JPH06342562A - Voice reproducing device - Google Patents

Abstract

PURPOSE:To enable a rate converting process for voice data with various sampling frequencies even by a rate conversion processing means having a limit in the processing ability. CONSTITUTION:By performing a thinning process which reduces the number of input voice data per unit time, with a data thinning means 31 in accordance with the rate ratio and the processing ability, the rate conversion is processed by the rate conversion processing means 21 even for the voice data with various sampling frequencies.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an audio reproducing apparatus having a speed varying function when reproducing a recorded audio signal.

[0002]

2. Description of the Related Art Conventionally, a voice reproducing apparatus having a speed varying function of a recorded voice signal has been put into practical use. Particularly in recent years, a device that performs a speed conversion processing function using digital signal processing has been put into practical use. The speed conversion process here means a process of changing only the speed while keeping the pitch constant. For the speed conversion processing, refer to, for example, "Digital Speech Processing", Sadahiro Furui, "TDHS Algorithm" of Tokai University Press.

An audio reproducing device having a variable speed function will be described below with reference to the drawings.

FIG. 3 shows the structure of a conventional audio reproducing apparatus having a variable speed function. In FIG. 3, reference numeral 1 is a reproduction mode selection means, and 2 is a speed conversion processing means.

The operation of the audio reproducing apparatus configured as described above will be described below. First, the audio data is directly supplied to the reproduction mode selection means 1 and the speed conversion processing means 2. The speed conversion processing means 2 performs speed conversion processing on the audio data based on the given speed ratio data and outputs it to the reproduction mode selection means 1. When the given speed ratio data is 1, the reproduction mode selection means 1 selects and outputs the audio data which is the normal speed, and when it is other than 1, selects the output signal of the speed conversion processing means 2. Output.

With this configuration, both the normal speed voice and the speed-variable voice can be heard as needed.

[0007]

However, in the above configuration, the speed conversion processing means 2 has a limited processing capacity, and when the sampling frequency of the input voice data is not constant, the voice data can be processed as it is. There are both cases that can not be processed and cases that cannot be processed. This is because the number of data to be processed per unit time is determined by the sampling frequency of the digital signal for which the speed conversion is performed. If the sampling frequency is doubled, the number of input data will be doubled, and the number of processes may increase by the square of 2 depending on the processing content.

Therefore, for example, when recording voice, 8k
When it is possible to record by encoding at 4 sampling frequencies of Hz, 16kHz, 32kHz, 64kHz, if you use the speed conversion processing means with the optimum processing capacity for 16kHz audio data, you can record at 32kHz, 64kHz. If the processed voice data is used as it is, the processing cannot be completed in time and the speed conversion processing cannot be performed. That is, the data of the sampling frequency, which requires a larger amount of processing, cannot be reproduced with variable speed. Conversely, sampling frequency 64kHz
Using a high-performance speed conversion processing means that has the optimum processing capacity for the voice data leads to an increase in cost, and the capacity is wasted when processing is performed for sampling frequencies other than 64 kHz. It will be.

The present invention solves the above problems and provides an audio reproducing apparatus capable of processing audio data and reproducing the audio at variable speeds even when the processing capacity of the speed conversion processing means mounted is limited. The purpose is to do.

[0010]

In order to achieve this object, the audio reproducing apparatus of the present invention thins out audio data at a predetermined rate per unit time and outputs data after passing through a filter satisfying the sampling theorem. Means and a speed conversion processing means for giving the number of pieces of data that can be processed per unit time to the data thinning means based on the input speed ratio and performing speed conversion processing on the thinned input data. The feature is that the reproduction speed can be changed.

[0011]

With this configuration, since the voice data subjected to the thinning processing according to the processing capability of the speed conversion processing means can be supplied to the speed conversion processing means, the voice data of various sampling frequencies can be processed.

Further, since the voice data in which the band for processing is selected according to the processing capability of the speed conversion processing means can be supplied to the speed conversion processing means, it is possible to process an audio signal which originally has a wide bandwidth. Become.

[0013]

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

FIG. 1 is a block diagram showing a first embodiment of the present invention. In the figure, 1 is a reproduction mode selection means, 21 is a speed conversion processing means, and 31 data thinning means. The operation will be described below.

In the present embodiment, it is assumed that the input audio signal has already been recorded on the audio storage medium and desired data can be read from this medium.

The speed ratio data is given to the speed conversion processing means 21 and the reproduction mode selection means 1. Data thinning means 3
The number 1 determines the number of pieces of audio data to be subjected to the speed conversion processing per unit time in consideration of the processing capacity of the speed conversion processing means 21 and the speed ratio data, and performs the audio data thinning processing. The thinned ratio Rdec is the speed conversion processing means 21.
Also give information. The data thinning means 31 also performs antialiasing filter processing on the thinned audio data to absorb a change in sampling frequency due to thinning, and supplies output data to the speed conversion processing means 21. The speed conversion processing means 21 performs speed conversion processing on the audio data based on the given Rdec and outputs it.

When the speed ratio is 1, the reproduction mode selection means 1 selects the audio data of the normal speed as output data, and when the speed ratio is other than 1, it selects the output data of the speed conversion processing means 21. Output. Thus, even when the number of input voice data per unit time is larger than the capacity of the speed conversion processing means 21, the bandwidth is reduced as compared with the original voice data, but the speed is reduced. A conversion process can be performed.

As described above, according to this embodiment, the number of audio data per unit time can be reduced by the data thinning means in consideration of the processing capacity of the speed conversion processing means and the speed ratio. It is possible to perform speed conversion processing on audio data of various sampling frequencies by using speed conversion processing means having a limited capacity.

FIG. 2 shows a block diagram of the second embodiment of the present invention. In the figure, 1 is a reproduction mode selection means, 22 is a speed conversion processing means, 32 is a bandwidth selection means, 41 is a first band synthesis filter, 42 is a second band synthesis filter, and 43 is a third band synthesis filter. , 51 is a first decoding means, 52 is a second decoding means, 53 is a third decoding means, and 54 is a fourth decoding means. The operation will be described below.

In the present embodiment, it is assumed that the voice data encoded by the 4-band sub-band encoding method has already been recorded in the voice storage medium and desired data can be read from this medium.

First, the operation during reproduction in the normal mode, that is, when the speed ratio is 1, will be described. The audio data divided into four bands are respectively the first to fourth decoding means 51 to 54.
And is decoded (for example, ADPCM decoding). The decoded data of each band is the first to third
The band synthesizing filters 41, 42 and 43 up-sample, at the same time they are filtered and synthesized to be reproduced as the sound of the original band.

On the other hand, when the speed ratio is other than 1, that is, in the speed conversion mode in which the speed is changed and reproduced, it is necessary to perform signal processing based on the speed conversion algorithm. The bandwidth and sampling frequency for signal processing are determined based on the data indicating the calculation capability of the speed conversion processing means 22 mounted on this apparatus and the data on the given speed ratio. Speed conversion processing means 22
Performs a speed conversion algorithm on the basis of the given bandwidth and sampling frequency, and outputs speed converted voice to the reproduction mode selection means 1. Reproduction mode selection means 1
When the speed ratio is 1, the output data of the band synthesizing filter 3 which is the normal speed is selected, and when the speed ratio is other than 1, the data processed by the speed conversion processing means 22 is selected and the output data is selected. And As described above, even when the sampling frequency of the input audio data is higher than the capacity of the speed conversion processing means 22, the bandwidth is reduced as compared with the original audio data, but the speed conversion processing is performed. You can

As described above, according to the present embodiment, the bandwidth and sampling frequency of the audio data processed by the bandwidth selection means can be reduced in consideration of the processing capacity and speed ratio of the speed conversion processing means. Therefore, it is possible to perform the speed conversion processing on the voice data of various bandwidths by using the speed conversion processing means having a limited capacity.

[0024]

As described above, according to the present invention, the data thinning means for thinning out the voice data at a predetermined rate per unit time and outputting the data after passing the filter satisfying the sampling theorem is provided. Even if the speed conversion processing means having a limit is used, it is possible to realize an audio reproducing device capable of performing speed conversion processing on audio data of various sampling frequencies.

Further, since the sampling frequency and the bandwidth are determined based on the processing capacity of the speed conversion processing means and the bandwidth selecting means for selecting the audio data to be subjected to the speed conversion processing is provided, the processing capacity is limited. Even if a certain speed conversion processing means is used, it is possible to realize an audio reproducing device capable of performing speed conversion processing on audio data of various bandwidths.

By adaptively setting the repeat section in which the repeated reproduction is performed according to the utterance speed of the audio signal, it is possible to provide the audio reproduction apparatus capable of performing the repeated reproduction more suited to human sense. That is, a relatively short repeat section length is set when the speech rate is low, and a relatively long repeat section length is set when the speech rate is fast. This is expected to make it easier to understand when there are missed words or unclear meanings and the user wishes to listen again.

Furthermore, by performing the speed conversion process in consideration of the vocalization speed, it is possible to obtain a repeated reproduced sound which is easy to understand.

[Brief description of drawings]

FIG. 1 is a configuration diagram of an audio reproducing device according to a first embodiment of the present invention.

FIG. 2 is a configuration diagram of an audio reproducing device according to a second embodiment of the present invention.

FIG. 3 is a block diagram of a conventional audio reproduction device.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 reproduction mode selection means 2 speed conversion processing means 21 speed conversion processing means 22 speed conversion processing means 31 data thinning means 32 bandwidth selection means 41 first band synthesis filter 42 second band synthesis filter 43 third band synthesis filter 51 first decoding means 52 second decoding means 53 third decoding means 54 fourth decoding means

Claims (2)

[Claims] 1. A data decimating means for decimating voice data at a predetermined rate per unit time and outputting after passing through a filter satisfying a sampling theorem, and processing per unit time based on an input speed ratio. An audio reproducing apparatus comprising: a speed conversion processing means for supplying a data number to the data thinning means and performing speed conversion processing on the thinned input data, wherein the reproduction speed can be varied with a constant pitch. 2. An n number of decoding means for decoding voice data coded after band division into n bands (n ≧ 2), and for synthesizing each band after decoding and returning to the original frequency band. The band synthesizing filter, the bandwidth selecting means for determining the sampling frequency and the bandwidth based on the processing capability of the speed converting processing means and selecting the audio data to be subjected to the speed converting processing, and the bandwidth selecting means. And a speed conversion processing means for performing speed conversion processing on the corresponding audio data at a corresponding sampling frequency and bandwidth, wherein the reproduction speed can be varied with a constant pitch.