JP3066452B2 - Sound characteristic conversion device, sound / label association device, and methods thereof - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to characteristic conversion for sound such as voice, musical sound, and natural sound, and more particularly to facilitation of the conversion operation. In addition, the present invention relates to a correspondence between a sound and a label suitable for the characteristic conversion.

2. Description of the Related Art It has been practiced to convert characteristics of sound such as voice to obtain desired characteristics. To convert a property,
It is common to modify the time domain waveform or frequency spectrum of a sound by modifying the waveform or spectrum. For example, take in an analog audio signal,
An operation of converting the digital data into digital data, subjecting the digital data to waveform deformation corresponding to the desired characteristic conversion, and then converting the digital data back into an analog signal is performed. Thereby, the characteristics of the audio signal can be converted to desired characteristics.

However, the conventional characteristic conversion as described above has the following problems. The processing of the characteristic conversion includes the time domain waveform of the sound, the frequency spectrum,
This is performed by displaying the parameters and the like of C) on the display and operating them. In order to obtain desired characteristics by this operation, there is a problem that it is necessary to have expertise in a time domain waveform, a frequency spectrum, parameters of a linear predictive analysis (LPC), and the like.
Further, there is a problem that sufficient training is required to perform a desired characteristic conversion even if the user has specialized knowledge.

DISCLOSURE OF THE INVENTION An object of the present invention is to solve the above-described problems and provide a sound quality conversion device and method that can be easily converted, and a sound and label association device and method suitable for these. I do.

A sound characteristic conversion device according to the present invention comprises: sound / label data holding means for holding sound data divided according to a predetermined division, and label data associated with each division of the sound data; When the decoration data is given, based on the decoration data, the display control means for visually modifying the label based on the label data and displaying the label on the display means, and the sound data associated with the label data. And conversion means for performing a corresponding characteristic conversion based on the decoration data given corresponding to the label data.

The sound characteristic conversion device of the present invention includes a sound data classification unit that classifies input sound data based on a sound segment, and a label data input with a delimiter code corresponding to the sound segment. Label data dividing means for dividing based on the delimiter code, and correspondence forming means for associating the divided sound data and the divided label data with each other are provided.

The sound characteristic conversion device of the present invention is characterized in that the visual modification to the label is a character decoration on the label.

The sound characteristic conversion device of the present invention is characterized in that the visual modification to the label is the order of the label.

The sound characteristic conversion method of the present invention associates sound data with label data, associates sound characteristic conversion contents with a modification process, and assigns a label represented by the label data to a given modification process. The sound data associated with the label data is subjected to characteristic conversion corresponding to the modification processing given to the label data.

The sound characteristic conversion method of the present invention classifies input sound data based on a sound segment, classifies label data according to the sound segment, and classifies the divided label data with the classified sound data. It is characterized by the fact that it is attached.

A sound characteristic conversion device according to the present invention comprises: sound / label data holding means for holding sound data segmented according to a predetermined segment and label data associated with each segment of the sound data; Conversion means for performing a corresponding characteristic conversion on the attached sound data based on the modification data given in correspondence with the label data.

The sound characteristic conversion method of the present invention associates sound data with label data, associates characteristic conversion contents with a modification process, and converts sound data associated with label data into label data. It is characterized in that characteristic conversion corresponding to the modification processing performed on the image is performed.

A system according to the present invention includes a transmitting device and a receiving device communicable via a communication path, and is a system for transmitting sound data from a transmitting device to a receiving device. Data input means for inputting label data and decoration data, and communication means for transmitting the label data and decoration data to the receiving device via a communication path. The receiving means comprises a label from the transmitting device. Communication means for receiving data and modification data; standard sound data generation means for generating standard sound data based on the label data; and converting sound characteristics of the standard sound data based on the modification data, and converting sound characteristic conversion data. Conversion means for generating.

The transmission method according to the present invention is a method for transmitting sound data from a transmission side to a reception side via a communication path. The transmission side inputs label data and modification data, and communicates the label data and modification data. The receiving side receives the label data and the modification data from the transmitting side, generates the standard sound data based on the label data, and transmits the standard sound data based on the modification data. It is characterized by converting sound characteristics and generating sound characteristic conversion data.

A sound / label associating device according to the present invention includes: sound data input means for inputting sound data; sound data classification means for classifying sound data based on a loudness of a sound represented by the sound data; Label data input means for inputting label data with a delimiter code at a position corresponding to the division, label data division means for dividing the label data based on the delimiter code, divided sound data and the divided sound data And association forming means for associating the label data with each other.

A sound / label associating device according to the present invention includes a sound data input unit for inputting sound data, a label data input unit for inputting label data corresponding to the sound data, and an average duration of each label represented by the label data. And detailed correspondence forming means for classifying the sound data in association with each label based on the sound data and the duration of the sound data.

The sound / label associating apparatus according to the present invention relates to label data and sound data associated by the association forming means, based on the average duration of each label represented by the label data and the duration of the audio data. There is provided detailed correspondence forming means for classifying data in association with each label.

A sound / label associating device according to the present invention includes a sound display unit for visually displaying a property of a sound represented by sound data, and a label display unit for displaying a label represented by label data. The sound display unit is characterized in that a delimiter mark indicating a sound delimiter is displayed based on a loudness of a sound.

The sound / label associating method of the present invention classifies sound data based on the loudness of sound represented by sound data, and converts label data with a delimiter code at a position corresponding to the division of the sound data. Then, based on the delimiter, the label data is divided, and the divided sound data and the divided label data are associated with each other.

The sound / label associating method of the present invention is a method of associating sound data with a label, in which an average duration time for each label is prepared in advance, and the average of each label represented by the label data is averaged. It is characterized in that sound data is classified in association with each label based on the duration and the duration of the sound data.

According to the sound / label association method of the present invention, an average duration for each label is prepared in advance, and for the associated label data and audio data, the average duration and audio for each label represented by the label data are recorded. It is characterized in that sound data is classified in association with each label based on the duration of the data.

The display method for sound / label correspondence according to the present invention is as follows.
A sound display unit for visually displaying the nature of the sound represented by the sound data, and a label display unit for displaying a label represented by the label data, wherein the sound display unit includes a sound display unit based on the volume of the sound. In this case, a delimiter mark indicating a sound delimiter is displayed.

In the present invention, label data refers to character strings, graphic strings, symbol strings, picture rows, and the like, and combinations thereof, which can be associated with sounds such as voices and natural sounds. And icon data corresponding to characters and the like.

The sound data is data that directly or indirectly represents the waveform of a sound, and includes, for example, data obtained by digitizing an analog waveform of the sound, data that represents the sound using LPC parameters, and the like.

Modification is a concept that includes emphasis, addition of an underline, addition of a sign and the like, and change of the order, etc., performed on a label in order to convert characteristics of a sound obtained based on sound data. Although the contents of the decoration data may indicate the contents of the property conversion, when visually modifying the label data based on the decoration data, it is preferable to indicate the contents of the visual modification. is there.

Sound characteristic conversion refers to changing some property of the sound, for example, pitch change, intensity change,
Addition of vibrato, change of frequency spectrum, change of duration, change of sampling interval, feminization, masculinization, clarification, unclearness, etc. This is a concept that also includes replacement of parts, partial deletion of sounds, and the like.

According to the sound characteristic conversion device of the first aspect and the sound characteristic conversion method of the fifth aspect, the label data is associated with the sound data, and the sound characteristic conversion content and the modification process are associated with each other, and are represented by the label data. Label is visually modified based on the applied modification processing, and the sound data associated with the label data is subjected to the characteristic conversion corresponding to the modification processing applied to the label data. It is characterized by performing. Therefore, the characteristic conversion for the sound can be performed only by visually modifying the corresponding label.

According to a second aspect of the present invention, there is provided a sound characteristic converting apparatus and a sound characteristic converting method, wherein input sound data is divided based on a sound segment, and label data is divided according to the sound segment. The classified label data is associated with the divided sound data. Therefore, it is possible to associate the two simply by inputting the sound data and the label data.

According to the sound characteristic conversion device of claim 7 and the sound characteristic conversion method of claim 8, while labeling data is associated with sound data, characteristic conversion contents and modification processing are associated with each other,
Characteristic conversion is performed on the sound data associated with the label data with respect to the modification processing performed on the label data. Therefore, it is possible to perform the characteristic conversion on the sound only by performing the modification process on the label data in which the syllable division is clearer than the sound data.

In the sound transmission system according to the ninth aspect and the sound transmission method according to the tenth aspect, label data and modification data are input on the transmitting side, and standard sound data is generated on the receiving side based on the label data. The data is subjected to sound characteristic conversion based on the modification data to generate sound quality conversion data. Therefore, a sound having desired sound characteristics can be sent only by sending the label data and the modification data.

The sound / label associating device according to claim 11 and the sound / label associating method according to claim 15 provide a sound volume represented by sound data by setting two thresholds relating to the sound volume, When the volume exceeds a threshold for more than a predetermined time continuously, the division is started. When the volume of the sound continuously falls below another threshold for more than a predetermined time, the classification ends. In such a way, the label data is divided and the label data with the delimiter attached to the position corresponding to the division of the sound data is received, and the label data is divided based on the delimiter, and the divided sound data and the divided sound data are divided. It is characterized in that label data is associated with each other. Therefore, when the sound data is separated from the divided sound data, the label data can be easily associated with the sound data.

A sound / label associating apparatus according to claim 12 and a sound / label associating method according to claim 16 are methods for associating sound data with a label, based on an average duration of each syllable of the label data. The sound data is classified according to each label on the basis of the average duration of the label and the duration of the sound data obtained in this way. Therefore, it is possible to easily associate the sound data with each label.

A sound / label associating device according to claim 14 and a display method for sound / label associating according to claim 18, comprising: a sound display unit for visually displaying a property of a sound represented by the sound data; A label display unit for displaying a label represented by the following formula. In the sound display unit, two loudness thresholds are set for the loudness of the sound, and the loudness of the sound is continuously longer than a predetermined time. When the threshold value exceeds one threshold, the segmentation starts, and when the loudness falls below another threshold continuously for a predetermined time or more, the segmentation ends. It is characterized in that a delimiter mark indicating a delimiter is displayed. Therefore, it is possible to input and display label data while confirming the separation position of sound data. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a diagram showing a display screen of a sound characteristic conversion device according to a position embodiment of the present invention. It is.

FIG. 2 is a diagram showing an overall configuration of a sound quality conversion device according to one embodiment of the present invention.

FIG. 3 is a diagram showing a hardware configuration when a CPU is used to realize the functions of FIG.

FIG. 4 is a flowchart showing the operation of the sound characteristic conversion device.

FIG. 5 is a flowchart showing the operation of the sound characteristic conversion device.

FIG. 6 is a diagram showing label data stored in association with audio data.

 FIG. 7 is a diagram showing a storage state of audio data.

 FIG. 8 is a diagram illustrating a label displayed on the CRT 16.

FIG. 9 is a diagram showing the correspondence between visual modification and sound quality conversion content.

FIG. 10 is a diagram showing label data to which decoration data is added.

FIG. 11 is a diagram showing a label to which visual modification has been applied.

FIG. 12 is a diagram for explaining division of audio data.

 FIG. 13 is a diagram showing a pitch conversion process.

 FIG. 14A is a diagram showing a sound source waveform before pitch conversion.

 FIG. 14B is a diagram showing a sound source waveform after pitch conversion.

FIG. 15A is a diagram showing audio data before power change and its short-term average power.

FIG. 15B is a diagram showing audio data after power change and its short-term average power.

 FIG. 16A is a diagram showing original audio data.

FIG. 16B is a diagram showing audio data in which the time length of sound has been changed.

FIG. 16C is a diagram showing audio data to which vibrato has been applied.

FIG. 17 is a diagram illustrating an example of a symbol used as a decoration for an icon.

 FIG. 18 is a diagram showing a change in the order of sounds.

FIG. 19 is a diagram illustrating an example of a delimiter process for Japanese.

 FIG. 20 is a diagram illustrating an example of classification for each label.

 FIG. 21 is a diagram illustrating an embodiment of a voice transmission device.

FIG. 22 is a diagram showing an example of data transmitted in the embodiment of FIG.

FIG. 23 is a diagram illustrating a table in which the decoration data is encoded.

BEST MODE FOR CARRYING OUT THE INVENTION FIG. 2 shows an overall configuration of a sound quality conversion device according to an embodiment of the present invention. The sound data classification means 2 classifies the input sound data based on sound delimiters. Label data to which a delimiter code corresponding to a sound delimiter is added is input to the label data dividing means 4. The label data classifying means 4 classifies the label data based on the delimiter. The segmented sound data and the segmented label data are input to the association forming unit 6, and are associated with each other for each segment. The associated sound data and label data are held in the sound / label holding means 8.

The display control means 10 receives the modification data for each section, modifies the corresponding label data, and displays the modified label on the display means 14. As a result, it is possible to easily confirm what kind of modification has been applied to which category. The conversion means 12 receives the modification data for each section, modifies the corresponding sound data, and outputs the modified sound data.

FIG. 3 shows a hardware configuration when the configuration in FIG. 2 is realized using a CPU. The bus line 40 includes a CRT 16 and a CPU 18 as display means, a memory 20 as sound and label data holding means, an input interface 22, and a hard disk 2
4. Output interface 26 and floppy disk drive (FDD) 15 are connected. A microphone 30 is connected to the input interface 22 via an A / D converter 28. Further, a keyboard 32 and a mouse 34 are also connected to the input interface 22. A speaker 38 is connected to the output interface 26 via a D / A converter 36. The hard disk 24 stores a program whose flowchart is shown in FIGS.
This program is installed on the hard disk 24 from a floppy disk (recording medium) by the FDD 15. Of course, it may be installed from a recording medium such as a CD-ROM. Memory 20
Is a sound / label data holding unit and is also used as a work area for executing a program.

The processing operation of the CPU 18 will be described with reference to FIGS. First, in step S1, an audio signal (analog audio data) is input from the microphone 30. CPU18
When an audio signal is input, the digital data processor fetches digital data (digital audio data) converted by the A / D converter. Further, the CPU 18 converts the waveform of the audio data into a CRT
It is displayed on the 16 sound display sections 80. This display state is shown in FIG.

Next, the digital audio data is divided based on sound delimiters (step S2). This division is performed as follows. For example, "Hi my name is John Nic
"e to meet you" is input. It is assumed that the digital audio data obtained at this time is as shown in the upper part of FIG. The upper part of FIG. 12 shows a waveform display of digital audio data. The CPU 18 calculates the short-time section average power based on the digital audio data.
The calculated short-time section average power is shown in the lower part of FIG.

Next, the CPU 18 determines two levels of the data level and the skip level.
Classification is performed based on one threshold. After the end of the classification, if the short-time section average power continuously exceeds the data level for 100 ms or more, the classification is started. In addition, after the start of classification, the average power for the short-time section is 80 ms
If the skip level is continuously lower than the above, the division is terminated. In this way, classification is performed. In this embodiment, the data level is 50 dB and the skip level is 40 dB.

Based on the above classification, as shown in FIG.
mS is the first category, 630mS to 1890mS is the second category, 2060mS to 239
0 ms can be determined to be the third category. The CPU 18 displays lines 84a, 84b, 84c, and 84d indicating the division positions on the waveform of the sound display unit 80 of the CRT 16 based on the determined division (see FIG. 1).

Further, the CPU 18 stores the divided digital audio data in the memory 20 (step S3). Each sound data stored in the memory 20 is shown in FIG. The first section is below address ADRS1, the second section is below address ADRS2,
The third section is stored at addresses ADRS3 and below, respectively.

Next, label data corresponding to the above voice data is input from the keyboard 32 to the label display section 82 of the CRT 16 shown in FIG. 1 (step S4). At this time, a punctuation mark is added as a delimiter code at the same position as the voice delimiter and input. For example, for the above audio data,
Enter "Hi, my name is John. Nice to meet you." The CPU 18 receives this input, and converts the label data into “Hi”, “my name is John”, “Nice to
meet you ".

In this embodiment, as shown in FIG. 1, lines 84a, 84b, 84c, and 84d indicating division positions for sound quality data are displayed. Therefore, when inputting label data, it is easy to input a delimiter code in association with the label data.

The CPU 18 stores the segmented label data in sequence in association with the segmented audio data (step S5). That is, as shown in FIG. 6, the head address of the corresponding audio data is stored together with each label data.

When the number of sections of audio data does not match the number of sections of label data, it is preferable to correct the number of sections of audio data based on the number of sections of label data. That is, the threshold value (data level and skip level) of the audio data division may be changed, the audio data may be divided again, and the number of divisions may be matched. Or,
The number of segments may be matched by newly setting or deleting the segment position of the audio data by inferring from the number of characters in the label data. The worker may correct the division using the mouse 30 and the keyboard 32.

Next, the CPU 18 displays a label based on the input label data on the label display unit 82 (see FIG. 1) of the CRT 16 (step S6). FIG. 8 shows the displayed label.
Next, the operator applies a predetermined visual modification to the displayed label in accordance with the content of each sound characteristic conversion. FIG. 9 shows an example of the correspondence between the visual modification and the sound characteristic conversion. If this is stored as a correspondence table, the correspondence between visual modification and sound characteristic conversion can be changed by changing this content. Note that the contents of FIG. 8 are displayed as icons on the CRT 16 as shown in FIG. 1, so that the operation is easy as guidance.

If you want to increase the power only for "my name is John", perform the following operation. First, the keyboard 32
Alternatively, using the mouse 34, the label display section 82 shown in FIG.
Select the "name is John" part. Next, select "m
y name is John ”icon 90
Is clicked with the mouse. As a result, as shown in FIG. 10, the decoration data “＼ emphasis” is added to “my name is John” in the memory 20. Here,
"@" Is a code indicating that the following character string is a control code (modification data).

In step S7, the CPU 18 displays the label modified based on the modification data on the label display section 82 of the CRT 16 (see FIG. 11). As is clear from FIG. 11, it is possible to easily confirm the location where the characteristic conversion is performed and the content thereof.

Next, the CPU 18 reads the first section of the label data shown in FIG. 10, and reads the corresponding audio data based on the start address ADRS1 (step S8). This allows
The digital audio data of “Hi” shown in FIG. 12 is read. Next, it is determined whether or not decoration data is added to the label data (step S9).
Here, since no decoration data is added, the process proceeds to step S11.

In step S11, the CPU 18 determines whether or not all the sections have been processed. If not, the process from step S8 is repeated for the next section (step S12). For the next segment, "my nam e is John"
Modification data “＼ emphasize” is added. Therefore,
The process proceeds from step S9 to step S10.

In step S10, a predetermined characteristic conversion is performed on the digital audio data “my name is John” for “＼ emphasis”. Here, the power of the audio data is increased according to the table of FIG.
The power is increased by increasing the amplitude of the waveform represented by the digital audio data. The sound data whose characteristics have been converted in this manner is again shown in FIG.
(The original audio data may be stored at another address in order to retain the original audio data.)

When the processing for all the sections is completed, the CPU 18
The digital voice data subjected to the sound quality conversion is output from the output interface 26 (step S13). FIG. 15A shows audio data before the characteristic conversion, and FIG. 15B shows audio data after the characteristic conversion. It can be seen that the power of "my name is John" has been converted to be larger. The digital audio data thus converted is converted into analog audio data by the D / A converter 36, and is output from the speaker 378 as audio whose characteristics have been converted. That is, "my nam
e is John "is enlarged and output.

As described above, the sound quality can be converted simply by applying visual modification to the label, and the operation is extremely easy. Further, it is possible to easily confirm what kind of sound quality conversion is performed for which section.

Note that the pitch can be raised in the same manner. In this case, after selecting the part whose pitch is to be increased, the icon 92 may be selected (see FIG. 1).

FIG. 13 shows the processing procedure for pitch rise. First, the CPU 18 performs linear prediction analysis (LPC) on the target digital audio data, and separates the audio data into sound source data and vocal tract transfer characteristic data. Next, the pitch of the separated sound source data is changed. After that, the digital voice data is re-synthesized with the vocal tract transfer characteristic data to obtain digital voice data with an increased pitch. As for linear prediction analysis,
For details, see "Linear Prediction of Voice" (JDMarkel: AHGray, Jr., translated by Kuki Suzuki, Corona). FIG. 14 shows a part of the digital voice data before the pitch rise and a part of the digital voice data after the pitch rise.

16A, 16B, and 16C show other conversion examples of sound characteristics. FIG. 16A shows audio data before conversion, and FIG. 16B shows “my
This is audio data after the time length of the sound has been changed for "name is john". The size of the label is processed so as to correspond to the time length.

FIG. 16C shows audio data of “my name is John” after being subjected to vibrato. The label is underlined. The type of vibrato may be changed according to the type of underline.

The description of all sound characteristic conversions is omitted because it is difficult to explain, but the present invention is directed to other sound characteristic conversions in general. When performing sound characteristic conversion in the frequency domain, a process may be performed by obtaining a frequency spectrum by FFT or the like.

The sound characteristic conversion includes a process of changing a sound order, a process of deleting a part of a sound, and a process of repeating, in addition to a process of mainly changing the sound quality as described above. For example, as shown in FIG. 18, the order of sound output may be changed by changing the order of labels associated with sounds. In this example, "Hi my name is John N
ice to meet you '' as `` Hi Nice to meet you my name i
s John ". Similarly, the sound can be deleted by deleting the label, or the sound can be repeated by duplicating the label.

Further, in the above-described embodiment, a description has been given using a character string as an example of a label. However, a symbol, a sign, or the like may be used, or an icon may be used. In addition, the method of modification is also performed by superimposing the mark for the male in FIG. 17A in the case of voice masculinization and the mark for the female in FIG. 17B in the case of feminization, over the target icon. Is also good.

Alternatively, the content of the sound quality conversion may be determined in association with the face photograph displayed on the screen, and the content of the sound quality conversion may be determined by selecting the face photograph with a mouse or the like.

Further, in the above-described embodiment, the description has been given of the voice.
The invention can be applied to all sounds such as natural sounds such as musical sounds, wind sounds, and wave sounds.

Further, in the above embodiment, the sound is input from the microphone 30, but the sound may be synthesized based on the label data. At this time, the basic sound is synthesized based on the label data, and the synthesized basic sound is subjected to characteristic conversion and output based on the modification content applied to the label data. Alternatively, the sound may be described and given as data using LPC parameters or the like.

In the above example, English speech was described, but the present invention is applicable regardless of language. FIG. 19 shows the processing state of the classification for the Japanese voice input “Yes, thank you, thank you”.

In each of the above embodiments, the sound quality is converted by modifying each section. However, if audio data is divided into syllables based on the number of label data in each division, sound quality conversion can be performed for each syllable. Such finer divisions will be described with an example of a Japanese voice input of "tatsusuru" (of course, it is applicable to other languages).

First, the average duration time when several subjects utter each element of the label is measured. This is stored in the hard disk 24 in advance as a table shown in Table 1.

First, the total time length T of the input voice data "Tatsuru" is actually measured. Here, for example, it is assumed that the actually measured total time length T is 802 ms. Next, each element of the label data “Tatsusuru” associated with the audio data is divided into categories according to Table 1. In other words, "ta""ssu"
It is decomposed into "R" and determined to be categories CV, DV and CV, respectively. The CPU 18 sums up the average durations t1, t2, and t3 of each element based on Table 1. Here, 204.Oms + 3
81.Oms + 204.Oms = 789.Oms is obtained. Further, based on the total time length t and the average time lengths t1, t2, and t3 of each element,
The time length ratios r1, r2, and r3 of each element are calculated. For example,
The time length ratio r1 of the element “ta” is 204.0 / 789.0. Similarly, the time length ratios r2 and r3 of the elements "ss" and "ru" are 381.0 / 789.0 and 204,0 / 789.0, respectively.

The time length ratios r1, r2,
Based on r3, the actually measured total time length T is allocated to each element. For example, the real time T1 allocated to the element “ta” is calculated by T1 = T · r1. Real time allocated to the elements "ss" and "ru"
Similarly, T2 and T3 are calculated as T2 = Tr2 T3 = Tr3. The real time T1 calculated in this way,
Based on T2 and T3, the audio data is divided as shown in FIG. By performing the above-described processing, the audio data can be classified in more detail (for each element of the label) and associated with the label (formation of detailed correspondence). Therefore, it is possible to perform sound characteristic conversion for each label element. For example, it is possible to apply vibrato to only “Ru” by modifying only “Ru” by underlining.

In this way, it is possible to classify each syllable in a simple manner. Note that each syllable may be more accurately estimated using a speech recognition technique.

In the above embodiment, the display control means 10 and the display means 14 are provided so that the modification can be performed while confirming the modification to the label data. However, even if these means 10 and 14 are not provided, it is possible to input the decoration data if the structure of the decoration data is known as shown in FIG. in this case,
Although the decoration cannot be confirmed by the display means 14,
It has the following effects.

Although it is possible to add modification data to sound data, it is difficult to perform sound quality conversion over a predetermined syllable range because the syllables are not clearly distinguished in the sound data itself. On the other hand, in the label data, the division between characters (corresponding to the division of each syllable) is clear,
It is easy to perform sound quality conversion over a predetermined syllable range. That is, it is easy to perform sound quality conversion on syllables in a desired range.

In the above embodiment, the CPU is used to realize the function of each block in FIG. 2, but a part or all of the CPU may be realized by hardware logic.

FIG. 21 shows one embodiment of the voice transmission system. Communication channel
The transmitting device 52 and the receiving device 60 are connected via 50. The communication path 50 may be wired or wireless. The transmitting device 52 includes a data input unit 54 such as a keyboard and a communication unit 56. In addition, the receiving device 60 includes a standard audio data generation unit 62, a communication unit 64, a conversion unit 66, and an audio output unit 68.

Hereinafter, a case where voice is transmitted from the transmitting device 52 to the receiving device 60 will be described as an example. First, the data input means
From 54, label data and decoration data as shown in FIG. 22 are input. The parts of “＼female” and “＼male” are the modification data, which determine the content of the sound quality conversion of the label data in the parentheses that follow. In this embodiment, it means to convert into a "＼female" female voice, and "、 male"
Means converting to a masculine voice.

Next, this data is transmitted by the communication means 56 to the receiving device 60 via the communication path 50. The communication means 64 of the receiving device 60 receives this and temporarily holds it. The standard audio data generating means 62 acquires the stored data and extracts only the label data from the stored data. Here, "Good morning" and "How are you?" Based on the label data, the standard voice data generating means 62 generates standard voice data corresponding to the label data by a voice synthesis method or the like.

On the other hand, the conversion unit 66 extracts only the decoration data from the data held in the communication unit 64. Here, “＼women” and “＼men” are extracted. The conversion means 66 converts the sound quality of the corresponding part of the standard audio data based on the modification data. The relationship between the decoration data and the content of the sound quality conversion is predetermined. Here, "Good morning" is converted into feminized voice data, and "Good morning" is converted into feminized voice data. The conversion means 66 outputs the sound quality conversion data thus obtained.

The audio output means 68 converts the sound quality conversion data into an analog signal and outputs the analog signal from a speaker.

Audio is transmitted from the transmitting device 52 to the receiving device 60 as follows. According to this embodiment, sound can be transmitted only by transmitting label data and modification data having a small data amount. In addition, not only a standard voice but also a voice having a desired sound quality can be transmitted based on the modification data.

In the conventional apparatus, audio data having a large data amount was transmitted, so that the transmission speed was low. However, according to this embodiment, this can be dramatically improved.

If the decoration data is complicated, a code may be added to the decoration data and stored in the receiving device 60, and only the code may be sent. For example, as shown in FIG. 23, the decoration data of {emphasis} italic ＼25 points,
It is convenient to memorize it with the symbol “A”.

Further, in order to confirm what kind of modification data is attached to the label data in the transmitting apparatus, the label may be modified by the modification data and displayed as in the embodiment of FIG. Good.

Note that the various modifications, applications, and extensions described in the embodiment of FIG. 2 can be applied to the present embodiment. For example, in this embodiment, the transmission target is audio, but the present invention can be applied to all other sounds.

──────────────────────────────────────────────────続 き Continuation of the front page (56) References JP-A-60-245000 (JP, A) JP-A-61-1730 (JP, A) JP-A-61-93484 (JP, A) JP-A-61-93484 100799 (JP, A) JP-A-1-219635 (JP, A) JP-A-5-143278 (JP, A) JP-A-6-337876 (JP, A) (58) Fields investigated (Int. ⁷ , DB name) G10L 13/00-13/06 G06F 3 / 16,15 / 20

Claims (19)

(57) [Claims] 1. A sound data segmented according to a predetermined segment and a label which is segmented by a delimiter code and indicates a transmission content of a sound indicated by the sound data as a character, a picture or a symbol for each segment of the sound data. Sound / label data holding means for holding the label data in the category in which the conversion of the sound characteristic is desired is given, when the decoration data is given, based on the decoration data,
Display control means for visually modifying a label based on the label data of the category and displaying the label on the display means, and giving sound data associated with the label data of the category in accordance with the label data of the category. A conversion means for performing a corresponding characteristic conversion based on the obtained modification data. 2. Sound data classification means for classifying input sound data based on sound divisions, and label data indicating, as characters or symbols, transmission contents of sounds indicated by the sound data. A label data dividing means for dividing the label data inputted with the corresponding delimiter code based on the delimiter code, a correspondence forming means for associating the divided sound data and the delimited label data with each other, and a label data. When the modification data is given to the sound data associated with the label data, display control means for visually modifying the label based on the label data and displaying the label on the display means based on the modification data. ,
A conversion unit for performing a corresponding characteristic conversion based on the modification data given in correspondence with the label data. 3. The sound characteristic conversion device according to claim 1, wherein the visual modification of the label is a character decoration on the label. 4. The sound characteristic conversion device according to claim 1, wherein the visual modification to the label is an order of the label. 5. The sound data is associated with a label indicating a transmission content of the sound indicated by the sound data as a character, a picture or a symbol, and the sound characteristic conversion content is associated with a modification process. The label represented by the data is visually modified based on the given modification processing and displayed, and for the sound data associated with the label data,
Performing a characteristic conversion corresponding to a modification process given to the label data. 6. The input sound data is classified based on a sound delimiter, and label data indicating, as a character or a symbol, the transmission content of the sound indicated by the sound data, the delimiter code corresponding to the sound delimiter. Is divided based on the delimiter code, and the divided sound data and the delimited label data are associated with each other. Based on the data, the label based on the label data is visually modified and displayed on the display means, and the sound data associated with the label data is
A sound characteristic conversion method characterized by performing a corresponding characteristic conversion based on decoration data given corresponding to label data. 7. A sound data segmented according to a predetermined segment and a label indicating the transmission content of the sound indicated by the sound data as a character or a symbol are associated with each segment of the sound data and segmented by a delimiter code. Sound to hold
A sound characteristic conversion device comprising: a label data holding unit; and a conversion unit that performs a corresponding characteristic conversion on sound data associated with the label data based on modification data given in correspondence with the label data. 8. For sound data classified according to a predetermined division, a label indicating a transmission content of a sound indicated by the sound data as a character or a symbol is associated with each division of the sound data, and separated by a delimiter code. A sound characteristic conversion method characterized in that sound data associated with label data is classified and the corresponding characteristic conversion is performed based on modification data given corresponding to the label data. 9. A system for transmitting sound data from a transmitting device to a receiving device, comprising a transmitting device and a receiving device communicable via a communication path, wherein the transmitting device comprises: Data input means for inputting label data divided into a plurality of divisions by a delimiter and modification data associated with the divisions; and communication means for transmitting the label data and the modification data to the receiving apparatus via a communication path. The communication means receives label data and modification data from the transmission-side device; the standard sound data generation means generates standard sound data based on the label data; and Conversion means for performing sound characteristic conversion on the standard sound data according to the classification of the label data associated with the decoration data, and generating sound characteristic conversion data. A sound transmission system comprising: 10. A method for transmitting sound data from a transmitting side to a receiving side via a communication path, wherein the transmitting side associates the label data divided into a plurality of sections by a delimiter code and the sections. Input the qualified data provided, and transmit the label data and the qualified data to the receiving side via the communication channel.The receiving side receives the label data and the qualified data from the transmitting device, and performs standardization based on the label data. Generating sound data, converting sound characteristics of the standard sound data based on the classification of the label data associated with the decoration data, based on the decoration data, and generating sound characteristic conversion data. Sound transmission method. 11. A sound data input means for inputting sound data, wherein the sound volume represented by the sound data is provided with two threshold values relating to the sound volume, and the sound volume continues for a predetermined time or more. The sound data segmentation is performed such that the segmentation starts when the threshold value exceeds one threshold value and ends when the loudness falls below another threshold value continuously for a predetermined time or more. Label data input means for associating a label indicating a transmission content of the sound indicated by the sound data as a character or a symbol with the sound data, and inputting label data with a delimiter sign at a position corresponding to the division of the sound data, A sound and / or sound generator comprising: label data classifying means for classifying label data on the basis of a delimiter code; and correspondence forming means for associating the classified sound data and the classified label data with each other. Bell associating device. 12. A sound data input means for inputting sound data, a label data input means for inputting as a label data a label indicating a transmission content of a sound indicated by the sound data as a character or a symbol, and inputting the label as label data. Detailed correspondence forming means for classifying sound data in association with each label based on the average duration of the label obtained based on the average duration of each syllable and the actual duration of the input sound data A sound / label associating device comprising: 13. The sound / label associating apparatus according to claim 11, wherein the label data and the sound data associated by the association forming means are associated with the label data obtained based on the average duration of each syllable of the label data. Detailed correspondence forming means for classifying sound data in association with each label based on the average duration and the duration of the sound data. 14. A sound display unit for visually displaying the nature of the sound represented by the sound data, and a label display unit for displaying a label represented by the label data. Two loudness thresholds are set for the loudness of the sound, and when the loudness of the sound exceeds one threshold continuously for a predetermined time or more, the segmentation is started. A sound / label correspondence, wherein a delimiter mark indicating a sound delimiter at the time of division is displayed so that the division is terminated when continuously falling below another threshold for a predetermined time or more. Mounting device. 15. When the sound volume represented by the sound data is provided with two threshold values relating to the sound volume, and the sound volume continuously exceeds one threshold value for a predetermined time or more. The sound data is divided so that the sound data is divided at the beginning of the sound data, and when the sound volume continuously falls below another threshold for a predetermined time or more, the sound data is divided. Delimiter is attached to the position, label data indicating the transmission content of the sound indicated by the sound data as characters or symbols, label data is divided based on the delimiter, the classified sound data and the classified label A sound / label associating method characterized by associating data with each other. 16. A method for associating sound data with label label data indicating, as characters or symbols, the transmission contents of the sound indicated by the sound data, wherein an average duration of each syllable of the label is prepared in advance. In advance, the average duration of the label is determined based on the average duration of each syllable of the label, and the sound data is calculated based on the determined average duration of each label and the actual duration of the sound data. A sound / label associating method, wherein the sound and the label are classified according to the label. 17. The sound / label associating method according to claim 15, wherein an average duration of each syllable of the label is prepared in advance, and the average duration of each syllable is related to the associated label data and sound data. A sound / label associating method, wherein sound data is classified in association with each label based on an average duration time of label data and a duration time of sound data obtained based on time. 18. A sound display unit for visually displaying a property of a sound represented by sound data, and a label display unit for displaying a label represented by label data. Sound loudness 2
The threshold is set, and when the loudness of the sound exceeds a threshold for more than a predetermined time continuously, the start of the segmentation is made. A display method for associating a sound with a label, wherein a break mark indicating a break of a sound at the time of division is displayed so that the division is terminated when the sound level falls below the threshold. 19. A storage medium in which a program for realizing a device or method according to claim 1 using a computer is described.