SU417965A3 - - Google Patents

Abstract

1,263,523. Speech synthesis. RCA CORPORATION. 12 Feb., 1969 [26 Feb., 1968 (2)], No. 7608/69. Heading H4R. A speech synthesizer has a library of phonemes stored at selected locations of a data store to be read out in any desired sequence to synthesize desired audible information. Each phoneme is truncated at the beginning and/or the end to provide continuity of formants across the junctions between the phonemes in the desired sequence. The phonemes are read out in sequence as determined manually, by punched card, paper, or magnetic tape, or computer output. Each phoneme to be reproduced is represented by three numbers, one defining the position of the phoneme in the library and the other two the beginning and end respectively of the desired portion of the phoneme. Preferably the phonemes are stored in digital, i.e. pcm, form on a magnetic drum with core memories used as an intermediate store during reading in and reading out of the stores. Fig. 9 shows a system used for recording phonemes the analogue waveforms of which are fed as 7 bit pcm signals into a core memory A. The core memory can store 4096 seven bit signals which represent one phoneme. The signals are fed into the core memory in defined locations commencing when a record command pulse 707 is received. When the core memory is full, flip-flop 709 is tripped and further storage ceases. After an index pulse is received from the drum 21 the core memory is read out in predetermined sequence to store on the drum 21, in locations defined by the sector pulses permanently recorded on the drum, the digital signals defining the phoneme. Further phonemes are recorded in further tracks of the drum. Fig. 10 shows apparatus for reading phonemes from the drum 821 under manual, 831, or paper tape, 833, control. The phenomes requested by the manual or paper tape control are read alternately into the core memories 17 and 827 and while one phoneme is being read into one core memory the desired portion of the previous phoneme, as defined by the second and third numbers read out of the manual or paper tape control, is read out from the other core memory, functions of the core memories being interchanged as soon as the phoneme has been read out of the core store for reproduction. Reading in from the drum to the core store takes place faster than reading out providing allowance for the truncation of the reproduced phoneme. It is suggested that the phoneme might be recorded in analogue fashion on the drum.

Description

one

The invention is from the field of telephone techie and can be used for c) speech synthesis.

The syntetic information is well known; the sound information contains a memory device and is connected to the controller, and a controller; a vert, the memory device has puncture blocks for a set of coded phonemes.

However, the known synthesizer does not allow satisfactorily solve the problem of combining elements of speech.

The aim of the invention is to improve the quality of avionic information while reducing the volume requirements and the value of lCLiOlvap.

This is achieved by the fact that in the proposed yiCTipoftcTBe controller, the controller is executed as a serial gateway, a sequence controller and an ombrerator with extraction from the selected phonemes within each selected block of the corresponding phoneme information that represents a continuous electrical signal denoting (initial and final positions of neighbors of phonemes in a series of phonemes issued by the synthesizer.

FIG. 1 and 2 (A block diagram of the proposed device.

The phoneme is pronounced into microphone 1 and amplified by amplifier 2. The amplifier output is connected to the input: m of the strobe circuit 3, which is turned on and off with the help of oscillator 4. A pulse sequence is received at the input of the strobe circuit 3, the repetition frequency of which is equal to the oscillator frequency 4, and the amplitude is The instantaneous output of the amplifier 2.

The output | Signal ic of amplifier 2 is rectified by diode 5, smoothed by low-pass filter b, and fed to sequence controller 7.

If switch 8 is in the "load" position, then the absence of a constant output signal from the low-pass filter causes the sequence switch

transfer to register 9 of the address of the first device 10 of memory the starting address of the next phoneme. The start address and the end address are taken from the reader 11, which is, on the inspirator, arranged for reading from a paper tape, a reader for punched cards or magnetic tapes, it can also be an electronic computer or control

device of another kind. The final address is stored in a regispree in the coitroller of the succession of justice.

Address register 9 consists of registers that can be unpicked by the output signal of generator 4 through gate circuit 12, which is given: By acting as a signal from the sequence controller during the accumulation of a phoneme.

Input: The signal is converted into a binary number with aialo-C | digital converter 13, which, together with the 3 0 | st | gating circuit, forms the second KOiinBepTOp. Each number can be represented by a seven-digit binary number from the zero value “0000000 to the maximum value“ 1111111. Therefore, the cereal zero level of the output signal from amplifiers 2 is equal to sixty four 100 1,000,000. (In reality, the zero level is counted with an accuracy of five centimeters from the center. The counting direction depends on the number of inversions through the amplifier. The reason for this shift is in the TGiM, which is the amplitude of the sound (WAVES caused by exhalation, is greater than the amplitude these waves are caused by the action of the muscles of the larynx).

The cross-binary number, received from its apalo-dual converter 13, is exchanged in the output register 14 of the memory under the action of synchronization and synchronization, its signal coming from generator 4 through the gate circuit 12. When the phoneme fills all its storage cells, the accumulation process is considered complete. The end point is determined using KiOMinapaTopa 15, which determines the equality of the address register, address address and end address stored in the controller, the sequence, and the corresponding signal is sent to the sequence controller.

When all the phonemes needed to reproduce are accumulated, switch 8 is shifted to the read position.

When speech is synthesized into a reader 11, a corresponding record is placed, K | 0tor contains two binary numbers for each played phoneme, 1 where the first number corresponds to the starting address, and the second to the final address | memory in the memory (the starting and ending addresses need not be equal to the first and last addresses of the calsda phoneme, respectively). After this, the indicated pairs of addresses of the required phonemes are set up in a sequence corresponding to the required for reproduction of the desired speech.

The initial and final addresses are determined so that the value of the binary number accumulated in the memory is equal to "1,000,000, and the formants of the phonemes for all frequencies are in contact with the corresponding ends of the preceding phonemes, regardless of the initial address

and the start of the odds. subsequent phoneme, depending on the final address.

With the corresponding command to the controller of the sequence from the reading device 5, the first two addresses are not affected, and the starting address is transferred to the address register 9.

The phoneme located at this address is retrieved and stored in the output

0 to the memory register, to the output of which the first converter 16 is connected (digital-to-analog converter). The output voltage at the last amplitude is determined by the magnitude of the binary number on its

5 entrance.

When the next pulse arrives from the generator 4 through the gate circuit 12, the address register increases by one, so the following d is extracted from the first 10 memory device: B | Oichnaya number, which enters the input of the first inverter 16. The output voltage amplitude of the converter The value of the binary numbers arriving at its input is continuously changing.

After that, the specified output voltage 1 is supplied to the corresponding transducer 17, for example, a loudspeaker, who generates the desired sounds.

0 The indicated increase in the address register continues until iHop until it becomes equal to the second address, i.e. the last address that is stored in controller 7 (Sequences.

5 When receiving a control signal signifying reaching the end address, the controller 7 of the sequence immediately causes the reader to read the next two addresses.

0 The phoneme corresponding to these addresses is extracted and transformed in the same way into an audible sound.

The extraction and transformation of the phoneme continues for as long as the controller

5 sequence does not accept the stop signal from the reader 11 or an external source.

In the proposed device, the first converter in Seb includes a second memory device, which includes the first 18 and second 19 nodes for accumulating one phoneme. The second memory device is connected in working mode with the controller, which sends a

5 about the desired sequence of signals, representing complete phonemes selected sequentially from the first memory device to the first 18 and second 19 nodes for accumulating phonemes, and along the chain

0 21, 23, 24, 26, 28 transmits the readings of the digitally coded signals of each phoneme, successively supplied by the first and second nodes to accumulate the phonemes in the sequence in which the selection took place from the first memory device. The subject matter of the invention 1. Sound information synthesizer containing inepiBoe ycTpiOEcTBO memory and connected to an iHHiM co-controller, the first converter, and the first device. Tests typing accumulation blocks for a plurality of coded phonemes, characterized in that, in order to improve the quality of synthesized patterns ““ Formations with the application of the requirement of the “01bjumu Accumulation and the size of the dictionary, indicated by the controller, for example, in the form of. Sequentially connected strobing circuit, the sequence controller AND KOMinapaToipa, and | y1po; m put The sequence controller and the comparator extract from the memory block a part of the specified background information (which is a continuous electronic signal | Cesky signal indicating the initial and final positions of the neighboring phonemes in the series of phonemes. 2. The synthesizer according to claim 1, which is different there that said first converter includes a second memory device that connects a second memory device, connected in working condition to the controller, „,,„ л ™ .-. tn .n-r onnfi of the mother and the circuit to output the coded phonemes accumulated there to translate "x IB continuous electrical signals from the second memory, the second memory device provides a temporary accumulation of more than one of the selected coded phonemes. 3. The synthesizer of claim 1, which is different from the fact that said second memory device includes the first and second nodes for accumulating respectively one complete phoneme, and said controller has a circuit for transmitting to the desired sequence. signals representing the respective corresponding phonemes that were chosen successively: from the first memory device to the specified first "to" nodes for accumulating the phonemes, and also to a chain for displaying the digital-coding signals of each phoneme sequentially transmitted by rvym and second nodes for storing the phonemes in the sequence in which the sampling phonemes from the first naiMHTH. 4 Synthesis ator pop. 1. that is, so that the phoneme accumulation unit in the first memory of the synthesizer contains a second converter connected to the first device with the formation of electric sn-10V from sound phonemes, while the controller connected to the converter arrangement: memory, has circuits for obtaining a sequential accumulation of received signals in the memory and representing. “Th phonemes in combination with group jl ,,, t, 1-. l .., potts1, G PPU. e in a given sequence, mi phonemes accumulated in a given sequence | MI SON-vi, UdiMJii. respectability. 5 The synthesizer of claim 1, wherein the second converter is made of consistently connected gating circuits and an analog-to-digital converter connected to the input of the first memory device.

/ J

J

Oz