JP2015206928A - Voice processor, voice processing program, and voice processing method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To suppress the fluctuation of the level of a recorded voice in the case of reproduction processing of a recorded voice, and to improve the operability of a voice processor without generating the volume adjusting operation of the voice processor in the case of reproduction processing of recorded data.SOLUTION: A voice processor 2 having a recording function includes voice input means 4, voice processing means 8, and storage means 10. The voice input means fetches voice having different levels, and generates recorded data. The voice processing means divides the recorded data generated by the voice input means into a plurality of frames for each group of voice, and determines the level of the voice included in the recorded data for each frame, and when the level of the recorded data is different from a reference level (output volume level), generates new recorded data by adjusting the level of the recorded data on the basis of the reference level. The storage means stores the recorded data generated by the voice processing means.

Description

  The technology of the present disclosure relates to a voice processing technology for recorded data.

  A voice processing device having a recording / playback function is used for communication such as voice mail for sending recorded voice messages, voice memos for recording voice messages, surrounding conversations and meetings, and the like. The As a recording / playback function of a voice processing device, for example, there is a function that performs playback processing by adjusting voice data in order to eliminate variations in volume of recorded voice data. Variations in sound volume and the like that occur in audio data are affected by, for example, the case where recording and playback are performed by different devices, and the surrounding environment where recording is performed.

  Regarding such recording data adjustment processing, it is known that the level of voice message is converted in accordance with the level information on the exchange side that detects the level information of the stored voice message and performs playback processing (for example, Patent Document 1). Also, it is known that when creating a voice mail, environment information at the time of recording is added, and noise removal processing is determined based on this environment information to edit the contents of the voice mail (for example, Patent Document 2). ).

JP-A-7-297828 JP 2004-236245 A

  By the way, in the recording process of voice by voice messages or voice memos, for example, when recording the utterances of multiple people, the level of the recorded voice is different depending on the voice level by each voice and the distance to the voice processing device to record. May vary greatly. For example, even when the voice of the same person is recorded, the voice level varies depending on the distance from the voice processing device to be recorded due to the action of the speaker, or the recording environment changes. There is a case. Further, when a plurality of recording data are created, the sound level differs depending on the sound collection environment and the state of the speaker for each recording data.

  Some audio processing devices having a recording function vary the microphone sound collection sensitivity (dynamic range) according to the level of the recorded sound level, for example, but it is not possible to suppress fluctuations in the recorded sound level. Therefore, if there is a change in the level of the uttered voice, the voice processing device records the voice with the fluctuation.

  When recorded data with variations in the sound level is reproduced, the user performs a volume adjustment operation in accordance with, for example, a change in volume due to a level change. That is, for example, when a low-level portion of sound is reproduced, the user decreases the sound volume and increases the sound volume because the reproduced sound becomes difficult to hear. In addition, when a sound with a high level is reproduced in a state where the sound volume is set to be high, the sound is output from the speaker at a high sound volume, and thus the user performs an operation for decreasing the sound volume. As described above, in the sound processing apparatus, there is a problem that a volume adjustment operation is always required when reproducing recorded data, and the user feels troublesome.

  Moreover, even if the audio processing device has a function of adjusting the level of recorded data to be reproduced in accordance with the recording environment and the recording level, for example, the level is uniformly adjusted for the recorded data that includes variations in the audio level. However, there is a problem that fluctuations in volume cannot be resolved.

  Therefore, an object of the technology of the present disclosure is to suppress fluctuations in the level of recorded voice during playback processing for recorded voice.

  Another object of the technology of the present disclosure is to improve the operability of the sound processing device without causing the sound volume adjusting operation of the sound processing device during the reproduction processing of the recorded data.

In order to achieve the above object, one aspect of the technology of the present disclosure is a voice processing device having a recording function, and includes a voice input unit, a voice processing unit, and a storage unit. The voice input means captures voices having different levels and generates recording data. The voice processing means divides the recording data generated by the voice input means into a plurality of frames for each group of voices, determines the level of voice included in the recording data for each frame, If the level is different from the reference level, new recording data in which the recording data level is adjusted is generated based on the reference level. The storage means stores the recording data generated by the voice processing means.

  According to the technique of the present disclosure, any of the following effects can be obtained.

  (1) When the recorded data is reproduced, fluctuations in the reproduction volume of the sound output from the speaker or the like are suppressed, and the recorded sound that is easy to hear during reproduction can be provided.

  (2) Since the fluctuation of the reproduction volume is suppressed, the volume adjustment operation of the audio processing apparatus that reproduces the recording data becomes unnecessary, and the convenience during audio reproduction can be improved.

It is a figure which shows an example of the audio processing apparatus which concerns on 1st Embodiment. It is a flowchart which shows an example of an audio | voice process. It is a figure which shows an example of the level adjustment of recording data. It is a figure which shows an example of the audio processing apparatus which concerns on 2nd Embodiment. It is a figure which shows an example of the division | segmentation state of recording data. It is a figure which shows an example of a volume level setting table. It is a figure which shows an example of a recording data table. It is a figure which shows the example of the production | generation state of the new recording data by volume level adjustment. It is a figure which shows the adjustment process of recording data. It is a flowchart which shows an example of an audio | voice process. It is a flowchart which shows an example of a file creation process. It is a flowchart which shows an example of a file level conversion process. It is a flowchart which shows an example of a sound recording data conversion process. It is a figure which shows the example of a state of the audio | voice process which concerns on 3rd Embodiment. It is a figure which shows the adjustment process of recording data. It is a figure which shows the other example of an audio | voice process. It is a figure which shows the other example of the adjustment process of recording data. It is a figure which shows the audio processing state example which concerns on other embodiment. It is a flowchart which shows an example of an audio | voice process.

  [First Embodiment]

  FIG. 1 shows an example of a speech processing apparatus according to the first embodiment. The voice processing device 2 is an example of a voice processing device according to the present disclosure.

  The voice processing device 2 is a device that records a voice including words spoken by a person such as a conversation and surrounding sounds and performs an editing process on the recorded voice. The voice processing apparatus 2 includes a voice input unit 4, a microphone 6, a voice processing unit 8, and a storage unit 10.

  The voice input means 4 is an example of a recording function that takes in external sound collected using a microphone 6 and generates recording data. Moreover, the voice input means 4 may be provided with a control function of the dynamic range (sound collecting ability) of the microphone 6, for example.

  For example, the microphone 6 collects sounds including sounds of different levels uttered by one person and voices of a plurality of persons having different levels of sound. This level of sound is, for example, the level of sound pressure or volume, and indicates the level of the audio signal captured by the microphone 6. The microphone 6 may be either one built in the audio processing device 2 or one that can be carried toward a speaker, or may be used by switching between a single or a plurality of microphones.

  The audio processing means 8 divides the captured recording data into frames for each set of predetermined audio and adjusts the volume level of each divided recording data file to generate a new recording data file. In the division of the recorded data, for example, the recorded data may be divided for each series of utterances, in addition to dividing the recorded data based on the person who speaks. In the level adjustment, the volume level is adjusted so as to eliminate the variation in the volume level between the recording data files divided into frames, and new recording data is generated. In the other voice processing means 8, for example, overall operation control of the voice processing device 2 may be performed.

  The storage means 10 is an example of means for storing new recording data for each frame generated by the sound processing means 8. For example, in addition to the memory built in the sound processing apparatus 2, a card-type IC memory that can be inserted and removed is used. , Semiconductor memory, magnetic disk, and the like. In addition, the storage means 10 may be a magnetic tape such as a cassette tape or DAT (Digital Audio Tape). The storage means 10 stores the newly generated recording data, and the audio processing device 2 reads the recording data file in response to a request for reproducing the recorded data.

  In the voice processing device 2, the voice processing means 8 suppresses the fluctuation of the voice level for each frame and adjusts the recorded data taken in by the voice input means 4 to an equivalent level.

  <About voice control processing>

  FIG. 2 shows an example of audio processing. The processing procedure and processing contents shown in FIG. 2 are examples, and the present invention is not limited to such a configuration.

  Audio processing is an example of the audio processing method or audio processing program of the present disclosure, and includes audio recording processing, recording data analysis and decomposition processing, recording data level conversion processing, and generation of new recording data.

  In the audio recording process, for example, when the audio process is started, the recorded data is stored in an audio buffer formed in the storage means 10, and a recording data file is generated (S1). Next, in parallel with the recording process or after the recording process is completed, the recording data is divided into frames for each unit of sound (S2). In this frame division, for example, a recording data file is divided based on a silent section of audio data as a reference and a timing when a series of utterances or speakers change.

  The sound level of the recording data file is determined for each divided frame (S3). The determined data is stored, for example, in the storage means 10 and tabulated. For example, as shown in FIG. 3, this table is divided into frames 12-1, 12-2, 12-3,... As a recording data table 12A. In each of the frames 12-1, 12-2, 12-3,..., Audio data is stored together with the detected audio level, for example.

  When the level of the recording data is determined, it is determined whether or not the level of the sound matches a reference level set in advance or set at a predetermined timing (S4). In this discrimination process, the difference between the recorded voice and the reference level is calculated. If there is no difference between the calculated levels as a recording data adjustment process (YES in S4), the sound level is maintained (S5).

  If there is a difference in level (NO in S4), the level is updated so that the audio level is equal to the reference level (S6). In the level update, the sound processing means 8 changes the level of the recording data table 12B of the storage means 10 as shown in FIG. The audio processing means 8 rewrites the table 12B so that the level of the table 12A is the same as or within a certain range as the reference level, and generates new recording data whose level is adjusted (S7).

  According to such a configuration, when the recorded data is reproduced, fluctuations in the reproduction volume of the sound output from the speaker or the like are suppressed, and a recorded sound that is easy to hear during reproduction can be provided. By suppressing the fluctuation in the reproduction volume, the volume adjustment operation of the audio processing apparatus that reproduces the recorded data becomes unnecessary, and the convenience during audio reproduction can be improved.

  [Second Embodiment]

  FIG. 4 shows an example of a speech processing apparatus according to the second embodiment. The configuration illustrated in FIG. 4 is an example, and the technology of the present disclosure is not limited to such a configuration.

  The voice processing device 20 has a recording function for the voice collected by the microphone 6 and a function for adjusting the recorded data. The voice processing device 20 includes, for example, a recording device, a PC (Personal Computer) capable of executing a voice processing program for a recording function and a recording program or recorded data, a portable information processing device, and the like.

  The audio processing device 20 includes, for example, a processor 22, a storage unit 24, an audio input / output unit 30, a volume level measurement unit 34, an audio amplification circuit 36, and a timer 37. The voice processing device 20 includes an operation unit 38, a display unit 40, a communication unit 42, and the like.

  The storage unit 24 is formed of, for example, a ROM (Read Only Memory) 26 that stores a program for operating the audio processing device 20 and recorded data, and a RAM (Random Access Memory) 28 that functions as an execution area of the program.

  The ROM 26 is a non-volatile memory that stores programs such as an OS (Operating System) of the sound processing device 20 and application programs for performing sound processing, as well as a recording data file, a recording data table 70 (FIG. 7), and the like. Is memorized. The ROM 26 may be formed of a magnetic disk such as an HDD (Hard Disk Drive), or a semiconductor memory such as a flash memory or an SSD (Solid State Drive).

  The RAM 28 forms an audio buffer as a work area for executing audio processing, and develops an audio processing program. The processor 22 is an arithmetic processing unit that executes a program, and executes voice processing using the program developed in the RAM 28.

  The voice input / output unit 30 is an example of a voice input / output unit of the present disclosure. The voice input / output unit 30 performs recording processing of voice input from the microphone 6, analysis of recorded data, division processing, and the like. Perform playback processing.

  The sound volume level measuring unit 34 is a part of the sound processing means and analyzes the sound volume level of the recording data captured by the sound input / output unit 30.

  The audio amplifying circuit 36 is an example of a circuit that forms part of the audio processing means of the present disclosure that adjusts the recording data so as to achieve a volume level set for each frame, and is formed by an amplifier, for example. . The audio amplification circuit 36 increases or decreases the volume level for the corresponding frame based on the instruction information in the generated recording data table 70.

  The communication unit 42 is an example of a unit that transmits / receives data to / from an external communication device using the communication antenna 44. The voice processing device 20 may transmit the recorded data and the recorded data table 70 subjected to voice processing, such as voice mail attached with a voice data file, to an external communication device via the communication unit 42. Further, the audio processing device 20 may receive a recording data file from, for example, an external communication device.

  <About imported recording data>

  FIG. 5 shows an example of a recording data division state. As shown in FIG. 5, for example, the recording data captured by the microphone 6 shows a high volume level when a person utters a voice in the surroundings, and a low volume level with respect to a voice uttered by silence or a long distance or a small voice. Show. Also, in the recorded data, for example, after a large volume level continues with time, a small volume level continues and a sound is collected. Such a group of voices indicates, for example, a state in which a person utters a series of words (conversation) or a state in which the person who has spoken has switched.

  When the recorded data with such a varying volume level is reproduced as it is, it is played back with a loud sound when the volume level is high, and a small sound when the volume level is low. The audio processing device 20 divides the frame for each unit of audio included in the recording data, and analyzes the audio level between the divided recording data files. The recording data is divided into frames based on, for example, fluctuations in the volume level, and in addition to the chunks in which the voice is emitted, the chunks of silence and voices below a certain level are also divided.

  In the audio processing device 20, the volume level is analyzed by the volume level measuring unit 34 for each recording data file. The volume level of audio varies even within one recorded data file. In the sound processing device 20, the volume level is set based on the maximum value or the average value of the volume levels in the recording data file generated by the volume level measurement unit 34.

  The volume level is set by using a volume level setting table 50 stored in the storage unit 24, for example, as shown in FIG. In this volume level setting table 50, for example, a predetermined threshold is set for the detected value of the volume level. For example, the threshold of the volume level can be divided evenly for the range of sound pressure included in the recorded data, or can be set by subdividing the range of sound pressure that tends to adjust the volume when the recorded data is played back. Good.

  In the audio processing device 20, a recording data table 70 in which volume level information and audio data are associated with each other is created for each frame as shown in FIG. 7 and stored in the ROM 26. The recording data table 70 is information obtained by analyzing the recording data, and also serves as instruction information when executing the reproduction processing of the recording data.

  In the recording data table 70, for example, for each frame, set volume level information is stored as the recording track A, and audio data including the uttered voice is stored as the recording track B. The recording data table 70 arranges frames according to the recorded time series.

  <About audio processing>

  FIG. 8 shows a new recording data generation state by adjusting the volume level.

  For example, as shown in FIG. 8A, the sound processing device 20 controls the setting when the sound volume level 3 is set in advance as the output sound volume level or is set by the user who performs the sound processing execution operation. Store as information. In the recorded data before conversion, as shown in FIG. 8B, for example, there are large and small variations in the volume level for each word group.

  In the recorded data, the utterance portions X1a and X3a are recorded at a high volume with respect to the set volume level 3, and the utterance portion X2a is recorded at a low volume. For example, as shown in FIG. 9A, the recording data is recorded in the recording data table 70 as volume level information 72, 74, 76 of frames 1, 3, 5 divided by a lump of words. 1 and volume level 4 are specified respectively.

  As shown in FIG. 8C, the audio amplifying circuit 36 reduces the volume of the utterance parts X1a and X3a recorded with a sound higher than the set volume level 3 in the recording data, and the utterance parts X1b and X3b. Convert to In addition, the sound amplification circuit 36 amplifies the volume of the utterance portion X2a having a volume smaller than the volume level 3 based on the volume level 3, and converts the voicing portion X2b into the utterance portion X2b. Then, the audio amplification circuit 36 generates new recording data with the volume level changed. Further, in the audio processing, the volume increase / decrease process is not performed for the frame specified as the silent section.

  In this audio processing, for example, as shown in FIG. 9B, the volume level 3 is set as the volume level information 72, 74, 76 in the recording data table 70, and a new recording whose volume is increased or decreased based on this setting information. Data is generated.

  <Specific examples of audio processing>

  10 to 13 are flowcharts illustrating an example of audio processing. The processing procedures and processing contents shown in FIGS. 10 to 13 are examples.

  This sound processing is an example of the sound processing method or the sound processing program of the present disclosure. The voice processing device 20 determines whether or not the recording function is started by, for example, pressing the operation unit 38 or operating the touch panel set on the display unit 40 as a voice input start determination (S11). The captured recording data is recorded, for example, in an audio buffer formed in the audio input / output unit 30 or the storage unit 24 (S12). In the audio buffer, for example, sound volume level waveform detection may be performed in association with time lapse information for recorded data.

  As the frame division process, the audio processing device 20 measures the volume level by using the volume level measuring unit 34 (S13), and discriminates the uttered part and the silent part. In the voice processing, for example, in order to extract a break portion of the voicing chunk, it is determined whether or not the sound volume level that is silent or less than the threshold has continued for 3 seconds or more as a predetermined time (S14). In the voice processing device 20, when silence has not continued for a predetermined time (NO in S14), the change in volume level is an inflection in the middle of a series of words, and the same person continues to speak and continue the conversation. Continue monitoring the volume level while recording.

  When the sound volume level is silent or less than the threshold value for a predetermined time (YES in S14), the sound processing device 20 creates a file in which the recorded sound is divided into frames (S15), and at the same time the sound volume level “0” for the soundless portion. "Is added (S16). In the frame division, for example, the recording data may be divided at the timing when the volume level becomes less than the threshold, that is, the timing when the predetermined time is started. Further, the audio processing device 20 creates recording data of a volume level “0” from the time when the frame is divided until the next audio input is detected.

  The voice processing device 20 determines whether or not the voice input is finished (S17). For example, when the user performs a recording function stop operation, the voice input is finished (YES in S17), and the stored voice is recorded. The process proceeds to file level conversion processing (S18). If the voice input has not ended (NO in S17), when the voice is detected again, the recording process and the detection of the silence or the volume level below the threshold are subsequently performed.

  Note that the level conversion process of the audio file is not limited to being executed after the recording process is completed. At the same time as the recording process, the audio processing device 20 may convert the level of the audio file for each frame of the recording data stored in the audio buffer.

  <About file creation processing>

  In the file creation process of S15, for example, as shown in FIG. 11, a recording data table 70 indicating the volume level associated with the audio data is created for the recording data file divided into frames. The recording data table 70 is set sound volume information when performing the reproduction processing of the recording data, and is created after the frame division processing so as not to include silence and audio less than a threshold value.

  For example, the sound processing device 20 reads the volume level setting table 50, compares the waveform information of the stored recording data with the threshold value of the volume level, and sets the set value of the volume level (S21). The set volume level is set in the recording data table 70 in association with the audio data (S22, S23). The created recording data table 70 (recording data file) is stored in the storage unit 24 (S24).

  <File level conversion process>

  In the file level conversion process, the sound level is set to a predetermined setting value for the created recording data file. In the audio processing device 20, for example, as shown in FIG. 12, the volume level is read for each created recording data file in the order according to the time frame of the entire recording data (S31). The voice processing device 20 reads the voice data according to the time frame (S32), converts the recorded data (S33), and creates new recorded data in which the converted volume level is associated with the voice data. In the recording data conversion process, the volume of the audio data is amplified or attenuated, and the set value of the volume level stored in the recording data table 70 is also converted.

  The file level conversion process is performed for each time frame, and is executed until the conversion process for all the frames is completed (S34). When all the frames are converted (YES in S34), the sound processing device 20 stores the converted recording data file in the storage unit 24 according to the time frame (S35).

  <Recording data conversion process>

  In the recording data conversion process, for example, as shown in FIG. 13, when the audio processing device 20 acquires the set output volume level (S41), the volume level of the recording data file is set to “3” for each frame of the recording data file. It is determined whether or not “0” (S42). In this determination, in addition to the case of complete silence (sound volume level “0”), it may be determined whether or not it is less than a threshold for determining silence.

  If the volume level of the recorded data is not “0” or greater than the threshold (NO in S42), it is determined whether or not the volume level of the recorded data file is smaller than the output volume level (S43). Alternatively, if the volume level of the recording data is “0” or less than the threshold value (YES in S42), the volume level (“0”) of the recording data file and the silent audio data are copied to the output file (S44). The output file indicates a new recording data table 70 formed in the storage unit 24, for example.

  If the volume level of the recording data file is smaller than the output volume level (YES in S43), the volume level of the recording data file is amplified to the output volume level (S45). Alternatively, when the volume level of the recording data file is larger than the output volume level (NO in S43), the volume level of the recording data file is attenuated to the output volume level (S46). In the recording data file conversion process, the difference between the volume level of the recording data and the set output volume level may be calculated, and the volume level may be amplified or attenuated based on this difference. Amplification or attenuation processing of the recording data file is executed by, for example, the audio amplification circuit 36.

  The audio processing device 20 stores the frame volume level of the output file for the amplified or attenuated recording data file (S47), and sets the volume level of the output file in the recording data table 70 (S48).

  According to such a configuration, when the recorded data is reproduced, fluctuations in the reproduction volume of the sound output from the speaker or the like are suppressed, and a recorded sound that is easy to hear during reproduction can be provided. By suppressing the fluctuation in the reproduction volume, the volume adjustment operation of the audio processing apparatus that reproduces the recorded data becomes unnecessary, and the convenience during audio reproduction can be improved. In addition, by monitoring the silent section and converting the volume level of only the portion including the audio data, it is possible to generate audio data that is easy to hear without amplification of noise during reproduction.

  [Third Embodiment]

  FIG. 14 shows an example of the state of audio processing according to the third embodiment. For example, as shown in FIG. 14A, the audio processing device 20 has an output volume level set as a volume level setting value. In the recording data before conversion, as shown in FIG. 14B, for example, there are large and small variations in the volume level for each word cluster.

  In the recorded data, the utterance portion X2a is recorded at a low volume with respect to the set output volume level. In this embodiment, a case is shown in which only a portion of the volume level of the recording data that is smaller than the output volume level is amplified.

  Therefore, as shown in FIG. 14C, the audio amplifying circuit 36 amplifies the utterance portion X2b having a volume smaller than the output volume level in the recording data based on the set output volume level, Convert to X2b. Further, as shown in FIG. 15A, the audio processing device 20 has volume level information 80A set for the frame 3 corresponding to the generation portion X2a of the recording data table 70 for the recording data file before conversion. Then, when the sound volume level of the generated portion X2a is converted, the sound processing device 20 sets the output sound volume level in the sound volume level information 80B as shown in B of FIG.

  <Other examples of audio processing>

  For example, as shown in FIG. 16B, the sound recording portions X1a and X3a are recorded at a louder volume than the set output volume level. For such recorded data, the audio processing device 20 attenuates the volume level for a frame having a volume level higher than the output volume level, for example, with respect to the set output volume level for the recorded data file.

  Therefore, in the audio amplifying circuit 36, as shown in FIG. 16C, the utterance portions X1a and X3a having a volume higher than the output volume level in the recording data are attenuated based on the output volume level, and the utterance portions X1b and X3b are attenuated. Convert to In addition, as shown in FIG. 17A, the audio processing device 20 sets the volume level for the frame 1 corresponding to the generation portion X1a of the recording data table 70 and the frame 5 corresponding to the utterance portion X3a of the recording data file before conversion. Information 82A and 84A are set. Then, when the sound volume levels of the generated portions X1a and X3a are converted, the sound processing device 20 sets the output sound volume level in the sound volume level information 82B and 84B as shown in B of FIG.

  In addition, about the audio | voice process with respect to recording data, the user of the audio | voice processing apparatus 20 may select amplification or attenuation arbitrarily, or may set beforehand. In addition, the audio processing device 20 selects, for example, audio processing, and performs sub-amplification processing or attenuation on the basis of a frame closer to the output sound volume level with respect to variations in sound volume of the utterance portions X1a, X2a, X3a. Processing may be set.

  According to such a configuration, the amount of variation in volume between frames can be reduced in the reproduction processing of recorded data, and the troublesome volume adjustment operation by the user can be eliminated. Also, by performing audio processing, variation in volume during playback processing can be suppressed, so that the user can easily understand the audio content between frames.

  [Other Embodiments]

  Examples of modifications described above are listed below.

  (1) In the above embodiment, the case where the audio processing device 2 performs audio processing on the recording data captured by the microphone 6 has been described, but the present invention is not limited to this. The audio processing device 2 may receive recording data from an external communication device, divide the frame of audio included in the recording data, and adjust the volume level for each frame.

  (2) In the above embodiment, the case has been shown in which the volume level is adjusted by dividing the frame for the sound included in one recording data recorded by the sound processing device 2 or received from the outside, but is not limited thereto. . The sound processing device 2 may analyze the variation of the sound volume level among a plurality of recorded data and perform the sound volume level adjusting process. As a result, when the recorded data is divided in advance by the recording device or the voice mail to which a plurality of recorded data is attached is reproduced continuously, it is possible to prevent the reproduction volume from varying for each recorded data.

  (3) In the above embodiment, the voice processing device 20 monitors the silent state having a volume less than the threshold while performing the voice recording process, and divides the frame when the silent state continues for a certain period of time. However, it is not limited to this. For example, the sound processing device 20 may divide the frame at regular intervals to create a recording data file, monitor the sound volume level for each frame, and perform sound processing to amplify or attenuate the sound volume level.

  As shown in FIG. 18A, the audio processing device 20 stores recording data having a variation in volume level over time. The audio processing device 20 creates a recording data file by cutting the frame every 3 seconds, for example, as the set time t for the recording data, and as shown in FIG. A recording data table 90 to be managed is created. In the recording data file, for example, audio data 1 included in the first utterance portion is recorded in frames 1 and 2 as audio data 1a and 1b. That is, since the audio data 1 is recorded across, for example, at least two frames, it has a length of 4 seconds or more. Next, in the recording data file, a silent section is recorded as frame 3, for example.

  In this way, the audio processing device 20 may create a recording data file based on the time of the timer 37 from the start of the recording process regardless of the silent period. Then, the sound processing device 20 may grasp the volume level of the created recording data file and may perform an amplification process or an attenuation process of the volume level.

  In the voice processing by the voice processing device 20, for example, as shown in FIG. 19, when the voice input is started (YES in S51), timing by the timer 37 is started (S52). The audio processing device 20 records the recording data input from the microphone 6 in the audio buffer (S53), and monitors whether or not the frame division set time t has elapsed (S54). When the set time t has elapsed (YES in S54), the recording data is frame cut (S55), and the process proceeds to a recording data file creation process (S56). After the recording data file is created, the recording data is converted into an output sound volume level as a file level conversion process for each recording data file (S57). The file creation process (S56) and the file level conversion process (S57) only need to be performed in the same manner as the above-described S15 and S18 (FIG. 10), and will not be described.

  Note that the audio processing device 20 may perform file creation processing and file level conversion processing simultaneously with the recording processing on the created recording data file. In the created recording data file, for example, continuous recording data including audio data may be combined into one recording data.

  (4) In the above embodiment, in the volume level conversion process for the recording data file, the volume level amplification or attenuation process is performed based on one output volume level. However, the present invention is not limited to this. In the volume level conversion process, a plurality of output volume levels may be set, and the conversion process may be performed based on any output volume level for each recording data file to be processed.

  For example, the user may set the output volume level for each file, or may set it based on time information of the recording process. Alternatively, the output volume level may be registered separately, or may be selected according to the voice quality of the speaker by analyzing the voice.

  According to such a configuration, the audio processing device 20 can execute the audio processing even when the capacity of the RAM that forms the audio buffer is small by creating the recording data file regardless of the arrival of the silent period. In addition, by making the length of each divided frame constant for the recorded data, it is possible to reduce the capacity burden on the audio buffer during recording. In addition, by dividing the recording data every predetermined time, the volume level can be analyzed finely in the section where the utterance time is long, and the output can be amplified or attenuated to reduce the variation of the volume, making it easy to grasp the volume. Data can be created.

  Next, the following additional notes are disclosed with respect to the embodiment described above. The technical idea related to the technology of the present disclosure can be grasped by various levels and variations from the superordinate concept to the subordinate concept, and the technology of the present disclosure is not limited to the following supplementary notes.

(Supplementary note 1) A voice processing device having a recording function,
A voice input means for capturing voice data of different levels and generating recording data;
The recording data generated by the voice input means is divided into a plurality of frames for each group of voices, the level of the voice included in the recording data is determined for each frame, and the level of the recording data is a reference level. If different, sound processing means for generating new recording data in which the level of the recording data is adjusted based on the reference level;
Storage means for storing the recording data generated by the voice processing means;
An audio processing apparatus comprising:

  (Additional remark 2) The said audio | voice processing means discriminate | determines the difference of the level of the said recording data taken in, and the said base level, and increases / decreases a level for every said frame based on this difference, It is characterized by the above-mentioned. Audio processing device.

  (Supplementary note 3) The voice processing according to supplementary note 1 or supplementary note 2, wherein the voice processing means monitors the volume level of the recorded data and adjusts the volume level of the recorded data based on the reference level. apparatus.

  (Additional remark 4) The said audio | voice processing means discriminate | determines whether the level of the said sound recording data is below a threshold value, and when it becomes below a threshold value, the said sound recording data are divided | segmented into frames, The additional notes 1 thru | or characterized by the above-mentioned. The speech processing apparatus according to any one of the supplementary notes 3.

(Supplementary note 5) Furthermore, a timer for measuring the recording time is provided,
The sound processing means divides the recording data to create a low level frame when the state where the recording data level is equal to or lower than the threshold value exceeds a predetermined time, and adjusts the level for frames other than the low level frame The speech processing apparatus according to appendix 4, wherein:

(Supplementary note 6) A voice processing method of a voice processing apparatus having a recording function,
Capture audio with different levels to generate recording data,
The generated recording data is divided into a plurality of frames for each unit of voice,
The level of the recording data is determined for each frame, and when the recording data level is different from a reference level, new recording data in which the recording data level is adjusted is generated based on the reference level.
A voice processing method including processing.

(Supplementary note 7) A voice processing program to be executed by a computer of a voice processing device having a recording function,
Capture audio with different levels to generate recording data,
The generated recording data is divided into a plurality of frames for each unit of voice,
The level of the recording data is determined for each frame, and when the recording data level is different from a reference level, new recording data in which the recording data level is adjusted is generated based on the reference level.
A voice processing program that causes a computer to execute processing.

The preferred embodiments of the configuration of the present disclosure have been described above. However, the technology of the present disclosure is not limited to the description of the above embodiment. It goes without saying that various modifications and changes can be made by those skilled in the art based on the gist of the technology described in the claims or disclosed in the specification. Needless to say, such modifications and changes are included in the technology of the present disclosure.

2, 20 Audio processing device 4 Audio input means 6 Microphone 8 Audio processing means 10 Storage means 12A, 12B Recording data table 12-1, 12-2, ... Frame 22 Processor 24 Storage section 26 ROM
28 RAM
30 Audio Input / Output Unit 32 Speaker 34 Volume Level Measurement Unit 36 Audio Amplifier Circuit 37 Timer 50 Volume Level Setting Table 70, 90 Recording Data Table 72, 74, 76, 80A, 80B, 82A, 82B, 84A, 84B Volume Level Information

Claims (5)

A voice processing device having a recording function,
A voice input means for capturing voice data of different levels and generating recording data;
The recording data generated by the voice input means is divided into a plurality of frames for each group of voices, the level of the voice included in the recording data is determined for each frame, and the level of the recording data is a reference level. If different, sound processing means for generating new recording data in which the level of the recording data is adjusted based on the reference level;
Storage means for storing the recording data generated by the voice processing means;
An audio processing apparatus comprising:   2. The audio processing apparatus according to claim 1, wherein the audio processing unit determines a difference between the level of the captured recording data and the reference level, and increases or decreases the level for each frame based on the difference. .   3. The voice processing means determines whether or not the level of the recorded data is equal to or lower than a threshold value, and divides the recorded data into frames when the level is equal to or lower than the threshold value. The voice processing apparatus according to 1. A voice processing method of a voice processing device having a recording function,
Capture audio with different levels to generate recording data,
The generated recording data is divided into a plurality of frames for each unit of voice,
The level of the recording data is determined for each frame, and when the recording data level is different from a reference level, new recording data in which the recording data level is adjusted is generated based on the reference level.
A voice processing method including processing. A voice processing program to be executed by a computer of a voice processing device having a recording function,
Capture audio with different levels to generate recording data,
The generated recording data is divided into a plurality of frames for each unit of voice,
The level of the recording data is determined for each frame, and when the recording data level is different from a reference level, new recording data in which the recording data level is adjusted is generated based on the reference level.
A voice processing program that causes a computer to execute processing.

Images