JPS5931136B2 - Translation playback device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates generally to recording and reproducing devices, and more particularly to recording and reproducing devices used by stenographers to convert spoken words into written words.

In this specification, the operation of converting spoken words into written words by writing or using a typewriter will be collectively referred to as the term "translator," and the person who performs the conversion will be referred to as a "translator." In recording and playback devices, the start is usually done manually (with a foot switch) to enable the stenographer to listen to one word group and translate that word group before hearing the next word group. - It is well known to play recordings in a stop-start manner. In order to use the manual operation switch while translating, it is necessary to intermittently perform typing or writing actions related to both performing the translation and operating the playback device, as well as listening to the playback sound. Translators require a high level of skill and concentration, so if they continue to work for long periods of time, the translator will become fatigued. It is an object of the invention to provide an automated system for controlling a reproduction device used for reproducing spoken words for trans-translation, with appropriate control of the reproduction speed adapted to the needs of the individual scribe. It is to provide a timer.

Another object of the invention is to avoid interrupting words or groups of words during automatic control of playback of recordings for trans-translation.

Another object of the invention is to allow selection of on-off time ratios and control of word and word group interruptions, allowing for a natural speed of reproduction that allows the translator to translate almost continuously. The purpose of the present invention is to control the playback of a recording/playback device.

Yet another object of the present invention is to provide the translator with on-off word clusters without the need for manual operations, other than the manual operations normally required to effect reproduction and the manual operation of automatic devices. The recording is played back almost continuously as presented.

Note that the term "translation" used in this specification indicates an operation of converting spoken words into written words through actions such as writing or typing.

Hereinafter, the present invention will be explained in detail with reference to the drawings.

The audio reproducer 11 shown in FIG. 1 is for reproducing audio recorded on a reel-to-reel type or force set type magnetic tape. This reproducer 11 normally provides an audio output to the translator's handset 12, and playback of the recorded audio is controlled by a start/stop control input.
This start/stop control input operates to enable or disable driving of the recording medium. That is, the start-stop control inputs can be electrical inputs for providing to a drive motor or for controlling a drive capstan that engages or leaves the tape. The start/stop control input normally applied to start/stop control input terminal 13 to drive the recording medium is provided from a foot switch for an office dictation machine. The switch input circuit leading to the input terminal 13 consists of a normally closed switch 14 and a series circuit of a normally open contact 16 and a parallel contact 15.

The normally closed switch 14 is for manually stopping the regenerator 11. (A conventional play switch and stop switch can also be provided). The input circuit to input terminal 13 is such that switch 14 is in its normally closed position and contacts 15 or 16 are closed.
is completed when either or both are closed. The switch 15 has two manual adjustment knobs 18 and 19.
The timer 17 is controlled by a timer 17 having the following values.

Knob 18 determines the time interval for timer 17 to maintain contact 15 open. Adjustment knob 19 selects the time interval determined by timer 17 for contact 15 to remain closed. Therefore, switch 14
is in the closed position, and the adjustment knobs 18 and 19 are in the timer 17 position.
When the set time is selected, the start-stop control circuit of the regenerator 11 includes a running time during which the contact 15 is in the closed state, determined by the selected position of the knob 19, and the start-stop control circuit of the regenerator 11 includes a running time during which the contact 15 is in the closed state, determined by the selected position of the knob 19, and is interrupted by the time that contact 15 is in the open state, defined by knob 18.
In this way, the regenerator 11 normally operates and stops operating at a predetermined on-off ratio, with the actual on-time and off-time being determined by the knobs 19 and 18, respectively. Knob 1
The on/off ratio determined by 8 and 19 is changed by a discontinuity detector 21.

This discontinuity detector 21 is coupled to line 22 to detect the audio signal extracted by the reproducer 11 from the recording medium. That is, when the reproducer 11 reproduces the audio from the recording medium, there is always an audio input from the line 22 to the discontinuity detector 21. Discontinuity detector 21 closes contact 16 in response to the level of the audio signal applied from line 22. Contact 16 opens whenever the level of the audio signal applied via line 22 falls below a certain level for a predetermined period of time. By controlling the time constant of the break detector 21 to match the normal break time between word groups in normal speech, the contact 16 can be controlled so that the contact 16 does not open in the middle of a word group. Since breaks within words are shorter than breaks between word groups, contact 16 does not open at breaks in the middle of words. Therefore, when the operating time determined by the adjustment knob 19 of the timer 17 ends and the break time determined by the adjustment knob 18 is started by opening the contact 15, the parallelism set by closing the contact 16 ends. In the closed circuit, the next break in the word group is line 2.
The regenerator 11 is kept in operation until detected by the absence of the signal applied via 2. In this way, the audio reproduction of the word group is completed before the drive of the recording medium is stopped. The remainder of the break time set by knob 18 will elapse before the break time ends and the regeneration drive is started. The contacts 15 are then closed for a preset operating time and then the above operation is repeated. If the total break time in this series of movements is too short for the translator, adjust the knob 18 to lengthen the break time.
Alternatively, a simple correction can be made by adjusting the knob 19 to shorten the operating time. The above explanation can be applied to any recording/playback device as long as it has a start/stop control function.

Other features and advantages of the present invention include, for example, U.S. Pat.
It will be understood from the description of an example of the application of the invention to an audio compressor/decompressor of the type disclosed in No. 6195. A controller 32 that controls the recording medium drive speed applied via line 33 and the frequency conversion applied via line 34 in conjunction with each other causes the audio compressor/decompressor 31 to perform the desired frequency conversion. Therefore, the voice can be reproduced at any desired speed while the frequency spectrum of the voice output is returned to the frequency spectrum of normal spoken words. For this purpose, the audio output of the regenerator 11 is applied via line 35 to an audio compressor/expander 31, where it undergoes frequency conversion before being returned to the regenerator 11 via line 36, and finally is provided to the handset 12. These added features not only allow the translator to control the on/off time ratio of playback without interrupting word clusters, but also allow the translator to control whether the translator's translation speed is faster or slower than the speaking speed. You can choose the playback speed higher or lower depending on how you like it. Further, after the translation is completed, the recording medium is played back at a high speed, for example, twice the normal speed, and pitch correction is performed by the compression/expansion device 31, so that errors in the translation can be corrected at high speed. At this time, the compression/expansion device 31
returns to normal speaking frequency. In addition, the frequency or pitch correction performed by the compressor/expander 31 aids in understanding the reproduced sound at an intelligible high burst rate, and also enables the translation of relatively long pronounced word sequences. An appropriately long piece can also be set using the knob 18. This can roughly be said to correspond to speed reading, in which a large amount of information is absorbed at high speed and the necessary pauses for retranslation are adjusted according to the translator's speed of retranslation. Next, with reference to FIG. 2, a logic diagram of a device constituting the device of the present invention using a monostable multivibrator will be explained.

The audio signal extracted from the recording medium being driven is applied to a dead zone comparator 42 via line 41. This dead zone comparator 42 does not generate an output for positive and negative signals below a certain absolute amplitude threshold, and outputs no output for signals greater than the positive or negative threshold. It produces a series of pulse outputs that span the audio signal amplitude. This pulse output is connected via line 43 to a retriggerable monostable multivibrator 44 and a non-retriggerable monostable multivibrator 4.
5 and 6 monostable multivibrator 44 is triggered by a pulse signal applied via line 43 to maintain one state, and when the pulse signal disappears, monostable multivibrator 44 maintains that state for a predetermined time. Then go back to the opposite situation. Typically, the time interval for monostable multivibrator 44 is approximately 200 ms, and its output ceases 200 ms after the pulses being applied via line 43 cease. The monostable multivibrator 45 can vary its output pulse length under the control of a controller 50 and is triggered by a pulse signal applied to its input terminal via line 43 to the length of the desired spoken word sample. Remain in one state for a corresponding selected period of time between approximately 2 and 8 seconds. After that time has elapsed, the monostable multivibrator 45 returns to its original state and outputs no more. The outputs of the monostable multivibrators 44 and 45 are applied to the AND gate 48 via lines 46 and 47 and the inverter, respectively, so when both outputs disappear, the output of the AND gate 48 is applied to the stop monostable multivibrator 49. Added.

The stop monostable multivibrator 49 is regulated by the controller 51 to generate an output pulse of a predetermined duration, e.g. 2 to 8 seconds, after which time the monostable multivibrator 49 returns to its original state. . The output pulse appearing on the output line 52 of the monostable multivibrator 49 is used to control the driving and stopping of the recording medium, and is also applied to the monostable multivibrator 45 to reset it, so that the monostable multivibrator 4
5 can be triggered on the next audio signal applied via line 43. Next, the operation of the logic circuit shown in FIG. 2 will be explained.

Assuming the recording media driver is running and an audio signal is present on line 41, dead zone comparator 42 generates a pulse on line 43 while the level of the audio signal is at an appropriate value. Another function of comparator 42 is to prevent signals from appearing on output line 43 due to noise occurring during playback of the recording. When monostable multivibrator 45 receives a pulse from line 43, this monostable multivibrator 45
is triggered to begin a timing cycle of approximately 2-8 seconds. The repeated appearance of audio pulses on line 43 continually retrigger monostable multivibrator 45 so that its output continues to produce a high level output for as long as such audio signal is present. . When the timing period of monostable multivibrator 45 expires, its output goes low to detect the next break in the audio signal. That is, when no audio signal is present on the output line 43 for about 200 ms, the level of the output of the monostable multivibrator 44 becomes low. Monostable multivibrator 44
, 45 are both at low level, the AND gate 48
provides a signal to trigger the monostable multivibrator 49, thereby causing a pulse signal appearing on the output line 52 lasting from 2 to 8 seconds to stop driving the recording medium. Then, of course, the audio signal applied to line 41 and the pulse applied to output line 43 disappear. When the pulse applied to the output line 52 disappears, a resetting signal is applied to the monostable multivibrator 45 via the line 53 to reset the multivibrator. respond to the next audio signal applied. Since the recording medium also starts when the pulse applied to the output line 52 disappears,
After a suitable break time, the audio signal appears again on line 41 and the pulse signal appears on line 43. Then, since the monostable multivibrator 45 has been reset, the above operation will be repeated from this point on. In this way, any break time added and the time between 2 and 8 seconds set by the monostable multivibrator 45 and the break in the audio signal determined by the monostable multivibrator 44 is reached. The recording medium is driven for the required time plus the required time, and after 2 to 8 seconds have elapsed and a break is detected, the recording medium driver stops operating for the time determined by the monostable multivibrator 49. I am made to do so. By adjusting the outputs of the monostable multivibrators 44 and 49, the length of the drive time and the length of the stop time can be selected depending on the speed of translation. In order to extend the running time of the recording medium when there are sufficiently wide gaps or silent periods in the spoken word, the second
The circuit shown in the figure can be modified.

In such cases, the translator is generally able to remember the content for a longer playback time than usual because the spaces between the words reduce the overall amount of information that needs to be stored until it can be translated. For this purpose, the switching gate 54
A positive potential for charging the timing capacitor of a monostable multivibrator 45 is applied to the input terminal of the monostable multivibrator 45, and a pulse is applied to its control input terminal via line 43. This additional circuit extends the timing of the monostable multivibrator 45 after it has been triggered. The reason is that the timing capacitor is charged only while a negative going word pulse is applied from line 43. During the positive going pulse plus the intermission period between the negative going word pulses, the switching gate 54 prevents the timing capacitor of the monostable multivibrator 45 from being charged, thereby causing the monostable multivibrator 45 to charge. Stable multivibrator 4
5. Extend the cycle time. Therefore, for speech that includes interruptions, a correspondingly long driving time is automatically provided by the monostable multivibrator 45. Otherwise, the operation of the circuit shown in FIG. 2 is the same as previously described. In another embodiment shown in FIG.
A logical detector detects the presence or absence of words in the audio signal reproduced from the recording medium, and integrates inversely depending on the presence or absence of words.

By selectively performing adjustable integrals separately for the two integral orientations, the translator can choose the length of the word groups to be reproduced and the length of the breaks between the word groups. .

In this way, it is possible to select a group of words that a translator can memorize, for example, as a group of words corresponding to a direction, and to select a group of words that a translator can memorize as a group of words corresponding to a direction, and to You can also select the pause time between words. The embodiment shown in FIG.
It includes a conventional regenerator 11 with 2.

The reproduced audio signal provided to the handset 12 is transmitted over line 64.
It is also applied to one input terminal of variable threshold amplifier 65 via . A variable voltage is applied to an input terminal 66 of this amplifier 65 by a variable resistor 67. Amplifier or threshold circuit 65 logically detects the presence of a word on line 64 as the audio level corresponding to a spoken word exceeds the threshold set by the operating voltage applied to input terminal 66. Then, the variable resistor 67 is set so that the level of the output of the amplifier 65 changes. At the end of a word, the voice signal disappears, and the input applied to input terminal 66 returns amplifier 65 to its normal state, producing an output on output line 68 at a level corresponding to the absence of a word signal on line 64. The output of amplifier 65 charges capacitor 71 when a word is present on line 64 and discharges capacitor 71 through a path including variable resistor 72 when no word is present on line 64 .

The word presence signal provided on output line 68 is applied to terminal 73 of comparator 74, but the absence of a signal on line 68 causes
Variable resistor 72 sets the time constant of the charge-discharge circuit such that the level of the signal applied to input terminal 73 decreases when a break that normally occurs between spoken words is met. In other words. ．． The input applied to input terminal 73 is a signal detecting the presence or absence of words that is averaged to eliminate breaks between syllables, which are typically shorter than breaks between individual words.

The output of comparator 74 is applied to an integrator.

This integrator is composed of an integrating amplifier 75 and an integrating capacitor 78. The output signal of the comparator 74 has its polarity inverted by an inverter 79 with a gain of 1. The level of the signal applied to the negative input terminal 76 of the integrator is determined by the potentiometer 8.
1. This control is performed using potentiometer 81.
This is done using a manual knob 82. An integral signal applied to the positive input terminal 77 of the integral amplifier 75 is provided from a potentiometer 83. This potentiometer 83 is adjusted by a knob 84. potentiometers 81 and 83
controls the level of logic signals from capacitor 74 representing the presence and absence of a word, respectively. Adjusting potentiometer 83, which causes integrator 75 to integrate in the positive direction, therefore controls the sample length or word cluster length that is reproduced before reaching the high integration endpoint set by Schmitt trigger 85. When no word is present, the signal of opposite polarity applied to the input terminal 76 integrates in the opposite direction, and when a low threshold is reached by integrating the signal generated by the setting of the potentiometer 31, the Schmitt trigger 85 is activated. The polarity of the output is inverted. The output of Schmitt trigger 85 is applied to power transistor 86. This transistor 86 is directly connected to the control line 63 of the motor. Normal manual regeneration switching is performed by a manual switch 10. The output terminal of Schmitt trigger 85 is connected to the output terminal of comparator 74 via voltage dividing resistors 91 and 92, and the common connection point of these voltage dividing resistors 91 and 92 is connected to input terminal 93 of Schmitt trigger 85. Such a connection prevents the motor from stalling even if a high integration level is reached in the middle of a word. If a word is present when the motor is rotating, the high level output of the Schmitt Trigger causes the output of comparator 74 to go high. Therefore, when the voltage dividing resistors 91 and 92 are at a high output level, the common connection point connected to the input terminal 93 of the Schmitt trigger 85 prevents the output of the Schmitt trigger 85 from going low. However, when integrator 75 reaches the high threshold level of Schmitt trigger 85, comparator 74
When an end-of-word is generated that causes the output of the integrator to go low, the Schmitt trigger 85 receives the output of the integrator and forces its output to a low level, causing the Schmitt trigger to disconnect the motor by making the transistor 86 non-conducting. 85. The operation of the circuit shown in FIG. 3 is believed to be clear from the above description and can be further explained with reference to FIG.

Each knob 82 of the potentiometers 81 and 82,
By adjusting 84, the translator can obtain a word group of the desired length including the desired length of break time,
and those parameters remain more or less constant regardless of changes in speaking rate or introductions during normal speech. As shown in FIG. 4, sufficiently choppy speech results in positive and negative integrals extending over an active period 101 followed by a standard stop period 102. Next operation period 1
03 has fewer breaks and therefore a shorter time, but period 1
Contains almost the same word content as 01. The subsequent outage period is of standard length. Assuming that the integral gain (slope between + and -1) is adjusted to keep the level below the stop threshold for a given distribution of speaking periods and breaks in speech, the playback device operates only for an extended period of time. On the other hand, by adjusting the integration rate, one can achieve the desired balance between word group length and pause time. In other words, these two types of adjustments allow the translator to match the flow rate of the words to the speed of the translation, allowing the device to produce word groups at a speed that allows smooth translation.

[Brief explanation of the drawing]

1 is a block diagram of an apparatus according to the invention together with associated apparatus for audio compression or reproduction; FIG. 2 is a block diagram of an alternative embodiment of the apparatus shown in FIG. 1; FIG. FIG. 4, a circuit diagram of another embodiment, is an integration timing diagram for explaining the operation of the circuit of FIG. 3. 11...Regenerator, 13...Start...
Stop control input terminal, 15... Timer switch, 17... Timer, 21... Discontinuation detector, 31... Audio compression/expansion device, 42...・
... Dead zone comparator, 44,45,49... Monostable multivibrator, 74... Comparator, 8
5...Shimitu trigger.

Claims (1)

[Scope of Claims] 1. A recording and reproducing device having a start-stop motor controller, and timing for operating the motor controller, including alternately repeated predetermined operating periods and stop periods for the motor. an apparatus, a break detector responsive to an audio signal being played from a recording medium to detect the absence of a signal during a time corresponding to a break between spoken words; circuitry coupled to the discontinuity detector for extending the period of operation of the motor beyond the normal start time of the preselected stop period until the next discontinuity is detected. A playback device for retranslation. 2. A means for operating a recording medium in order to obtain an audio signal from the recording medium, a signal transponder for stopping the operation of the recording/reproducing device for a predetermined stop period, and a means for operating the recording medium to obtain an audio signal from the recording/reproducing device; a detector for detecting the presence of the audio signal; means for starting a predetermined time interval in response to a first detection of the audio signal; and timing a break detection period in response to the presence of the audio signal; means for extending a predetermined time period beyond the absence of said audio signal; and means for generating a stop signal responsive to the end of both said predetermined time period and said predetermined time interval; and means for applying a signal to the signal response means for stopping the operation of the regenerator. 3. a recording/reproducing device having a start/stop control motor; a circuit device coupled in response to an audio signal reproduced from a recording medium to logically detect the presence and absence of a word; an integrator responsive to said circuit arrangement to generate an integrated value by integrating in one direction and integrating in the other direction in the absence of a word, and setting integration boundary points in said two directions; means for stopping the motor in response to the integrated value reaching an integration boundary point during the presence of a word; and means for stopping the motor in response to the integrated value reaching an integration boundary point during the absence of a word; and a motor controller that stops the motor when the motor is stopped.