JP4736331B2 - Acoustic signal playback device - Google Patents

Description

  The present invention relates to a sound signal reproducing apparatus suitable for producing a homepage with BGM (background music) on the Internet.

  Conventionally, various methods are used for compression of an acoustic signal. As a method for compressing and encoding an acoustic signal, MP3 (MPEG-1 / Layer3), AAC (MPEG-2 / Layer3), and the like have been put into practical use. Such a compression encoding method makes it possible to handle an acoustic signal as small data, and contributes to the efficiency of data recording and transmission.

Recently, not only lossy encoding methods such as MP3 and AAC as described above, but also lossless encoding methods that can be completely restored have been developed and are used for management of acoustic materials (for example, Patent Document 1).
JP 2000-82199 Gazette

  However, in the conventional lossless encoding method as described above, the condition parameter is not changed midway, and the same condition parameter encoding is performed throughout. Therefore, there is a problem that the condition parameter corresponding to the section cannot be set. Further, in the conventional method, since the acoustic signal to be encoded must be read at once, it is difficult to encode a large volume of acoustic data.

  For this reason, the present applicant has developed a time-series signal encoding device and reproducing device that can change the encoding condition parameter for each predetermined section and can encode even a large amount of data. is doing.

  However, such a playback device can be realized with a stand-alone terminal device, but cannot be played back on the terminal device side via a network such as the Internet.

  An object of the present invention has been made in view of the above-described problems, and an object of the present invention is to provide a reproducing apparatus capable of reproducing an acoustic signal via a network.

According to a first aspect of the present invention for achieving the above object, in a system in which a server and a terminal device are connected via a network, the server includes a plurality of music materials composed of acoustic signals compressed by encoding. a generating means for generating an object tags having information plurality select the music material to be played, the object tag, comprising means for writing a predetermined position of the HTML document, and said terminal apparatus, said HTML When the document is accessed, in accordance with the description of the object tag , the compressed block is read for each of the compressed audio signals corresponding to the plurality of selected music materials , the compressed block is decoded, and the decoded plural And a means for reproducing the uncompressed block by synthesizing the waveform . The server or the terminal device can set the matrix.

  According to the present invention, when a terminal device accesses a server via a network, a predetermined acoustic signal is reproduced.

(0.1 Outline of Embodiment)
Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the drawings. FIG. 1 is a diagram showing a system in which a sound signal reproducing apparatus according to this embodiment is incorporated. A web server 103, a terminal device 105, a terminal device 107, and the like are connected to the network 101.

Web server 103, the terminal device 105 has a manufacturer 10 9 homepage. The web server 103 provides a home page created by the producer 109 via the network 101 such as the Internet. The terminal device 105 is a computer or the like necessary for producing this homepage. The terminal device 107 is a computer or the like owned by a general user.

  FIG. 2 is a block diagram showing the configuration of the terminal device 105. The terminal device 105 includes a control unit 111 such as a central processing unit, a storage unit 113 such as an HDD, a media input / output unit 115 that reads and writes media such as a flexible disk and a CD-ROM, and a modem connected to the network 101. And a communication unit 117 such as a LAN board, an input unit 119 such as a keyboard, a mouse or a microphone, a printing unit 121 such as a printer, a display unit 123 such as a CRT or a liquid crystal display, and an audio output unit 125 such as a speaker. These are connected by a bus 125.

  FIG. 3 is a block diagram showing the configuration of the web server 103. The web server 103 includes a control unit 131 such as a central processing unit, a storage unit 133 such as an HDD, a media input / output unit 135 that reads and writes media such as a flexible disk and a CD-ROM, and a modem connected to the network 101. And a communication unit 137 such as a LAN board, an input unit 139 such as a keyboard, a mouse or a microphone, a printing unit 141 such as a printer, and a display unit 143 such as a CRT or a liquid crystal display device. It is connected.

The configuration of the terminal device 107 is the same as the configuration of the terminal device 105 shown in FIG.
Next, the processing operation of terminal apparatus 105 in the present embodiment will be described. FIG. 4 is a flowchart showing processing of the terminal device 105. A program for executing the processing shown in FIG. 4 is plugged in the terminal device 105 and the terminal device 107 in advance. When the producer 109 produces a homepage or the like, an HTML document is created. In the present embodiment, by incorporating a program for reproducing an acoustic signal in the HTML document, when a general user terminal device 107 or the like accesses a home page, character information and image information are displayed on the screen. The sound signal attached to this screen is reproduced like BGM.

  The process shown in FIG. 4 is a flowchart showing the process of the terminal device 105 for creating a homepage incorporating this acoustic signal.

In response to an instruction from the producer 109, the control unit 111 of the terminal device 105 opens a player screen (step S401). That is, the control unit 111 performs a predetermined display on the display unit 123 of the terminal device 105 according to the plugged-in program. When the producer gives a predetermined instruction on this screen, the control unit 111 causes the display unit 123 to display the player screen. The player screen is a screen for selecting music material.
FIG. 5 is a diagram showing a player screen 151 displayed on a part of the display unit 123 of the terminal device 105. On the player screen 151, a track type display unit 153, a reproduction non-reproduction display unit 155, a volume designation unit 157, and a matrix 159 are displayed.

The track type display area 153 displays five track names. That is, track 1, track 2,..., Track 5 are displayed.
The reproduction / non-reproduction display unit 155 distinguishes between the case where the track is reproduced and the case where the track is not reproduced.
The volume designation unit 157 designates the volume of each track with a numerical value from “0” to “100”.

  The matrix 159 is a 5 × 5 matrix, and includes 25 “11”, “12”,..., “21”, “22”,..., “53”, “54”, “55”. It consists of elements, one element represents one music material, and each material is different.

In accordance with the instruction from the producer 109, the control unit 111 sets the matrix 159 (step S402). That manufacturer 10 9 using a mouse or the like to the matrix 159 that is displayed, selecting one of the materials for each track, the control unit 111, for example, the case shown in FIG. 5, "12" , “22”, “ 31”, “43”, and “51” are recognized to be selected.

Next, when the producer 109 presses an OK button 161, the control unit 111 automatically generates an object tag specifying the player of the plugged-in program and copies it to the clipboard on the memory (step S403). .
FIG. 6 shows an HTML 16 having an object tag.
<object calssid = ”player” ... </ object> is the contents of the object tag.
calssid is a serial number indicating a plugged-in program as a player.

<param name = “autoPlay” value = ”On” turns on / off the designation of automatic reproduction, and in this case, on is designated.
<param name = “Directory” value = “AAAAAA” specifies the material placement location, and AAAAAA is a URL indicating the placement location.
<param name = ”File” value = ”ois” specifies acoustic material identification information.
<param name = ”Extention” value = ”enc” specifies lossless compression and non-compression. Lossless is specified for enc, and non- compression is specified for wav.

<param name = “Sector” value = “2,2,1,3,1” indicates the default value of the selected music material. As shown in FIG. Column 2 is specified, column 2 is specified for track 2, column 1 is specified for track 3, column 3 is specified for track 4, Since the “1” column is designated for the track 5, “2, 2, 1, 3, 1” is set as a default value.
By performing such designation, it becomes possible to arrange music.
<param name = “MatrixSize” value = “5” indicates the number of music materials and indicates the number of tracks to be synthesized.

As shown in FIG. 6, the object tag may be sandwiched between <div> tags. The <div> tag is written by the producer 109.
The <div> tag indicates the position of the upper left corner of the player bar indicating the playback device on the home page. In the case of FIG. 6, the player bar is set to be displayed at a position of 40 pixels from the upper end and 10 pixels from the left end.

FIG. 7 is a diagram showing a screen 171 of the terminal device 105 on which the object tag shown in FIG. 6 is displayed. An area 173 for displaying the object tag and the like, a copy button 175, and a file save button 177 are displayed. The copy button 175 is a button for copying the object tag in the area 173 to a predetermined position in the HTML document. The file save button 177 saves the object tag displayed in the area 173 at an arbitrary location by dragging and dropping. When arrangement is determined before creating the HTML document, the object tag can be saved as a file in an arbitrary folder.

Further, the producer 109 creates an HTML document 181 with a text editor using the terminal device 105 (step S404). The HTML document 181 constitutes a home page, and FIG. 8 is a diagram showing the HTML document 181.
When the copy button 175 shown in FIG. 7 is pressed, the control unit 111 can paste the object tag in the HTML document 181 described above (step S405).

FIG. 9 is a diagram showing an HTML document 181 in which an object tag is pasted. As shown in FIG. 9, HTML 162 having an object tag is pasted into the HTML document 181, and as shown in FIG. 10, a home page is displayed on the display unit 123 of the terminal device 105. The player bar 185 is displayed such that the upper left end of the player bar 185 is positioned at 40 pixels from the upper end of the home page screen and 10 pixels from the left end. That is, the player bar 185 is displayed at a position of 40 pixels from the upper end and 10 pixels from the left end indicated by the <div> tag.
The producer 109 stores the HTML document 181 obtained through the above processing in the web server 103 (step 406).

  When accessing the web server 3 from the terminal device 107 of a general user and browsing the home page, a home page screen as shown in FIG. 10 is displayed. When the music playback switch 187 of the player bar 185 is clicked, the object tag 162 designates it. The sound signal is synthesized and reproduced.

  That is, since the URL of the material arrangement location is described in the object tag 162, the material is selected in the server having the URL according to the default value of the arrangement. The selected material is sent to the terminal device 107 and synthesized by a plug-in program.

  That is, a general user can listen to an acoustic signal at the same time as viewing a homepage screen. This sound signal is preset by the producer 109, but when the setting switch 189 of the player bar 183 is clicked from the terminal device 107, the player screen 159 is displayed, and the setting of the matrix 159 is newly set on the terminal device 107 side. To perform another arrangement. Note that the process of reproducing the acoustic signal according to the matrix setting will be described later.

  Further, as shown in FIGS. 11 and 12, an object tag 191 and a metafile 193 may be provided separately. FIG. 11 shows an object tag 191, and a metafile 193 is designated in the object tag.

FIG. 12 shows the metafile 193. In the metafile 193, automatic reproduction designation on / off, material arrangement location designation (URL), and the like as shown in FIG. 6 are written.

  As described above, according to the present embodiment, it is possible to embed a sound signal reproducing device in a website (HTML). This playback apparatus can play back high-definition audio signals of CD (compact disc) or higher.

The general user can reproduce the sound signal by accessing the server via the Internet. Conversely, from the server side, the music file can be distributed to the general user via the Internet. it can.
The reproduced music can be arranged by the producer on the server side, and the user can also specify the arrangement.

As shown in FIGS. 11 and 12, when a metafile is created, a copy button 175 and a play list save button 178 are displayed on the screen 171 as shown in FIG.
When the play list save button 178 is clicked, the object tag 191 and the metafile 193 are saved.

  Also, in the player screen 151 shown in FIG. 5, as shown in FIG. 14, a file instruction unit 160 is provided beside the matrix 159, and instead of setting the matrix 159, the file designation unit 160 uses the music material (reproduction file). May be set.

FIG. 14 shows a case in which, for example, the track 3 is not designated by the matrix 159 but is designated by the file designation unit 160. That is, when the button 162-3 of the file specifying unit 160 is clicked, a dialog 164 is displayed, and the music material (playback file) displayed in the dialog 164 is selected.
Note that the music material (reproduction file) may be set by the file designation unit 160 not only for the track 3 but also for other tracks.

Next, encoding and reproduction of an acoustic signal according to the present embodiment will be described in detail. The encoded data that is the target of this playback apparatus is obtained by compressing and encoding the original acoustic signal by the acoustic signal encoding apparatus shown below.
This encoding device is realized by the terminal device 105.

(1.1 Configuration of encoding device)
FIG. 15 is a configuration diagram of an audio signal encoding device. In FIG. 15, 10 is a block dividing means, 20 is a low-order fixed bit deleting means, 30 is an inter-channel computing means, 40 is a sample sequence rearranging means, 50 is a signal flat part processing means, 60 is a correlation frame detecting means, and 70 is A prediction error conversion means, 80 is a polarity processing means, 90 is a variable length coding means, and 100 is a code output means.

  In FIG. 15, the block dividing means 10 has a function of dividing a digital acoustic signal, which is a sample string obtained by sampling, into a predetermined number of samples and reading it into a work memory. The lower fixed bit deletion means 20 has a function of deleting a predetermined number of lower bits considered to have been added to match the number of bits when a plurality of acoustic signals are synthesized. The inter-channel calculation means 30 has a function of performing a correlation calculation between each channel of a sample string composed of a plurality of channels. The sample sequence rearrangement means 40 has a function of separating the sample sequence constituting the block into a main sample sequence which is a sample sequence obtained based on recording and a sub sample sequence obtained by interpolating the main sample sequence. Have. The signal flat part processing means 50 has a function of detecting a flat part having a constant signal value and efficiently encoding the sample sequence for each channel.

  The correlation frame detection means 60 sets a predetermined section as a frame for each sample sequence, detects a correlation frame in which all corresponding sample values are the same between frames, and is backward in time ( A function of deleting a correlation frame located in the future). The prediction error conversion means 70 has a function of converting the value of each sample into a prediction error value using a linear prediction error method. The polarity processing means 80 has a function of performing processing for dividing a bit string of each sample in which a positive / negative value is represented by a complement expression into one bit representing a positive / negative polarity and another bit string. The variable length encoding means 90 has a function of encoding each sample value with a variable bit length. The code output means 100 has a function of outputting the data encoded in units of each block and the data obtained by the above means to a single encoded file while maintaining the divided block form. The apparatus shown in FIG. 1 is actually realized by a computer and a dedicated software program installed in the computer.

(1.2 Processing operation of encoding device)
Next, the processing operation of the acoustic signal encoding apparatus shown in FIG. 15 will be described. First, the block dividing means 10 blocks in units of a predetermined number of samples from the head of the sample sequence constituting the acoustic signal, and reads it into the work memory for each block. The number of samples for one block can be set in advance. The number of samples for one block is preferably slightly longer than that of the first chapter of music, that is, about 5 to 10 seconds. Even if the length is the same in time, the number of samples varies depending on the sampling frequency, so the setter sets in advance in this system a number of samples that is about 5 seconds to 10 seconds. For example, if the sampling frequency of the acoustic signal to be encoded is 48 kHz, 480000 samples may be set as one block to make one block 10 seconds. The block dividing means 10 sequentially reads the set number of samples as one block into the work memory.

  Next, the lower fixed bit deletion means 20 separates a predetermined number of lower bits of each sample of the sample sequence read as one block. This is performed in order to remove redundant lower-order bit components when data having a quantization bit number of 16 bits is converted to 24 bits in order to match a high-definition audio signal. If this process is not performed, the amount of encoded information will increase 3/2 times. In addition, even when the acoustic signal of the underlying material is quantized with high-definition 24-bit data, redundant lower-order bit components are generated only in specific blocks due to the performance of the A / D converter and editing processing. In some cases, the process of detecting and deleting redundant lower fixed bits in units of blocks by the lower fixed bit deleting means 20 may be effective. If the low-order bits are not fixed but significant data, the low-order fixed bits can be separated instead of being deleted, and the separated low-order bit data array can be recorded separately as part of the output code data. In this case, the processing load after the prediction error conversion means 70 in the subsequent stage is reduced. The lower fixed bit deletion means 20 can be set in advance as to whether or not to operate.

Subsequently, the inter-channel calculation means 30 performs a correlation calculation process between the channels. Specifically, first, the difference calculation between the sample of channel ch1 and the sample of channel ch2 at the same time is performed, and the difference value is recorded as a new sample value of channel ch2. That is, the sample values _{x L} of the original channel ch1, the sample values of the original channel ch2 and _{x R,} and a new value of the sample _{x ch2} = x L -x x _ch2 calculated in _R is the channel ch2 Become. However, if the sum of the absolute values of x _ch2 is calculated over one block and it is larger than the sum of the absolute values of the sample values of the original channel ch2, the sample value of the channel ch2 is changed. Absent. This is because the present invention is intended for data compression, and therefore it does not make sense to increase the amount of data. When the channel ch2 is updated, a new value of the channel ch1 sample is calculated. Specifically, x _ch1 = x _R + x _ch2 / 2 is calculated using x _ch2 which is the new sample value of the calculated channel ch2. The calculated x _ch1 is recorded as a new sample value of the channel ch1. The new sample value x _ch1 is mathematically (x _L + x _R ) / 2, but if this is done, an error will occur in the calculation by the computer. Therefore, after calculating x _ch2 once, x _ch1 Is calculated.

From the sample string of each channel obtained in this way, at the time of decoding, it can be restored as follows. _First, to restore the sample values of the original channel ch2 as _{x R = x ch1 -x ch2 /} 2. Subsequently, using the sample values _{x R} channel ch2 restored _by x L _{= x} R + x _ch2 restores the sample values of the original channel ch1.

  When the processing by the inter-channel operation means 30 is finished, the sample row rearrangement means 40 performs sample row rearrangement processing. The processing by the sample row rearrangement means 40 is effective when handling work data in which a plurality of acoustic signals are mixed as acoustic signals. Specifically, a normal acoustic signal having a sampling frequency of 48 kHz and a quantization bit number of 16 bits and a high-definition acoustic signal having a sampling frequency of 96 kHz and a quantization bit number of 24 bits are mixed. The acoustic signal obtained by mixing acoustic signals having different sampling frequencies in this way is handled by unifying the sampling frequency into a high-definition acoustic signal. In this case, an acoustic signal with a sampling frequency of 48 kHz is interpolated with an average value of adjacent samples to match the number of samples with the acoustic signal with a sampling frequency of 96 kHz.

  Such an acoustic signal is schematically shown in FIG. In FIG. 16A, the numbers in parentheses are sample numbers assigned in ascending order from 1, and x indicates the value of the sample. For such a sample sequence, the sample sequence rearrangement means 40 performs four types of processing. The first is to calculate the difference between the odd-numbered samples and the average value of the even-numbered samples on both sides. Second, the difference between the even-numbered samples and the average value of the odd-numbered samples on both sides is calculated. The third is to calculate the difference between the odd-numbered sample and the previous even-numbered sample. Fourth, the difference between the even-numbered sample and the immediately preceding odd-numbered sample is calculated. The sample rows after each calculation are schematically shown in FIGS. 16B to 16E, respectively. In the examples of FIGS. 16B and 16D, even-numbered samples that are not subjected to calculation are moved in the past, and in the examples of FIGS. 16C and 16E, odd-numbered samples are moved in the past. It is shown in the state.

  As a result of this difference calculation, the sample having the smallest difference value is the sub sample sequence, and the sample that is not calculated in this case is the main sample sequence. In the example of FIG. 16, the values in the latter half of the arrays in FIGS. 16B to 16E are compared. For example, when the odd-numbered sample is obtained by interpolation using both adjacent samples, the latter half of the array shown in FIG. When the even-numbered sample is obtained by interpolation using both adjacent samples, the latter half of the array shown in FIG. When the odd-numbered sample is interpolated with the same value as the immediately preceding sample, the latter half of the array shown in FIG. When the even-numbered sample is interpolated with the same value as the immediately preceding sample, the latter half of the array shown in FIG. For example, when there are many values near 0 in the second half of the array shown in FIG. 16B, the set of even-numbered samples is separated as a main sample column, and the set of odd-numbered samples is separated as a sub-sample column.

  Further, in the processing of the sample row rearranging means 40, as shown in FIGS. 16B to 16E, the main sample is moved in the past in time and the sub sample is moved in the future in time. However, the main sample and the sub sample may be separated and handled. For example, when the odd number is a sub sample, the main sample and the sub sample are separated as shown in FIG. In this embodiment, in order to prevent the amount of data from increasing due to performing linear prediction on the sample sequence including the sample obtained by interpolation using the original sample, the main sample is used. Distinguish columns from subsample columns. Therefore, if linear prediction can be performed separately for the main sample sequence and the sub sample sequence, even if one sample sequence as shown in FIGS. Two sample rows as shown in FIG.

  Next, the signal flat portion processing unit 50 performs signal flat portion processing on the sample sequence. The signal flat portion refers to a portion where the same signal level continues. In particular, it often appears in a silent portion where the signal level is “0” and a saturated portion where the absolute value of the signal level is maximum. The silence part occurs when the sound is actually silent or when the sound is not recorded very low, but the saturation part occurs during the process of recording the signal and A / D conversion. Regardless of whether the same signal level is continuous in the silent part, the saturated part, or otherwise, the signal flat part continuously records the same signal level for a predetermined time (a predetermined number of samples). For this reason, this portion is easily compressed data. Specifically, three values of the start time position of the signal flat portion, the number of samples with the same signal level, and the signal level (sample value) are separated from the sample sequence of each channel as signal flat portion data and recorded. To do. The signal flat portion is deleted from the sample sequence of each channel. This is schematically shown in FIGS. 17 (a) and 17 (b). FIG. 17A shows a sample sequence before the signal flat portion processing. In FIG. 17A, the shaded portion indicates a flat signal portion. By the processing of the signal flat part processing means 50, the signal flat part is separated from the original sample sequence, as shown in FIG. However, the separated signal flat portion is recorded in a format as shown in FIG. 17C in order to restore it to the original state at the time of decoding.

  As described above, the signal flat portion data is recorded for each signal flat portion with the three attributes of the start time (sample number), the number of samples, and the sample value. Here, the head time is the time from the start position of the signal, and in the example of FIG. 17C, it is recorded with the sample number from the head. As described above, when the sample number is divided by the sampling frequency, it is converted into time. The number of samples is information indicating how long the sample value continues. Note that the end time of the signal flat portion may be recorded instead of the number of samples. The sample value indicates the digitized signal level. Here, since quantization is performed with 16 bits, the maximum value is “32767” and the minimum value is “−32768”. That is, “0” indicates a silent portion, and “32767” and “−32768” indicate a saturated portion. However, the signal flat portion is not unconditionally processed. This is because the purpose is to compress the data, and it is meaningless if the signal flat portion data becomes larger than the reduction amount of the sample string. Therefore, the signal flat portion data is generated and separated from the sample sequence of each channel only when a predetermined number or more of samples serving as the signal flat portion are continuous.

Subsequently, the correlation frame detection means 60 sets a frame having a predetermined section length for the sample string of each channel, and compares the set frames. In this embodiment, the frame length is fixed over the entire interval from the start time to the end time of the sample string. Specifically, one frame is 512 samples. Correlation frame detection means 60 sets 512 samples from the head of the sample sequence of each channel as one frame, and obtains a correlation frame in which all samples match between frames. Specific procedures will be described with reference to the flowchart of FIG 8.

  First, the correlation frame detection means 60 performs framing in units of a predetermined number of samples (step S1). In this embodiment, the frame length is fixed length 512 samples in any block. As shown in FIG. 19A, the correlation frame detecting means 60 sets 512 samples from the head of the sample sequence as one frame in each block.

  Next, a search is made for a frame in which all sample values constituting each frame match. Specifically, as shown in FIG. 19B, first, among the set frames, the last frame in time in the block is set as a target frame for searching for a correlation frame. Next, within a predetermined search range, a sample having the same value as the value of the first sample of the target frame is searched while going back in time (step S2). For example, as shown in FIG. 20A, it is assumed that the target frame is composed of 512 samples mT to mT + 511. In this case, first, a sample that is the same as the sample value x (mT) of the first sample mT of the target frame is searched. Further, the search is performed in order of the sample mT-1 and the sample mT-2. In FIG. 6, m indicates the m-th frame from the beginning, and T indicates the frame length (512 samples in this embodiment).

  If a matching sample t is found (step S3), it is then compared whether the next sample t + 1 of the sample t matches the second sample mT + 1 of the target frame. In this way, as long as the sample values match, the subsequent samples are compared (step S4). In step S4, the process is repeated as long as the values of x (t + p) and x (mT + p) match. For example, in the example shown in FIG. 20B, since x (t) to x (t + 8) coincide with x (mT) to x (mT + 8), the process of step S4 is continued with p = 9. Will be. If all x (t + p) and x (mT + p) from p = 0 to p = 511 match (step S5), the sample string is set as a correlation frame for the target frame, and the first sample number of the correlation frame and the target frame Is recorded as frame correlation data in association with the first sample number, and the target frame is deleted from the original sample sequence (step S6). If it does not match all the samples of the target frame, the search is further made in time to determine whether there is a sample whose value matches the first sample of the target frame. If there is no matching correlation frame even after a predetermined number of samples, the search for the correlation frame related to the target frame is stopped, and the correlation frame search is performed using the frame immediately before the target frame as a new target frame. When the processing for one target frame is completed, the process returns to step S2, and the processing is continued using the immediately previous frame as a new target frame (step S7). In this way, correlation frame detection processing is performed using all frames except for a frame located in the vicinity of the first sample in the block as target frames.

  Looking at the entire sample sequence in the block, if a correlation frame corresponding to the target frame is detected as shown in FIG. 19 (c), the target frame is deleted as shown in FIG. 19 (d). . At this time, frame correlation data as shown in FIG. 19E is recorded so that it can be completely restored at the time of decoding. As shown in FIG. 19 (e), in the frame correlation data, the head sample number of the target frame and the head sample number of the correlation frame are recorded in association with each other.

  Subsequently, the prediction error conversion unit 70 converts the value of each sample of the sample sequence (the main sample sequence and the sub sample sequence when processing by the sample sequence rearrangement unit 40 is performed) into a prediction error value. The calculation of the prediction error value in a certain sample is performed using the values of one or a plurality of samples immediately before being located in the past in time. In the present embodiment, a method of dynamically changing the number of samples immediately before use is used. Hereinafter, such adaptive linear predictive coding will be described. An outline of the process of adaptive linear prediction encoding performed by the prediction error conversion means 70 is shown in the flowchart of FIG. First, a linear prediction error corresponding to each prediction calculation formula is calculated using a plurality of prediction calculation formulas prepared in advance (step S11). Specifically, the following [Formula 1] to [Formula 11] are prepared as prediction calculation formulas for calculating the prediction error of the sample number t.

[Formula 1]
e0 (t) = x (t) -e0 (t-1) / 2

[Formula 2]
_{e1 (t) = x (t} ) -a 11 · x (t-1) -e1 (t-1) / 2

[Formula 3]
_{e2 (t) = x (t} ) -a 21 · x (t-1) -a 22 · x (t-2) -e2 (t-1) / 2

[Formula 4]
_{e3 (t) = x (t} ) -a 31 · x (t-1) -a 32 · x (t-2) -a 33 · x (t-3) -e3 (t-1) / 2

[Formula 5]
_{e4 (t) = x (t} ) -a 41 · x (t-1) -a 42 · x (t-2) -a 43 · x (t-3) -a 44 · x (t-4) - e4 (t-1) / 2

[Formula 6]
_{e5 (t) = x (t} ) -a 51 · x (t-1) -a 52 · x (t-2) -a 53 · x (t-3) -a 54 · x (t-4) - a ₅₅ · x (t−5) −e5 (t−1) / 2

[Formula 7]
_{e6 (t) = x (t} ) -b 11 · x (t-1) -e6 (t-1) / 2

[Formula 8]
_{e7 (t) = x (t} ) -b 21 · x (t-1) -b 22 · x (t-2) -e7 (t-1) / 2

[Formula 9]
_{e8 (t) = x (t} ) -b 31 · x (t-1) -b 32 · x (t-2) -b 33 · x (t-3) -e8 (t-1) / 2

[Formula 10]
_{e9 (t) = x (t} ) -b 41 · x (t-1) -b 42 · x (t-2) -b 43 · x (t-3) -b 44 · x (t-4) - e9 (t-1) / 2

[Formula 11]
_{e10 (t) = x (t} ) -b 51 · x (t-1) -b 52 · x (t-2) -b 53 · x (t-3) -b 54 · x (t-4) - b ₅₅ · x (t−5) −e10 (t−1) / 2

  In the above [Equation 1] to [Equation 11], e0 (t) to e10 (t) are prediction errors in the sample at time t according to each prediction calculation formula, and x (t) to x (t-5) are times. It is a sample value in t-t-5.

“A ₂₁ · x (t−1) + a ₂₂ · x (t−2)” in [Expression 3], and “a ₃₁ · x (t−1) + a ₃₂ · x (t−) in [Expression 4]. 2) + a ₃₃ · x (t−3) ”,“ a ₄₁ · x (t−1) + a ₄₂ · x (t−2) + a ₄₃ · x (t−3) + a ₄₄ · x (t−4) ”,“ a ₅₁ · x (t−1) + a ₅₂ · x (t−2) + a ₅₃ · x (t−3) + a ₅₄ · x (t−4) ”in [Formula 6] ) + A ₅₅ · x (t−5) ”,“ b ₂₁ · x (t−1) + b ₂₂ · x (t−2) ”in the above [Formula 8],“ b ₃₁ · x in the above [Formula 9] (T−1) + b ₃₂ · x (t−2) + b ₃₃ · x (t−3) ”,“ b ₄₁ · x (t−1) + b ₄₂ · x (t−2) ”in [Expression 10] above. + B ₄₃・x (t−3) + b ₄₄ · x (t−4) ”,“ b ₅₁ · x (t−1) + b ₅₂ · x (t−2) + b ₅₃ · x (t−3) ” ) + B ₅₄ · x (t−4) + b ₅₅ · x (t−5) ”is a linear prediction component based on the past 2 to 5 samples. Using this linear prediction component and the prediction errors “e1 (t−1) / 2” to “e10 (t−1) / 2” (error feedback component) calculated in the immediately preceding sample, a prediction error at time t e0 (t) to e10 (t) are calculated.

As an initial value in the above coefficient _{a 11 ~a 55, a 11 =} 1, a 21 = 2, a 22 = -1, a 31 = 3, a 32 = -3, a 33 = 1, a 41 = 4 , A ₄₂ = −6, a ₄₃ = 4, a ₄₄ = −1, a ₅₁ = 5, a ₅₂ = −10, a ₅₃ = 10, a ₅₄ = −5, and a ₅₅ = 1. The initial values of the coefficients b _{11 to} b ₅₅ are as follows: b ₁₁ = 1, b ₂₁ = 2, b ₂₂ = −1, b ₃₁ = 3, b ₃₂ = −3, b ₃₃ = 1, b _{41 = 4, b 42 = -6} , b 43 = 4, b 44 = -1, b 51 = 5, b 52 = -10, b 53 = 10, b 54 = -5, b 55 = 1 value of Each is set. In the present embodiment, these coefficients are dynamically changed according to the set mode. FIG. 22 shows a linear coefficient setting mode that can be set in this system. In FIG. 22, “initial fixed value” indicates that the initial value is used as it is for all samples in the block. “Initial optimum value calculation” indicates that an optimum value is calculated through the entire sample sequence in the block, and the calculated value is used for all samples in the block. The “user setting initial fixed value” indicates that a value uniquely set by the user is used for all samples in the block. “Sequential optimum value calculation” indicates that the coefficient is updated in units of a predetermined number of samples using the initial value. In the present embodiment, using the mode 2, the coefficient of the a _ij sequence is set to “initial fixed value”, and the coefficient of the b _ij sequence is set to “sequential optimum value calculation”. Here, “sequential optimum value calculation” will be described. Specifically, the “sequential optimum value calculation” determines the coefficients b _{11 to} b ₅₅ using the following [Equation 12] using the Levinson-Durvin algorithm.

[Formula 12]
φ (k) = 1 / (NK) · Σ _{j = 1, NK} x (j) · x (j + k)
k _i = − {φ (i) + Σ _{j = 1, i−1} b _j (i−1) · φ (i−j)} / E (i−1)
b _i (i) = k _i
b _j (i) = b _j (i−1) + k _i · b _i−j (i−1) where 1 ≦ j ≦ i−1
E (i) = (1−k _i ² ) E (i−1)

In the above [Equation 12], φ (k) is a sample shifted by k samples within the range of the maximum value K (5 in the above example) in N samples x (j) (j = 1,..., N). The autocorrelation value with the column. Note that N is a sufficiently large value with respect to K (for example, when K = 5, N = 32768). [Formula 12] is recursively repeated from i = 1 to i = K, and finally obtained b _j (K) is a coefficient corresponding to the past K samples, and is obtained in each phase. The intermediate result b _j (i) is the coefficient b _ij . In step S1, using the coefficients determined by the above [Equation 12], calculation is performed using the equations [Equation 7] to [Equation 11]. The calculation according to [Equation 12] is actually performed in step S17 described later. In addition, since the values for several past samples are required to determine the coefficient, the first N-1 samples should be calculated using the above initial coefficients [Formula 7] to [Formula 11]. become.

  Returning to the flowchart of FIG. 21, the linear prediction error that minimizes the cumulative error, which is the cumulative absolute value of the prediction error value for each prediction calculation formula, is selected as the prediction error of the sample (step S12). Here, the concept of cumulative error is used. Specifically, the cumulative values of the past samples of the prediction error calculated by each prediction calculation formula [Formula 1] to [Formula 11] are set as A0 to A10. Then, a prediction error corresponding to the smallest one of the accumulated errors A0 to A10 is selected. For example, it is assumed that A2 is the smallest among A0 to A10. In this case, the prediction error e2 (t) calculated by [Formula 3] is selected as the prediction error e (t) to be encoded. The selected prediction error e (t) is replaced with the original value x (t) of the sample and the subsequent processing is performed.

  Subsequently, the absolute values of the prediction errors e0 (t) to e10 (t) are added to the accumulated errors A0 to A10 (step S13). Specifically, as shown in the following [Equation 13], the variables A0 to A10 that are accumulated error values are updated. At the same time, every time processing of each sample is performed, processing is performed in which the counters C1 and C2 are added one by one.

[Formula 13]
A0 ← A0 + | e0 (t) | A1 ← A1 + | e1 (t) |
A2 ← A2 + | e2 (t) | A3 ← A3 + | e3 (t) |
A4 ← A4 + | e4 (t) | A5 ← A5 + | e5 (t) |
A6 ← A6 + | e6 (t) | A7 ← A7 + | e7 (t) |
A8 ← A8 + | e8 (t) | A9 ← A9 + | e9 (t) |
A10 ← A10 + | e10 (t) |

  Subsequently, it is determined whether the counter C1 has exceeded a predetermined number of times (step S14). In this embodiment, this predetermined number is set as 100 times. That is, it is determined whether or not the counter C1 exceeds 100.

  As a result, if the counter exceeds 100, the accumulated error is halved (step S15). Specifically, as shown in [Formula 14] below, variables A0 to A10 that are accumulated errors are divided by two. At the same time, the counter C1 is reset to zero. That is, A0 to A10 here are not cumulative errors in a pure sense, but are moving averages of cumulative errors. In the present embodiment, the maximum 100 samples immediately before are accumulated, but the previous ones are processed in half. Thereby, the influence of the sample separated in time is made small.

[Formula 14]
A0 ← (A0) / 2 A1 ← (A1) / 2
A2 ← (A2) / 2 A3 ← (A3) / 2
A4 ← (A4) / 2 A5 ← (A5) / 2
A6 ← (A6) / 2 A7 ← (A7) / 2
A8 ← (A8) / 2 A9 ← (A9) / 2
A10 ← (A10) / 2

  Subsequently, it is determined whether or not the counter C2 has exceeded a predetermined number (step S16). In the present embodiment, this predetermined number is set to 32768. That is, it is determined whether or not the counter C2 exceeds 32768.

As a result, the counter C2 is If exceeds 32768, performs recalculation of the coefficients _b 11 _{~b 55} (step S17). Specifically, the coefficients b _{11 to} b ₅₅ are recalculated using the above [Equation 12]. At the same time, the counter C2 is reset to zero.

  By executing the processing of step S11 to step S17 over all samples in the sample sequence in the block, the values of all samples are replaced with the target error e (t) from the original amplitude value x (t). Become. In the present embodiment, in particular, it is possible to calculate a prediction error with higher accuracy by dynamically changing the coefficients of a plurality of prediction formulas.

  Subsequently, the polarity processing means 80 performs positive / negative polarity processing of each sample in the block. The value of each sample is replaced with the prediction error from the amplitude value by the prediction error conversion means 70, but the bit format of each sample remains the same. Normally, when being calculated by a computer such as a computer, each data is processed in units of 32 bits and expressed using a two's complement expression. This is converted into a positive / negative signed absolute value expression, and the absolute value portion is moved one bit higher, and the positive / negative sign bit is moved to LSB (least significant bit). FIG. 23 schematically shows how the bit processing unit 80 converts the bit configuration. FIG. 23A shows a bit configuration before processing, and FIG. 23B shows a bit configuration after processing. The reason why the positive and negative code bits are moved to the LSB in this way is to make it easier to detect the bit length of each sample in the later processing of the variable length encoding means 90.

  Next, the variable length coding means 90 performs processing for converting each sample into a variable length. The variable length coding in this embodiment employs a method generally called Golomb coding. Specifically, the bit component constituting one sample is divided into an upper bit component and a lower bit component, the lower bit component is left unchanged, and the upper bit component is a numerical value obtained by decimal conversion of only the upper bit. It is an array in which bit “0” is arranged and separator bit “1” is added at the end. For example, consider an 8-bit bit component “00101000”. At this time, if the lower bit component is 4 bits, the lower bit component is “1000”. Since the upper bits are “0010”, “2” pieces of “0” obtained by decimal conversion are arranged and converted to “001” by adding “1” at the end. As a result, the 8-bit bit string “00101000” is converted into a 7-bit bit string “0011000”. In this embodiment, the bit length of the lower-order bit component that makes the bit component unchanged before and after conversion is made variable for each sample.

  Hereinafter, the process performed by the variable-length encoding unit 90 will be specifically described. FIG. 24 is a flowchart showing an outline of variable length coding. First, an average bit length Bf that is a moving average of the bit lengths of past samples is calculated (step S21). The average bit length Bf is obtained by dividing the cumulative bit length RB, which is a cumulative value of the past bit length, by a counter C3 based on the past number of samples. That is, Bf = RB / C3. Since the accumulated bit length RB is 0 in the initial state, when processing a sample with t = 1, the bit length Bd (t) of the sample with t = 1 is set as an initial value. The initial counter C3 = 1 is set.

  Subsequently, the bit length Bd (t) of the sample at time t is calculated (step S22). For samples after t = 2, after calculating the average bit length Bf, the bit length Bd (t) of the sample is calculated. This bit length Bd (t) is easily calculated by converting the bit configuration by the polarity processing means 80. As a result of the conversion to the bit configuration as shown in FIG. 23B, the bit length starts from the point where the bit “1” appears at the head in the bit configuration of each sample. Next, the bit length Bv of the changing unit is calculated (step S23). This is calculated by subtracting the average bit length Bf from the bit length Bd (t) of the sample. Subsequently, the data code is output (step S24). Specifically, “0” is output for the numerical value obtained by decimal conversion of the upper Bv bits, then the separator bit “1” is output, and the lower Bf bits are output as the unchanged part. The code output is performed as recording to an external storage device such as a hard disk or a CD-R. Next, the bit length Bd (t) is added to the cumulative bit length RB (step S25). At the same time, each time processing of each sample is performed, a process of incrementing the counter C3 one by one is performed. Subsequently, it is determined whether or not the counter C3 exceeds a predetermined number (step S26). Here, about 100 is set as the predetermined number. Therefore, it is determined whether or not the counter 4 exceeds 100. As a result, if the counter exceeds 100, the cumulative bit length RB is halved (step S27). Specifically, the variable RB that is the cumulative bit length is divided by two. At the same time, the counter C3 is halved. As described above, encoding with variable bit length is performed for each sample.

  Subsequently, the code output means 100, while maintaining the divided block form of the variable length encoded data of each block output from the variable length encoding means 90, together with each data obtained by the above means, Record sequentially in one encoded file.

(1.3 Structure of encoded data)
The encoded data obtained as described above is stored as needed in a storage device such as a hard disk connected to the computer, and then stored in a format corresponding to a necessary storage medium. Here, FIG. 23 shows an outline of the overall configuration of the finally obtained encoded data. FIG. 25A shows the overall schematic configuration, FIG. 25B shows the schematic configuration of the encoded data in units of blocks, and FIG. 25C shows the schematic configuration of the encoded data in units of channels in each block. As shown in FIG. 25A, high-speed mode identification data, the number of blocks, block length, and encoded data for each block are recorded as the entire encoded data. The high-speed mode identification data is 1-bit data indicating whether the mode is the high-speed mode or the normal mode. In the high-speed mode, encoding processing is performed without using a work memory for each single sample without performing block division. I do. The processing according to the present invention is performed in the normal mode. The number of blocks is 2-byte data indicating the total number of blocks of encoded data. The block length is 4-byte data indicating the number of samples in one block. The encoded data 1 to n are encoded data of each block.

  As shown in FIG. 25B, encoding condition parameters and channel data are recorded as encoded data of each block. The encoding condition parameter is data of 73 bits at maximum in which the encoding condition parameter for each block is recorded. As the encoding condition parameter, whether the lower fixed bits have been deleted / separated by the lower fixed bit deleting means 20, or if deleted / separated has been performed, the sample string rearrangement means 40 rearranges the sample strings. Whether the signal flat part processing means 50 has been processed, the correlation frame detection means 60 has been processed, the update interval of the linear coefficient in the prediction error conversion means 70, and so forth. Each channel data is encoded data for each channel, and when a stereo sound signal is encoded as in this embodiment, two channels are recorded as shown in FIG.

  As shown in FIG. 25 (c), as the encoded data of each channel, lower fixed bit data, sample rearrangement state, signal flat portion data, frame correlation data, and prediction error variable length encoded data are recorded. . The lower fixed bit data is recorded when separated without being deleted by the lower fixed bit deleting means 20. The sample rearrangement state is 2-bit data indicating which of the four states shown in FIGS. 16B to 16E. The signal flat portion data is data as shown in FIG. 17C obtained by the signal flat portion processing means 50. The frame correlation data is data as shown in FIG. 19 (e) obtained by the correlation frame detection means 60. The prediction error variable length encoded data is variable length encoded data obtained by the variable length encoding means 90. By the encoding process as described above, the encoding condition parameter can be changed for each predetermined section, and even a large amount of data can be encoded.

(2. Processing of acoustic signals for using this playback device)
The encoded data obtained as described above is reproduced by the acoustic signal reproducing apparatus according to the present invention. In the acoustic signal reproducing apparatus according to the present invention, a plurality of acoustic signals encoded by the encoding apparatus are synthesized and reproduced. In order to perform the synthesis smoothly, it is necessary to process the reproduction times of the encoded data to be synthesized to be the same. This is performed by increasing the number of samples of other encoded data by interpolating so as to match the encoded data based on the reference with one encoded data as a reference. Further, in this embodiment, each encoded data is created by dividing it into parts such as melody, chord, rhythm, etc. so that the user who reproduces can freely change the composition of music. Each encoded data has the structure shown in FIG.
This playback device is realized by the terminal devices 105 and 107.

(3. Reproduction of acoustic signal)
Hereinafter, a sound signal reproduction apparatus according to the present invention will be described.
(3.1 Configuration of playback device)
FIG. 26 is a block diagram showing an embodiment of a sound signal reproducing apparatus according to the present invention. 26, 1 is a compressed block reading means, 2 is a block decoding means, 3 is a synthesis ratio setting means, 4 is a waveform synthesis processing means, 5 is a synthesis block storage means, 6 is a sound device driver, 7 is a sound device, 8 Is a timer.

  The compressed block reading unit 1 has a function of reading data in compressed block units from a compressed encoded data file. The block decoding means 2 has a function of decoding the read compressed block and restoring it to an uncompressed block that is in a state before compression encoding. The composition ratio setting unit 3 has a function of setting a ratio at which a plurality of uncompressed blocks are to be combined. The waveform synthesis processing unit 4 has a function of synthesizing a plurality of uncompressed blocks decoded by the block decoding unit 2, so-called digital waveform data, at a synthesis ratio set by the synthesis ratio setting unit 3. The synthesized block accumulating means 5 has a plurality of buffer memories for accumulating synthesized synthesized blocks, and the synthesized blocks accumulated in these buffer memories are converted into FIFO (first-in first-out) method, that is, first. It has a function of processing incoming information in a way that goes out first. That is, the synthesis block accumulating unit 5 has a function of accumulating the synthesis blocks input from the waveform synthesis processing unit 4 in the input order and passing them to the sound device driver 6 in that order. The sound device driver 6 has a function of driving the sound device 7 to reproduce the synthesized block sound, and the sound device 7 has a function of D / A converting the synthesized block that is digital data and reproducing it as sound. Have. That is, the sound device driver 6 and the sound device 7 function as synthetic block reproduction means. The timer 8 is a timer used for timing the reproduction of the acoustic signal by the sound device and the reproduction of the acoustic signal of the external device, and is shared with a timer that performs time management in the computer.

(2.2 Processing operation of playback device)
Next, the processing operation of the playback device shown in FIG. 26 will be described. First, the user of the playback device sets a combination ratio of a plurality of encoded data by the combination ratio setting means 3. Specifically, a setting screen as shown in FIG. 5 is displayed on the display device connected to the playback device, and the user is allowed to set on the setting screen. In FIG. 5, five tracks can be selected, and a volume designation section 157 and a matrix 159 are provided on the right side of the track name. The example of FIG. 5 shows a state in which all five tracks are checked, and thereby encoded data for five tracks is read and combined. Further, in the example of FIG. 5, the volume designation is the maximum value “100” for all five tracks, and the five encoded data are synthesized at the same ratio. The matrix 159 designates input of an encoded data file in which an acoustic signal to be reproduced after being decoded and synthesized is compressed. The user sets a track to be used, a volume for each track, and a reproduction file name (encoded data file name) on the player screen 151 as shown in FIG. The set volume for each track is given from the composition ratio setting means 3 to the waveform composition processing means 4 as a composition ratio of the encoded data set for each track. Also, the encoded data corresponding to the reproduction file name set in the matrix 159 is read by the compressed block reading means 1.

  The compressed sound data, which is encoded data, has a structure as shown in FIG. 25 and is recorded in units of blocks. Therefore, first, the compressed block reading means 1 reads the encoded data in units of blocks.

  Subsequently, the encoded data read by the block decoding means 2 is decoded in units of blocks. The decoding process is basically realized by performing the reverse process of the process performed by the encoding apparatus. Here, the characteristic part of the decoding process will be described below. First, the block decoding means 2 determines which part of the read bit string is a change part and which part is an invariant part, and restores a fixed bit length sample.

  Here, an outline of the fixed bit length sample restoration processing is shown in the flowchart of FIG. First, a bit string corresponding to the changing unit is extracted from the bit string of the input code data, and is decoded to a fixed length (step S31). The input bit string is followed by 0 from the beginning and 1 appears. This is because variable length coding was performed according to such a rule at the time of the above coding. Therefore, the block decoding means 2 decodes the bit string when “1” appears according to a rule reverse to that at the time of encoding. For example, when a bit string “0011000...” Comes, it is determined that the first “1” appears up to “001” as a change part, and two “0” s continue. Decode into “0010” (in case of 4 bits).

  Next, an average bit length Bf, which is a moving average of the bit lengths of past samples, is calculated (step S32). This is exactly the same process as the variable length encoding step S21. Subsequently, the bit string of the invariant part is extracted (step S33). This is performed by extracting a bit string corresponding to the calculated average bit length Bf. For example, in the bit string “0011000...” As described above, assuming that up to “001” is the changed part and Bf = 4, the subsequent “1000” is extracted as the invariant part.

  Subsequently, the bit string of the changing part decoded in step S31 and the bit string of the invariant part extracted in step S33 are connected to output a fixed-length sample (step S34). The sample restored to the fixed length is separately subjected to polarity restoration processing. Next, the bit length Bd (t) is added to the cumulative bit length RB (step S35). At the same time, each time the processing of each sample is performed, a process of incrementing the counter C one by one is performed. Subsequently, it is determined whether or not the counter C has exceeded a predetermined number (step S36). Here, about 100 is set as the predetermined number. Therefore, it is determined whether or not the counter exceeds 100. As a result, if the counter exceeds 100, the cumulative bit length RB is halved (step S37). Specifically, the variable RB that is the cumulative bit length is divided by two. At the same time, the counter C is halved. The processing from step S35 to step S37 is exactly the same as the processing from step S25 to step S27 by the variable length encoding means 90 of the encoding device.

  If fixed-length samples are output by the above processing, the block decoding means 2 then restores the positive / negative polarity of each error sample. This is the reverse of the processing performed by the polarity processing means 80 of the encoding device. Specifically, the sign bit of positive / negative is moved to the MSB (least significant bit), the absolute value part is moved down by 1 bit, and then the sample value expressed by the positive / negative signed absolute value is replaced with 2's complement. Convert to an expression using. When the state of bit configuration conversion is schematically shown, the bit configuration shown in FIG. 23B is converted to the bit configuration shown in FIG. This processing makes it easy to perform a bit arithmetic operation by a computer.

  Further, the block decoding unit 2 restores the error sample value in which the prediction error value is recorded to an independent value independent of the past sample value. The outline of the restoration processing to this independent sample is shown in the flowchart of FIG. First, it is determined which prediction calculation formula is applied among a plurality of prediction calculation formulas prepared in advance (step S41). As the prediction calculation formula, any one of [Formula 1] to [Formula 10] used at the time of encoding is used. Which one of [Formula 1] to [Formula 11] is selected is determined based on the accumulated error value corresponding to each prediction formula. The accumulated errors A1 to A10 are the same as those used at the time of encoding, and are all 0 in the initial state.

  In the above [Equation 1] to [Equation 11], e0 (t) to e10 (t) are prediction errors in the sample at time t according to each prediction calculation formula, and x (t) to x (t-5) are times. It is a sample value in t-t-5.

As described above, “a ₂₁ · x (t−1) + a ₂₂ × x (t−2)” in [Expression 3] and “a ₃₁ · x (t−1) + a ₃₂ in [Expression 4] above. X (t−2) + a ₃₃ · x (t−3) ”,“ a ₄₁ · x (t−1) + a ₄₂ · x (t−2) + a ₄₃ · x (t−) in the above [Formula 5] 3) + a ₄₄ · x (t−4) ”,“ a ₅₁ · x (t−1) + a ₅₂ · x (t−2) + a ₅₃ · x (t−3) + a ₅₄ · x (t−4) + a ₅₅ · x (t−5) ”,“ b ₂₁ · x (t−1) + b ₂₂ · x (t−2) ”in [Expression 8], and in [Expression 9] above. “B ₃₁ · x (t−1) + b ₃₂ · x (t−2) + b ₃₃ · x (t−3)”, “b ₄₁ · x (t−1) + b ₄₂ · x” in [Expression 10] above (T- _{) + B 43 · x (t} -3) + b 44 · x (t-4) ", the" _b 51 · _x in [Equation 11] _{(t-1) + b 52} · x (t-2) + b 53 · x “(T−3) + b ₅₄ · x (t−4) + b ₅₅ · x (t−5)” is a linear prediction component based on the past 2 to 5 samples. Using this linear prediction component and the prediction errors “e1 (t−1) / 2” to “e10 (t−1) / 2” (error feedback component) calculated in the immediately preceding sample, the sample e at time t The value of (t) is substituted as any one of e1 (t) to e10 (t) into any prediction formula determined in step S41, and x (t) is calculated (step S42).

  Subsequently, the absolute values of the prediction errors e1 (t) to e10 (t) are added to the accumulated errors A1 to A10 (step S43). Specifically, as shown in the above [Equation 13] used in the encoding apparatus, the variables A1 to A10 that are accumulated error values are updated. At the same time, each time the processing of each sample is performed, a process of incrementing the counter one by one is performed.

  Subsequently, it is determined whether the counter has exceeded a predetermined number of times (step S44). In this embodiment, this predetermined number is set as 100 times. That is, it is determined whether or not the counter exceeds 100.

  As a result, if the counter exceeds 100, the accumulated error is halved (step S45). Specifically, as shown in the above [Equation 14] used in the encoding apparatus, the variables A1 to A10 that are accumulated errors are divided by two. At the same time, the counter is reset to zero. That is, A1 to A10 here are not cumulative errors in a pure sense, but are moving averages of cumulative errors. In the present embodiment, the maximum 100 samples immediately before are accumulated, but the previous ones are processed in half. Thereby, the influence of the sample separated in time is made small.

  By executing the processes in steps S41 to S45 on all the error samples that have been read, the values of all the samples are restored from the original prediction error e (t) to the amplitude value x (t).

  As described above, by performing a decoding process corresponding to the encoding process performed in the encoding apparatus, the samples of the original digital sound signal are restored in units of blocks, and an uncompressed block is obtained. However, in the reproduction apparatus according to the present invention, the above decoding process is performed in parallel on each read encoded data. As a result, a plurality of uncompressed blocks are obtained at the same time.

  Subsequently, the plurality of uncompressed blocks obtained are synthesized by the waveform synthesis processing means 4 to generate a synthesized block. Specifically, it is performed by adding each sample constituting each non-compressed block multiplied by the synthesis ratio. As a result, a composite block is obtained.

  The synthesized block obtained by synthesizing by the waveform synthesis processing unit 4 is accumulated in the synthesized block accumulating unit 5. In this embodiment, since it is possible to store up to 4 blocks in the synthetic block storage means, the processing by the sound device driver 6 is not started until 4 blocks are stored. As shown in FIG. 29, when four synthesized blocks are accumulated in the synthesized block accumulating unit 5, the sound device driver 6 reproduces the first block among the synthesized blocks accumulated in the synthesized block accumulating unit 5. Specifically, the sound device 7 performs D / A conversion on the synthesized block data and outputs it to the speaker. The synthesized block that has been acoustically reproduced is deleted from the synthesized block storage means 5.

  When the synthesized block is deleted and there is room in the synthesized block storage unit 5, the synthesized block synthesized by the waveform synthesis processing unit 4 is input to the synthesized block storage unit 5. As a result, the combined block storage means 5 is stored up to the maximum capacity again. The synthesized block synthesized by the waveform synthesis processing means 4 is actually put into the synthesized block storage means 5 when the CPU functions as a synthesized block throwing means. This synthesis block input means not only simply inputs the synthesis block into the synthesis block storage means 5, but also when the synthesis block storage means 5 has no free space, the compressed block reading means 1, the block decoding means 2, and the waveform synthesis processing. A message for interrupting the processing is sent to the means 4 to control the input of the composite block to the composite block storage means 5.

  On the other hand, the sound device driver 6 sequentially plays back the first block among the synthesized blocks stored in the synthesized block storage means 5. At this time, the sound device driver 6 sends a message permitting execution of each process to the synthesis block input unit, the compressed block reading unit 1 and the block decoding unit 2 every time the sound reproduction of one synthesis block is finished. .

  Here, the outline of the processing in the playback apparatus is organized and shown in the flowchart of FIG. First, the synthesis block input unit searches for a vacant buffer memory in the synthesis block storage unit 5 (step S51). If there is no free buffer memory, a message for interrupting the processing is sent to the compressed block reading means 1, the block decoding means 2, and the waveform synthesis processing means 4, and the reception of the reproduction end message from the sound device driver 6 is awaited. (Step S52). If there is a playback end message from the sound device driver 6, it is deleted from the buffer memory storing the synthesized block for which playback has ended (step S53). Since the reproduction end message from the sound device driver 6 is simultaneously transmitted to the synthesis block input means, the compressed block reading means 1, the block decoding means 2, and the waveform synthesis processing means 4, the compressed block reading means 1 and the block decoding means 2 are also transmitted. Then, the waveform synthesis processing means 4 restarts the process, and decoding of the uncompressed block and synthesis of the uncompressed block are performed (step S54). Subsequently, the composite block is stored in an empty buffer memory (step S55). On the other hand, the sound device always searches the buffer memory in the synthesized block storage means 5 (step S56), and if a synthesized block exists, the synthesized block is reproduced (step S57). Waiting for playback of one synthesis block (step S58), when playback is completed, a playback end message is transmitted to the synthesis block input means, compressed block reading means 1, block decoding means 2, and waveform synthesis processing means 4 (step S59). .

The figure which shows the system with which the reproducing | regenerating apparatus of the acoustic signal which concerns on this Embodiment was integrated. Block diagram showing the configuration of the web server 103 The block diagram which shows the structure of the terminal device 105 The flowchart which shows the process of the terminal device 105 The figure which shows the player screen 151 Diagram showing object tags The figure which shows the screen of the terminal device 105 on which the object tag was displayed Figure showing an HTML document Diagram showing an HTML document with an embedded object tag Figure showing homepage screen Diagram showing object tags Figure showing a metafile The figure which shows the screen of the terminal device 5 on which the object tag was displayed The figure which shows the player screen 151 Functional block diagram of an audio signal encoding device The figure which shows the mode of the rearrangement of the sample by the sample row rearrangement means 40 The figure which shows the mode of the process by the signal flat part process means 50 Flowchart showing processing by correlation frame detection means 60 The figure which shows the mode of the sample row | line | column by the process of the correlation frame detection means 60 The figure which shows the mode of the sample compared by the process of the correlation frame detection means 60 The flowchart which shows the process by the prediction error conversion means 70 Diagram showing settable linear coefficient setting mode The figure which shows the mode of a bit structure conversion by the polarity processing means 60 The flowchart which shows the process by the variable length encoding means 90 The figure which shows the whole structure of the encoding data obtained as a result of encoding Functional block diagram of a sound signal reproducing apparatus according to the present invention. The flowchart which shows the outline | summary of the process which restores the variable length sample in the block decoding means 2 to fixed length. The flowchart which shows the outline | summary of the decoding process from the prediction error in the block decoding means 2. The figure which shows the reproducing | regenerating apparatus of the acoustic signal of the state in which the synthetic block was accumulate | stored Flow chart showing processing operation of acoustic signal reproducing apparatus

Explanation of symbols

1 ... Network 3 ... Web server 5, 7 ... Terminal equipment

Claims (7)

In a system in which a server and a terminal device are connected via a network,
The server is
Generating means for generating an object tag having information for selecting a plurality of music materials to be reproduced from a plurality of music materials composed of sound signals compressed by encoding ;
Means for writing the object tag in a predetermined position of the HTML document;
Comprising
The terminal device is
When accessing the HTML document, in accordance with the description of the object tag, for each of the compressed audio signals corresponding to the plurality of selected music materials , the compressed block is read, and the compressed block is decoded and decoded. and means for reproducing a plurality of uncompressed blocks and waveform synthesis was,
An acoustic signal reproducing apparatus comprising:   2. The coding is characterized in that lossless compression is performed by adaptive linear predictive coding so that a block of a given acoustic signal is divided and each block can be completely restored. Sound signal reproducing device. Further comprising means for setting a matrix;
Said generating means in response to a matrix that is set, the reproducing apparatus of the audio signal according to claim 1, wherein the generating the object tags having information for selecting music material to be reproduced. The sound signal reproducing apparatus according to claim 3, wherein the server can set the matrix. The sound signal reproducing apparatus according to claim 3, wherein the terminal device can set the matrix. A program causing a computer to function as the server according to claim 1 . A program causing a computer to function as the terminal device according to claim 1 .

Images