select a preset number of sampling points in the reference sub-frame, and adjust signal amplitude of each of the selected sampling points in such a way that a fifth difference is greater than a seventh threshold, in which the fifth difference is a difference between the signal amplitude value of each of the selected sampling points and a maximal amplitude value of the reference sub-frame.

0152In this embodiment, the time envelope corresponding to the transient signal in the high-frequency band signal obtained through decoding at the decoding end is modified, such that the difference between the amplitude values of the time envelopes of all sub-frames corresponding to the transient signal is more distinct, thereby improving the quality of the transient signal recovered at the decoding end; moreover, in this embodiment, before the time-domain signal is modified by using the time envelope, the amplitudes of the sampling points before the time-domain signal of the sub-frame having the maximal time envelope are decreased, so as to highlight the characteristics of the transient signal, thereby significantly improving the output effect of the transient signal in the output signal.

0153<figref idref="f0010">FIG. 11</figref> is a schematic structural view of a transient signal processing system according to a ninth embodiment. As shown in <figref idref="f0010">FIG 11</figref>, the transient signal processing system includes a transient signal encoding device 111 and a transient signal decoding device 112.

0154The modification of the time envelope of the transient signal may be performed at the encoding end.

0155The transient signal encoding device 111 is configured to obtain a reference sub-frame where a time envelope having a maximal amplitude value (that is, a maximal time envelope) is located from time envelopes of all sub-frames of an input transient signal, adjust an amplitude value of the time envelope of each sub-frame before the reference sub-frame in such a way that a first difference is greater than a preset first threshold, and write the adjusted time envelope into bitstream, in which the first difference is a difference between the amplitude value of the time envelope of each sub-frame before the reference sub-frame and the amplitude value of the maximal time envelope.

0156The transient signal decoding device 112 is configured to modify a pre-obtained time-domain signal according to the time envelope in the received bitstream, so as to obtain a recovered transient signal.

0157In this embodiment, the time envelope of the transient signal is modified at the encoding end, and the difference between the amplitude value of the time envelope having the maximal amplitude value among the time envelopes of all sub-frames of the transient signal and the amplitude values of other time envelopes is enlarged, so as to highlight the characteristics of the transient signal, thereby improving the quality of the transient signal recovered at the decoding end.

0158In the transient signal processing system of this embodiment, as for the specific detailed structure of the transient signal encoding device 111, reference can be made to the description of the embodiments in <figref idref="f0007">FIGs. 7</figref> and <figref idref="f0008">8</figref>, and as for the specific principle of the modification of the time envelope of the transient signal, reference can be made to the description of the embodiments in <figref idref="f0001 f0002 f0003">FIGs. 1 to 3</figref>, which will not be repeated herein.

0159Alternatively, the modification of the time envelope of the transient signal may be performed at the decoding end.

0160The transient signal encoding device 111 is configured to write a time envelope of each sub-frame of a transient signal in a bitstream.

0161The transient signal decoding device 112 is configured to obtain a reference sub-frame where a maximal time envelope having a maximal amplitude value is located from time envelopes of all sub-frames of a signal in the received bitstream, adjust an amplitude value of the time envelope of each sub-frame before the reference sub-frame in such a way that a first difference is greater than a preset first threshold, and modify a pre-obtained time-domain signal according to the adjusted time envelope to obtain a recovered transient signal, in which the first difference is a difference between the amplitude value of the time envelope of each sub-frame before the reference sub-frame and the amplitude value of the maximal time envelope.

0162In this embodiment, the time envelope of the transient signal is modified at the decoding end, and the difference between the amplitude value of the time envelope having the maximal amplitude value among the time envelopes of all sub-frames of the transient signal and the amplitude values of other time envelopes is enlarged, so as to highlight the characteristics of the transient signal, thereby improving the quality of the transient signal recovered at the decoding end.

0163In the transient signal processing system of this embodiment, as for the specific detailed structure of the transient signal decoding device 112, reference can be made to the description of the embodiments in <figref idref="f0009">FIGs. 9 and 10</figref>, and as for the specific principle of the modification of the time envelope of the transient signal, reference can be made to the description of the embodiments in <figref idref="f0004 f0005 f0006">FIGs. 4 to 6</figref>, which will not be repeated herein.

0164It should be understood by persons of ordinary skill in the art that the accompanying drawings are merely schematic views of preferred embodiments, and modules or processes in the accompanying drawings are not indispensable for implementing the present invention.

0165It should be understood by persons of ordinary skill in the art that, modules in a device according to an embodiment may be distributed in the device of the embodiment according to the description of the embodiment, or correspondingly disposed in one or more devices different from this embodiment. The modules of the above embodiment may be combined into one module, or further divided into multiple sub-modules.

0166The sequence numbers of the above embodiments are merely for the convenience of description, and do not imply the preference among the embodiments.

0167Persons of ordinary skill in the art should understand that all or a part of the steps of the methods according to the embodiments may be implemented by a program instructing relevant hardware. The program may be stored in a computer readable storage medium. When the program is run, the steps of the methods according to the embodiments are performed. The storage medium may be any medium that is capable of storing program codes, such as a ROM, a RAM, a magnetic disk, and an optical disk.

0168Finally, it should be noted that the above embodiments are merely provided for describing the technical solutions of the present invention, but not intended to limit the present invention, which is defined by the appended claims.