JPH0210959B2 - - Google Patents

Description

[Detailed description of the invention]

Description of the Technical Field The present invention is an audio data output method for reproducing and outputting a human voice in response to an instruction to compose a necessary edited sentence from a storage device in which a human voice is divided into units such as words or phrases and recorded in advance. Regarding. Technical background of the invention and its problems Voice output devices have been put to practical use in various fields as voice response or voice guidance. For example, there are voice responses regarding the operation of a vending machine, voice guidance regarding the operation of an elevator, voice guidance regarding the operation of a supervisory control device for factory equipment, etc., and voice responses regarding a warning of an accident in the equipment. Voice output is easy for humans to understand, has few errors, and can convey a large amount of information in a certain amount of time, making it excellent for prompt reporting. Also, it is better to use audio rather than text and graphics for instructive information or warning information for humans. Conventional voice output devices used for voice response or voice guidance store as many types of voice information as possible (word-by-word or phrase-by-phrase sounds) and output them in a short time (within a few seconds) based on an external command. , has been improved with the aim of creating a device that can edit text into the desired text and reproduce it. For example, audio data units for various words and phrases are stored in memory as shown in FIG.
In other words, there is a voice “No. 1” between address a ₀ and length l.
digitized audio data units are stored sequentially in byte units. Similarly, length m from address _b
A data unit of voice "No. 2" is stored in between, and as a whole, various voice data units are sequentially stored as shown in the figure. Conventionally, when a memory with such a configuration is to be edited into a sentence by combining words according to an external command, a table like Table 1 is set up, and the corresponding sentence is edited by inputting the sentence number from the outside. and is output for playback.

[Table] That is, when outputting the sentence with sentence number 1, the word number is read from the combination word number list table of sentence 1 as shown in Table 1, and the start address of the audio data unit corresponding to that word number and its Read the length, that is, the memory size. Then, data corresponding to the memory size from the start address of the corresponding audio data unit is converted into an audio waveform and output. For example, at the beginning, the data unit "No. 1" is determined by the start address a ₀ of word number 0 and the size l.
is read and reproduced, and then the data unit "pump" is read and reproduced according to the start address d ₀ and size n of word number 3, and each word is sequentially reproduced and output. Outputs the sentence number 1 “It has started.” In this way, in order to read out the audio data unit of each word, a storage table for the memory start address (a ₀ , b ₀ ...) and a storage table for each memory size (l, m, ...) are required. Ta. The start address is essential for reading the audio data, but since a memory size table is also required, the creation of that table and the program for processing the table become complicated, and the table's worth of data becomes more complicated. There are drawbacks such as the need for extra space in ROM4. To this end, one possible method that does not require a large memory size is to insert an "end code" at the end of the audio data of each word. However, even with this method, if there is a code that does not exist in the audio data (for example, data in bytes), it can be distinguished as an "end code", but if you look at the audio data in bytes, it is expressed in hexadecimal. This method is not possible if all numbers from 00 to FF exist. OBJECT OF THE INVENTION An object of the present invention is an audio data output method that enables reading of corresponding audio data units without providing a data size table, facilitates table creation, etc., and reduces memory space. Our goal is to provide the following. SUMMARY OF THE INVENTION The present invention sequentially stores sound data units for each word or phrase in a memory, and when reading out a sound data unit corresponding to the content of a sound output command from this memory, reads the sound data unit at the beginning of the sound data unit in the memory. Specify an address, and sequentially read the data after this first address while incrementing the address, and each time this address is incremented, this address becomes the audio data stored next to the audio data unit currently being read in the memory. The audio data output method is characterized in that it is determined whether or not the first address of the unit matches, and if they match, the reading of data is stopped. Embodiment of the Invention An embodiment of the present invention will be described below with reference to the drawings. FIG. 2 is a diagram showing the configuration of an audio reproduction/output device that executes the method of the present invention. In the figure, 1 is a central processing element (hereinafter referred to as CPU), and a random access memory (hereinafter referred to as CPU) is connected via an address data and control bus (hereinafter simply referred to as bus) 2.
3. Read-only memory (hereinafter referred to as RAM) 3. Read-only memory (hereinafter referred to as
ROM) 4, audio data ROM 5, parallel-to-serial converter (hereinafter referred to as P-S converter)
6. Connect to input port 10 and output port 11, respectively. 7 is an audio reproducer, and P-S converter 6
receives the signal from the speaker 9 via the amplifier 8.
make it work. Here, CPU1 passes through bus 2 as mentioned above.
It is connected to RAM3, ROM4, etc. and exchanges data with them. In other words, CPU1 is ROM
4, inputs an external audio output command via the input port 10 as necessary, and outputs audio data corresponding to the input data. The audio data is read from the ROM 5 in bytes. Then, the data is converted into serial data for a fixed period via a P-S converter 6, reproduced into an audio waveform by an audio reproducer 7, amplified by an amplifier 8, and outputted as audio from a speaker 9. The audio data ROM 5 stores various audio data in the order shown in FIG. That is, as shown in FIG. 3, from address _a0 to address al, data units obtained by digitizing voice "1" are sequentially stored in byte units. Similarly, from address b ₀ to address bm, the voice “No. 2”
data units are stored, and as a whole, various audio data units are sequentially stored in the same manner as that shown in FIG. 1 described in the prior art. In addition, the audio player 7 is a CVSD (Continuously
It consists of a Uariable Slope Delta Modulatar) decoder and a filter circuit, and reproduces serial digital data into an analog waveform (audio waveform). The reproduced audio waveform is amplified by amplifier 8,
Audio is output from the speaker 9. In the above configuration, when reading out an audio data unit, a table shown in Table 2 is constructed in the ROM 4, and by inputting a sentence number from the outside, the corresponding sentence is edited and reproduced.

[Table] The difference between Table 2 and Table 1 (conventional example) is that the memory size table is omitted. That is, in order to obtain the same effect without requiring a memory size, the present invention assigns a word number to each word in the order in which audio data units are stored, and stores the start address of the corresponding word in that order, as shown in FIG. Create a starting address table that looks like this.
In this way, the audio data unit of a certain word number (for example, word number 0) will start from the audio data storage start address and end one place before the audio data storage start address of the next word number (word number 1). It turns out. The readout operation in units of audio data in this case will be explained with reference to the flowchart shown in FIG. In this case, the audio data of the word is output using the start address of each audio data unit of the word number and the next word number. For example, when outputting the word "No. 1" with word number 0, the start address (a ₀ ) of the audio data unit corresponding to the word is first read out from the start address table shown in FIG. 4 and set (step 51). ). Then that address (a ₀ )
Read and output audio data from (step
52, 53). Also, the address is incremented for each output (step 54). Furthermore, each time this address is incremented, it is compared whether the incremented address matches the next word, that is, the first address (b ₀ ) of word number 1, and a ₀
+ x = b ₀ (Here, x = 1, 2, 3...) Repeat reading and outputting the audio data (step
55). Therefore, the audio data is output until one address before the start address of the next word, that is, the final address of the audio data unit corresponding to that word. The above explanation relates to the output of one word, but the same applies when outputting a certain sentence, and the output of each word may be performed in the order shown in Table 2. Effects of the Invention As described above, according to the present invention, the
It is possible to obtain an audio output device that requires less ROM and is easy to program.

[Brief explanation of drawings]

Fig. 1 is a diagram showing the stored contents of the memory in the conventional method, Fig. 2 is a block diagram showing an example of the configuration of an audio output device for carrying out the method of the present invention, and Figs. 3 and 4 are the memories used in the method of the present invention. FIG. 5 is a flowchart showing an embodiment of the audio data output method according to the present invention. 1...Central processing element, 2...Address. data and control bus, 4...ROM for programs and tables, 5...audio data ROM, 7...audio reproducer.

Claims (1)

[Claims] 1. Sequentially store voice data units for each word or phrase in memory, and when reading a voice data unit corresponding to the content of a voice output command from this memory, specify the start address of this voice data unit in the memory, and The data after the first address is sequentially read while incrementing the address, and each time this address is incremented, this address matches the first address of the audio data unit stored next to the audio data unit currently being read in the memory. 1. An audio data output method characterized by determining whether or not they match, and stopping reading of data if they match.