JP2006222568A - Narration support device, and document editing method and program thereof - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a narration support device capable of adjusting in detail and confirming reading timing. <P>SOLUTION: In the narration support device 9, a voice processing program 12 performs correction editing for adjusting reading timing of a voice waveform of a voice file 32 wherein a reading voice of a narration document displayed on a display 5 is recorded according to a command entered from a control input section 4 and a voice/character processing program 13 copies the display position of a mark M indicating the reading timing displayed in a narration document from the voice file 32 corrected and edited by a voice recognizing function and information on reading time length to a document file 31 of the narration document giving guide display of the reading timing. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a narration support apparatus used for reading a narration document for television broadcasting, and a document editing method and program thereof.

  In a television broadcast news program or the like, an announcer reads out a manuscript, but the manuscript must be read out in accordance with an allowable time. In preparation for adjusting the reading speed, the announcer scans the original, but the actual time adjustment depends on the skill level of the announcer. Therefore, there is an original reading device (hereinafter referred to as a narration support device) similar to a music karaoke that calculates the number of original characters and time, adjusts the reading speed, and displays the original characters on the screen at a timing in accordance with the speed (hereinafter referred to as a narration support device). For example, Patent Document 1).

However, this method is sufficient to keep the reading time of the entire manuscript within a predetermined time, but the program provider displayed on the TV screen is read out in synchronism with the movement of the screen, landscapes, news programs, etc. There was a problem that was not suitable for adjusting the timing of details such as video explanations and narrations. Also, there was no rehearsal function to check the state of reading-out timing, etc., and a correction function.
JP 7-67034 A (page 3, FIG. 2)

  Conventional document readers are suitable for reading program providers displayed on a TV screen in synchronization with the screen, or adjusting detailed timing such as explaining the scenes or explaining video narration in news programs. There was no problem.

  The present invention has been made to solve the above problems, and an object of the present invention is to provide a narration support apparatus, a document editing method, and a program thereof that can be confirmed along with detailed adjustment of reading timing.

  In order to achieve the above object, a narration support device according to the present invention is a narration support device for displaying a narration document read out by a television broadcast, and a document file in which information of the narration document is written, and a voice read out from the narration document A database for storing a recorded audio file, a control input means for inputting a command for editing the audio file, and an audio for an announcer to read the narration original. An audio input unit that converts the audio file stored in response to the command into an audio output unit that reproduces and outputs the audio file, and the voice waveform of the narration document or the voice file of the narration document Refer to reading timing A display that displays a screen together with a time scale, a timer, a syllable that reads out the narration document read out by the input command, or a sentence for each unit of the sentence, and each sentence and each sentence A text processing program for performing a process of editing the mark indicating the read-out timing on the screen in accordance with the time scale and storing the mark in the database, reading the voice file from the database by the command, and displaying the screen on the display A control processing unit comprising: a speech processing program for performing processing of editing the reading timing of each sentence and adjusting the time length of each sentence to be read and adjusting the waveform to be stored in the database It is characterized by .

  The narration support apparatus of the present invention also provides a narration support apparatus for editing a narration support apparatus for displaying a narration document read out by a television broadcast, wherein the narration support apparatus includes a document file in which information of the narration document is written. A database for storing a voice file in which the voice of the narration manuscript is recorded; a control input means for inputting the manuscript file or a command for editing the voice file; and an announcer A voice input unit that converts a voice to be read into a digital voice signal of the recorded voice, a voice output unit that plays back and outputs the stored voice file according to the command, and the narration document or the voice file The sound waveform is A display that displays a screen together with a time scale for referring to the reading timing of the reading of the original document, and a control processing unit, wherein the control processing unit reads a syllable or phrase that reads out the narrated original read out by the input command. A command for editing the voice file is input, edited into the narration manuscript displayed on the screen with each mark from the start of the narration and each reading timing and displayed on the screen. The voice file is read from the database by referring to a built-in timer, the voice waveform of the voice file is displayed on the display, and the reading timing of each sentence written to the voice file is read. Processing to adjust the display of the reading timing of the narration document displayed on the screen according to the reading timing of each sentence edited by editing the time length of each sentence and storing it in the database It is characterized by performing.

  Further, the program of the narration support device of the present invention includes a database that stores a document file in which information of a narration document is written, a sound file in which a sound that reads out the narration document is recorded, the document file, Based on the command from the control input means for editing the audio file, the narration manuscript is read as a syllable or a sentence for each unit of the narration, and each sentence from the start of the narration and each reading timing are displayed together with a mark. When the command for editing the voice file is input, the voice file is read from the database, and the voice waveform of the voice file is displayed on the display. The above is displayed The voice waveform is adjusted and the reading timing of each sentence written in the voice file and the time length for reading each sentence are edited by referring to a timer and stored in the database, and the voice waveform is adjusted. A program for a narration support apparatus that performs a process of adjusting the display of the reading timing of the narration document displayed on the screen according to the reading timing of each sentence edited in this way.

  According to the present invention, it is possible to provide a narration support device capable of confirming, together with detailed adjustment of the timing of a reading guide for a narration document, by recording and editing the sound of the narration document being read down.

  Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing a functional configuration of a narration support apparatus according to an embodiment of the present invention.
In FIG. 1, a narration support device 9 includes a control processing unit (hereinafter referred to as CPU) 1, a timer 2, a database 3, a control input unit 4 including a keyboard, a mouse, and the like connected to each other by an internal bus or the like. A display 5, an audio input unit 6, an audio output unit 7, and a communication IF (interface) 8 are provided. The narration support device 9 may be an information terminal such as a personal computer or a workstation as long as it has the above-described functional configuration.

  The CPU 1 reads the document file 31 stored in the database 3, the sentence processing program 11 for processing the audio file 32, the voice processing program 12, the document sentence displayed on the display 5, the timing mark, and the audio waveform of the audio file 32. Are provided with a voice / character processing program 13 for controlling the synchronization between and a work memory 14.

  The timer 2 operates as a reference clock for the original file 31 and the audio file 32 to operate in synchronism with respect to the reading timing of the original and the rising of the audio waveform, which will be described later.

  The database 3 stores a manuscript file 31 read by the announcer and a sound file 32 that records the sound of the announcer reading the manuscript.

  The control input unit 4 receives commands such as text input of a narration document, character display position, and processing editing of an audio file by an operator. Then, the data and commands are output to the internal bus via the internal bus, and further, the CPU 1 performs predetermined processing. The CPU 1 writes and saves the text edited and created by the text processing program 11 in the database 3 as a manuscript file 31.

  The display 5 displays the narration original read by the announcer and the audio waveform of the audio file 32 that records the original read by the announcer. The narration manuscript and audio file are displayed on each screen, and on these screens, the timing to read each narration sentence from the beginning of reading the narration manuscript, in other words, the elapsed time or the remaining time until the start of reading The time scale is displayed. Although the narration document and the audio waveform may be displayed on the same screen, the operation of the narration support apparatus will be described below by taking the case of displaying on each screen as an example.

  The audio input unit 6 outputs a digital audio signal obtained by A / D converting the audio input from the microphone 61 to the internal bus. The CPU 1 inputs a digital audio signal from the internal bus by the audio processing program 12, generates an audio file 32, writes and stores it in the database 3.

  The voice output unit 7 restores the voice file 32 read out by the voice processing program 12 of the CPU 1 to D / A conversion and voice in response to a command from the control input unit 4 and outputs the voice file 32 from the speaker 71.

  The communication IF 8 receives a start (queue) signal for changing the character display of the document in synchronization with the broadcast video screen.

FIG. 2 is an example of an image display surface of the display 5 on which a narration document read by the announcer is displayed.
In FIG. 2, the screen is marked with manuscript sentences b1 to bn edited by the text processing program 11 and the voice processing program 12, and the reading start timing (hereinafter abbreviated as reading timing (t1 to t3)). A screen button B for inputting commands for recording, reproduction, and editing is displayed on the left side of the screen. The screen button B is operated by pointing or clicking with the mouse or keyboard of the control input unit 4.

  The manuscript sentences b1, b2, and b3 respectively read “This program is xyz aimed at tomorrow,” “OO company,” and “The sponsors you see are sent” respectively. Read out at t3). Depending on the reading speed (reading timing), the character color is changed from “blue” indicating that it has not been read out to “orange” indicating that it has been read out.

  The time scale C is a time scale for notifying the announcer of the time until the start of reading of the displayed document text on the screen. In FIG. 2, the color of the bar cb is, for example, rightward as time elapses from the narration start time. From left to right, it changes from “blue” unread to “orange” to read. The order of the timing of reading out the original text in accordance with this bar is indicated by marks M from “Δ1” to “Δ3”. For example, under the “Δ1” position, “This program shows tomorrow. The target xyz is arranged.

  These changes in the character color and the bar cb are executed when the sentence processing program of the CPU 2 refers to the timer 2.

  Note that the display screen in FIG. 2 may be such that the elapsed time from the start of reading is a number, a clock part CL such as a clock hand, and a cue lamp Q that is lit when the reading start timing comes.

  The announcer reads these manuscript sentences b1 to bn while looking at the display 5, and at that time, the “REC” of the screen button is pressed by the mouse of the control input unit 4 as a “record” command, or When “REC” is input from the keyboard, the reading voice is written and stored in the database 3 as an audio file 32.

  The voice processing program 12 can read out the recorded voice file 32 and edit and correct the utterance timing. In addition, the voice / character processing program 13 performs a process of synchronizing the voice waveform utterance start timing of the voice file 32 with the mark M of the reading start timing displayed on the narration document by using the voice recognition function.

FIG. 3 is an example of a sound waveform displayed on the display 5 when the sound file 32 is edited by the sound processing program 12.
In FIG. 3, the waveform editing screen displays voice waveforms w1 to w3, and the voice waveforms w1 to w3 correspond to the original sentences b1 to b3, respectively. The speech waveform is a vibration waveform at the time of utterance and has a shape of a group for each syllable in sentence units. A time scale TS is displayed on the voice waveform. Although this time scale TS is synchronized with the time scale C in FIG. 2, a time scale that is slightly smaller than the time scale C is written so that the timing adjustment of the audio waveform can be finely adjusted.

  Further, the waveform editing screen is provided with an editing button E on the screen, which is used for editing the utterance timing and generation time length of the later-described speech waveform.

  For example, the voice processing program 12 extends or shortens the free time between the voice waveforms w1 and w2, in other words, a timing setting function for moving the sound generation start timing, and a time for lengthening or shortening the time when the voice is uttered. It has a compander function.

  In the conventional recorder, when the recording time length and the playback time length are different, the sound pitch corresponding to the sound frequency changes in proportion to the speed. However, the time compander function compares the playback time length and the recording time length, in other words, the playback speed and the recording speed, and if the occurrence time is each article, the audio information is thinned out or inserted when it becomes longer By using the encoded speech correction technique, the same effect as changing the reading time can be created by extending or shortening the utterance time without changing the speech pitch.

  Therefore, the narration support apparatus 9 according to the embodiment of the present invention records the original read once by the announcer and edits the audio file, thereby correcting the reading timing and the reading time length. Can be created. Then, the corrected file can be used as a document reading voice.

  Further, the voice / character processing program 13 detects the start timing of the voice waveform set in the voice file 32 by the voice recognition function, and corrects the mark M indicating the read-out timing of the narration document to the corrected document file 31. It is possible.

FIG. 4 is a flowchart for explaining the operation procedure of the narration support device 9 in the embodiment of the present invention.
Hereinafter, with reference to FIGS. 1-4, the process of each structure of the narration assistance apparatus 9 and an operation | movement procedure are demonstrated.

  In FIG. 1, an operator (manuscript creator) writes a draft of a narration draft created by inputting a text from the control input unit 4 as a manuscript file 31, for example, named “Advertisement 1-L” in the database 3 ( Step s1) in FIG.

  In this manuscript, manuscript sentences b1 to bn are described which are divided for each sentence to be read out. The reading timing of each of the manuscript sentences b1 to bn may be set in advance by the manuscript creator, but here, the sound file created by recording the manuscript read by the announcer is edited to set the sound and timing. The procedure for creating a document will be described by taking the case of creating a document as an example.

  The announcer operates the control input unit 4 to read the document file 31 named “advertisement 1” from the database 3 and display it on the display 5 (step s2).

FIG. 5 is a screen display example when the original file 31 of “advertisement 1-L” before correction is displayed on the display 5.
In FIG. 5, the bar cb of the time scale C above the original sentences b1 and b2 is a number from “Δ1” to “Δ3” indicating the reading order and the reading timing set by the sentence processing program 11. Is displayed. At the same time as the start of the video screen, the announcer inputs a screen start “REC” command from the control input unit 4 (step s 3), and starts recording the document draft. After the start of recording, the voice read out from the original is input from the microphone 61 to the voice input unit 6 (step s4).

  The voice input unit 6 converts the voice input from the microphone 61 into digital voice and outputs it to the internal bus (step s5), and the voice processing program 12 of the CPU 1 sequentially writes the digital voice into the voice file 32 of the database 3 (step s5). s6). When the end of the sentence has been read, when a recording end command, for example, “END” is input from the control input unit 4 (step s7), the recording is ended, and the audio file 32 is, for example, “advertisement 1-V”. Is given and written in the database 3 and stored (step s8).

  Subsequently, when a playback command, for example, “PLY”, for example, is input from the control input unit 4 in order to audition the reading result from the announcer (step s9), the CPU 1 reads the audio file 32 of “advertisement 1-V”. Is read from the database 3 and input to the audio output unit 7, and the audio reproduced by the audio output unit 7 is output from the speaker 71 (step s10).

  As a result of the trial announcement, the announcer determines that it is necessary to correct the reading to increase the interval between the original sentence b1 and the original sentence b2 and to increase the reading time length of “xyz aiming for tomorrow” of the original sentence b2. (Step s11 is Yes).

  When the announcer inputs a correction edit command “EDT” of the audio file 32 from the control input unit 4 for editing to correct the timing (step s12), an audio file with the name of the advertisement 1-V ″ is displayed on the display 5. 32 voice waveforms are displayed (step s13).

FIG. 6 is a screen when the audio waveform of the audio file 32 of the “advertisement 1-V” before correction is displayed on the display 5.
In FIG. 6, the recording start timing is indicated by a mark “Δ0”, and thereafter, the audio waveforms w1 to w3 are displayed in which the original sentences b1 to b3 are read out. For example, the speech processing program 12 observes the envelope of the speech waveform, detects its rise timing, and outputs a detection signal to the internal bus.

  When the voice / document processing program 13 receives this detection signal, it reads the reading timing (t1) time with reference to the timer 2 at the rising portion of the first voice waveform. Since the rising timing of the envelope corresponds to the reading start timing of the original text, the reading start timing of “Δ1” of the original text b1 is compared with the original file 31 of the “advertisement 1-L”. The mark M is assigned to the position of the time scale TS and the reading timing (t1) of the bar cb in FIG.

  The same processing is performed for other manuscript sentences, and the marks “Δ1” to “Δ3” and the audio waveforms w1 to w3 are displayed in correspondence with each other on the screen of FIG.

  The reading timings (t1 to t3) for the original sentences b1 to b3 are 0.7 seconds, 2.5 seconds, and 4.7 seconds, respectively, after the start of recording, when compared with the time scale TS. Further, for example, it can be read from the display of the voice waveform that the reading time length of the document sentence b2 is about 0.8 seconds.

  In FIG. 6, the mark M of “Δ0” is shown at the recording start timing, and the reading timing (t1) of the first document sentence b1 becomes the mark M of “Δ1”. It is also possible to adjust the (time) so as to be synchronized with the cue signal for starting the video. In this case, the video start queue is input to the narration support device 9 by the communication IF 8 and the “M” mark “M” is set at the read start timing (t1) of the original text b1 with the “M” mark “M” as a reference. Is adjusted prior to setting the reading timing (t2) of the document sentence b2.

  For example, when reading is started 2 seconds after the start of the video, each mark M and each audio waveform is dragged by dragging the mark M of “Δ1” to the position of 2 seconds on the time scale TS with the mouse of the control input unit 4. The whole is shifted to the left, that is, adjusted to be delayed by 1.3 seconds. Then, the reading timing (t2) of the document sentence b2 is 2.5 + 1.3 = 3.8 (seconds), and the reading timing (t3) of the document sentence b3 is 6.0 seconds.

  Now, the announcer operates the control input unit 4 to lengthen the interval between the original sentences b1 and b2, places the pointer on the mark M of “Δ2”, and refers to the time scale TS to the left side of the screen. The utterance timing is delayed by 200 msec by dragging (step s14).

  Since the voice / character processing program 13 stores each mark M and the reading start position of each of the original sentences b1 to b3 in association with each other, the rising position of the voice waveform recorded in accordance with the movement of the mark M, that is, the reading-out is read out. Timing (t1) also moves.

  For example, “D2” may be input as a command for delaying the reading timing. In this case, “D” in “D2” means delaying, and “2” means one step in the number of unit steps of delay, and here means 100 ms delay. As a result of this processing, the reading timing (t2) of “OO Company” in the original sentence b2 is delayed by 200 milliseconds corresponding to two steps (if the blank time is shortened on the contrary, as a command to advance, for example, “A2”) Is entered.) As a result, the reading timing (t2) of the original sentence b2 is 4.0 seconds, but the reading timing (t3) of the original sentence 3 is held at 6.0 seconds.

  Further, the voice / character processing program 13 sets the marks M of “Δ1” and “Δ3” and their marks so that the rising timings of the voice waveforms coincide with the reading timings (t1) and (t3) of the voice waveforms w1 and w3. Processing to move to the rising position is performed (step 14-1).

Now, for example, the announcer drags on the speech waveform w2 of the read-out portion of “OO Company” of the document sentence 2 displayed on the display 5 in order to make correction to increase the read-out time length (step s15). Then, the end point is indicated and an extension section is set. Then, "EXP" of the edit button E on the screen is clicked once as an extension command (step s16). Each time this click is performed, for example, 10% time is set longer. (If it is shortened, click “COM” of the edit button E on the screen once.)
As another method of increasing the reading time length, for example, “E10” may be input from the control input unit 4. “E” of “E10” means extension of the reading time length, and “10” means that the reading time length is increased by 10%. As a result, the speech processing program 12 of the CPU 1 extends the readout time length to 0.9 seconds by inserting a correction bit into the speech signal of the readout portion of “xyz aiming for tomorrow” (step s17).

  Then, the announcer inputs an “END” command to end editing from the control input unit 4, and the CPU 1 overwrites and stores the audio file 32 of “advertisement 1-V” with the modified content, and temporarily ends the audio file editing (step) s18).

  When this “END” command is input, the narration document is displayed on the display 5 again. The voice / character processing program 13 of the CPU 1 reads the timing (t1) of the mark M of “Δ1” corrected from the overwritten voice file 32 from the time scale TS. Then, the corrected reading timing (t1) time and the mark M of “Δ1” of the cue timing are written in the bar cb in the original file 31 of “advertisement 1-L”. Then, the reading time length of the document sentence b2 is read from the voice waveform w1 and written into the document file 31 of “advertisement 1-L” as the attached information of the document sentence b2.

  Subsequently, for the reading timings (t2) and (t3), the time information, the correction of the marks M of “Δ2” and “Δ3”, the writing of the reading time are performed, and the correction contents of the audio file 32 are changed. Copied (step s19). Then, the document file 31 of “advertisement 1-L” in the database 3 is also overwritten and stored (step s20). As a result, the screen display of each mark M on the display 5 is as shown in FIG.

  Here, when the announcer inputs the “PLY” command again to confirm the correction contents, the audio output unit 7 outputs the audio of the overwritten “advertisement 1-V” audio file from the speaker 71. Also, the sentence processing program 11 compares the edited document file 31 with the timer 2 and starts the control of the bar cb and the character eclipse as well as the start of recording and reproduction. On the screen of the display 5 in FIG. As time elapses from the start of output, the bar cb and the text of the original text are displayed so that the color of the read-out portion changes.

  For example, when two commands “PLY” and “EDT” are simultaneously input, the document display screen and the audio waveform display screen of FIGS. 5 and 6 may be displayed vertically on the same screen. When the audio file 32 is read and recorded and reproduced, the text processing program 11, the audio program 12 or the audio text processing program 13 refers to the timer 2 on the bar cb and time scale, and starts reproduction ( Alternatively, the time from the start of reproduction may be indicated together with the cue start). This time-lapse display method is such that the aforementioned color change or vertical line moves. Also, the time course may be displayed as a number or a clock hand.

  If the announcer further changes the reading timing or reading time length of the portion of the document sentence b2, the above-described step s14 or step s15 and the subsequent steps are repeated for recorrection. Even in this case, the read-out timing of “OO Company” of the original sentence b3 is retained as long as the read-out timing stored before the re-correction is stored unless the correction command for the original sentence b3 is input. The voice / character processing program 13 performs correction processing so as to be limited only to the part concerned.

  When the reading timing (t1) is set in advance by writing and storing the position of “Δ1” in the original file of “Advertisement 1-L”, the lead time from the start of the video to the start of reading is memorized from the beginning. can do.

  Further, if necessary, the announcer performs the same procedure as the original sentence b2 even if attached to the original sentences b1 and b3, so that each processing program becomes the original file 31 of the “advertisement 1-L” and the “advertisement 1−1”. The V "audio file 32 is edited.

  Further, when the final narration original is displayed and the original is read out, if the reproduced sound of the audio file 32 is output as a small guidance voice from the speaker 71, the voice is read out from the narration original displayed on the display. It is possible to increase the guidance effect because it is output in synchronization with the output. Further, the finally edited audio file 32 of “advertisement 1-V” may be used as it is as the audio material for broadcasting.

  In this case, the command for starting the narration is, for example, that “PLY” is input from the control input unit 4 or is based on a video start cue signal from the communication IF 8. Is output. Further, instead of being output from the speaker 71, the reproduced audio may be output as a broadcast audio material via the communication interface 8 as a digital audio signal or to an external device (not shown) as an audio signal.

  In the second embodiment, the configuration of the narration support device 9 is configured by omitting the voice / character processing program 13 from the configuration of the first embodiment shown in FIG.

  Accordingly, in the second embodiment, the document file 31 and the voice file 32 are created in the same manner as in the first embodiment, but the correspondence between the voice waveform and the mark M is automatically achieved by omitting the voice / character processing program 13. Will not be done. As a result, the operations in steps s14-1, s19 and s20 in FIG. 4, that is, the matching operation for associating the speech waveform in FIG. 6 with the mark M to be matched therewith, and display of the mark M on the speech waveform screen from the work screen. The operation of reading the position and resetting the mark M in accordance with the time scale C of the narration original is performed by the operator or the announcer operating the control input unit 4 to input a command and setting the necessary parameters in the text processing program 11. There must be. However, since the sound file editing and playback sound by the sound processing program 12 is the same as in the first embodiment, the sound of the modified and changed sound file can be used as a model or the like as in the first embodiment.

  As described above, according to the present invention, it is possible to provide a narration support device capable of confirming, along with the detailed adjustment of the timing of a reading guide for a narration document, by recording and editing the sound of the narration document read down. .

The block diagram which shows the function structure of the narration assistance apparatus which concerns on Example 1 of this invention. A screen display example of a display showing a narration manuscript. An example of the screen display of the voice waveform read out from the narration manuscript. The flowchart explaining the operation | movement procedure of the narration assistance apparatus in Example 1 of this invention. An example of a screen display of a document before correction of reading timing. An example of the screen display of the audio waveform of the manuscript read by the announcer.

Explanation of symbols

1 Control processing unit (CPU)
Chapter 11 Processing Program 12 Voice Processing Program 13 Voice / Character Processing Program 14 Work Memory 2 Timer 3 Database Storage Unit 31 Original File 32 Audio File 4 Control Input Unit 5 Display 6 Audio Input Unit 7 Audio Output Unit 8 Communication IF (Interface)
9 Narration support device C, TS Time scale M mark

Claims (16)

In a narration support device that displays a narration manuscript read out by a television broadcast,
A database for storing a manuscript file in which information of a narration manuscript is written, and a sound file in which a sound of reading out the narration manuscript is recorded;
Control input means for inputting a command for editing the original file and the audio file;
An audio input unit for converting an audio from which the announcer reads the narration document into a digital audio signal of the recorded audio;
An audio output unit that reproduces and outputs the stored audio file as audio by the command;
A display for displaying the voice waveform of the narration document or the voice file together with a time scale for referring to the reading timing of the narration document;
A syllable to be read out by the timer and the narration document read out by the input command, or a sentence for each phrase unit, and a mark indicating the reading timing of each sentence and each sentence are adjusted to the time scale. A sentence processing program for performing processing for screen display editing and storing in the database, and reading each voice file from the database by the command, and adjusting the voice waveform displayed on the display by the command. A narration support apparatus comprising: a speech processing program that performs a process of editing the time to read out and the time length of each sentence read out and storing the edited time in the database.   The narration support apparatus according to claim 1, wherein each of the programs refers to the timer to display a lapse of time from the start of the narration on the display.   3. The time lapse is displayed by at least one of a color change of a bar displayed on the time scale and a color change of a character read out in each sentence. Narration support device. In a narration support device that displays a narration manuscript read out by a television broadcast,
A database for storing a manuscript file in which information of a narration manuscript is written, and a sound file in which a sound of reading out the narration manuscript is recorded;
Control input means for inputting a command for editing the original file and the audio file;
An audio input unit for converting an audio from which the announcer reads the narration document into a digital audio signal of the recorded audio;
An audio output unit that reproduces and outputs the stored audio file as audio by the command;
A display for displaying the voice waveform of the narration document or the voice file together with a time scale for referring to the reading timing of the narration document;
A syllable to be read out by the timer and the narration document read out by the input command, or a sentence for each phrase unit, and a mark indicating the reading timing of each sentence and each sentence are adjusted to the time scale. A sentence processing program for performing processing for screen display editing and storing in the database, and reading each voice file from the database by the command, and adjusting the voice waveform displayed on the display by the command. A speech processing program for performing processing for editing the time length of each sentence to be read out and storing it in the database, and the display timing for each sentence edited by adjusting the speech waveform Read-out timing of each sentence Narration support apparatus characterized by comprising a control unit and a voice-character processing program for performing processing to tailor the mark indicating grayed. The sentence processing program and the speech processing program refer to the timer, display the passage of time from the start of narration in accordance with the time scale, and display the speech waveform on the screen for displaying the speech waveform. Corresponding to each mark indicating the same, the marks are displayed on the screen in the same manner,
The voice / character processing program reads the reading timing of each sentence by voice recognition that detects the rising of the voice waveform from the voice file, and the reading timing is read with reference to the timer. The narration support apparatus according to claim 4, wherein each mark is displayed on the screen in the order at a time corresponding to a reading-out timing.   6. The narration support apparatus according to claim 5, wherein each of the programs refers to the timer to display a lapse of time from the start of the narration on the display.   The time elapsed is displayed by at least one of a color change of a bar displayed on the time scale and a color change of a character read out in each sentence. Narration support device.   When a command for starting narration is input from the control input unit, the audio output unit reads out the edited and stored audio file from the database, and generates reproduced audio as a guide audio for reading out the narration document. The narration support apparatus according to claim 1, wherein the narration support apparatus outputs a broadcast sound material in place of an announcer's read-out sound. The narration support device further includes communication interface means,
8. The narration support apparatus according to claim 6, wherein a cue signal of a signal for starting a video corresponding to narration is input as a command for starting the narration through the communication interface means. In a method for editing a narration support device for displaying a narration manuscript read out by a television broadcast,
The narration support device edits the original file or the audio file, a database that stores a manuscript file in which information of the narration manuscript is written, an audio file in which a voice that reads out the narration manuscript is recorded, and Control input means for inputting a command, a voice input unit for converting a voice that the announcer reads the narration document into a digital voice signal of the recorded voice, and playing back the voice file stored by the command as voice An audio output unit for outputting, a display for displaying the audio waveform of the narration document or the audio file together with a time scale for referring to the reading timing of the narration document, and a control processing unit,
The control processing unit
The narration manuscript to be read out by the input command is changed to a syllable to be read out or a sentence for each unit of the syllable, and each sentence from the start of the narration and each reading timing are edited into the narration manuscript displayed on the screen together with marks. Stored in the database
When a command for editing the voice file is input, the voice file is read from the database by referring to a built-in timer, and the voice waveform of the voice file is displayed on the display and written to the voice file. The reading timing of each sentence and the time length of each sentence read out are edited and stored in the database, and the narration displayed on the screen according to the reading timing of each sentence edited by adjusting the speech waveform A document editing method for a narration support apparatus, characterized in that a process for adjusting a display of a document reading timing is performed. The control processing unit
With reference to the timer, the passage of time from the start of narration is displayed according to the time scale, and the same mark corresponding to each mark indicating the reading timing of each sentence on the screen displaying the speech waveform Are sequentially displayed on the screen, and the reading timing of each sentence is read by voice recognition that detects the rising of the voice waveform from the voice file, and each mark is set to the time at a time corresponding to the read reading timing. The narration support apparatus document editing method according to claim 10, wherein the screen is displayed in accordance with a scale.   12. The time lapse is displayed by at least one of a color change of a bar displayed on the time scale and a color change of a character read out in each sentence. Manuscript editing method of narration support device. From a database that stores a document file in which information of a narration document is written, a sound file in which a voice that reads out the narration document is recorded, the document file, and a control input means for editing the sound file Based on the command
A syllable to read the narration manuscript, or a sentence per phrase unit, and edit each narration manuscript from the start of the narration and the respective reading timing to the narration manuscript displayed on the screen together with marks, and store in the database
When a command for editing the audio file is input, the audio file is read from the database, the audio waveform of the audio file is displayed on the display, and the displayed audio waveform is adjusted to the audio file. The read-out timing of each written sentence and the length of time during which each sentence is read out are edited by referring to a timer and stored in the database, and the read-out of each sentence edited by adjusting the speech waveform A program for a narration support apparatus that performs processing for adjusting display of a reading timing of a narration document displayed on the screen according to timing. The program is
With reference to the timer, the passage of time from the start of narration is displayed according to the time scale, and the same mark corresponding to each mark indicating the reading timing of each sentence on the screen displaying the speech waveform Are sequentially displayed on the screen, and the reading timing of each sentence is read by voice recognition that detects the rising of the voice waveform from the voice file, and each mark is set to the time at a time corresponding to the read reading timing. 14. The program for a narration support apparatus according to claim 13, wherein the narration support apparatus displays the screen in accordance with a scale.   The program refers to the timer according to at least one of a color change of a bar displayed on the time scale and a color change of a character to be read out of each sentence. The narration support apparatus program according to claim 14, wherein the narration support apparatus program is displayed.   The narration support apparatus program according to claim 13, wherein the program edits the length of time to be read out without changing the pitch of the voice.

Images