JPS6248824A - Voice data storage system - Google Patents

Abstract

PURPOSE:To improve the economy by storing one voice data in a memory at each 2-sampling period, extracting it at each 2-sampling period so as to interpolate the intermediate value at each sampling period thereby halving the storage capacity of the voice data. CONSTITUTION:A stored voice V1 is sampled, subject to PCM coding and reaches a reception register together with a clock signal ck1 and stored synchronously with a clock signal. A counter circuit 4 is advanced by a clock signal ck2 having a period twice the sampling period and inputs a count in a memory 2 as a write address a1 via a selection circuit 5. As a result, the voice data V2 at each 2-sampling period is stored sequentially in the memory 2. A counter circuit 6 is advanced by a clock signal ck3 having a period twice the sampling period, reads the count in the memory 2 via the selection circuit 5 and inputs it as an address. As a result, the voice data V3 is read from the memory 2, stored in a transmission register 3, the signal is extracted twice synchronously with a clock signal ck4 of one sampling period and outputted as a reproduced voice V4.

Description

[Detailed Description of the Invention] [Summary] In an audio storage device, one audio data is stored in a memory every two sampling periods, the stored audio data is extracted every two sampling periods, and the extracted audio data is stored in a memory. By interpolating the intermediate value for each sampling period from , it is possible to reduce the storage capacity of audio data by half.

[Industrial application field]

The present invention relates to an audio data storage method that makes it possible to reduce the required storage capacity of audio data in an audio storage device that stores and reproduces PCM encoded audio data.

One of the functions of private branch exchanges, which have become increasingly diverse in recent years, is a variety of voice guidance functions (so-called voice guidance functions).
is being put into practical use.

In this type of voice guidance function, the user of the private branch exchange plays back PCM-encoded voice data that has been registered in advance in the voice storage device at any time, so that desired subscribers can listen to it.

It is strongly desired that the storage capacity for storing audio data in such an audio storage device be reduced as much as possible in order to improve the economic efficiency of the audio storage device.

[Conventional technology]

In a conventional audio storage device, all audio data that has been sampled at a predetermined sampling period and PCM encoded is stored in a built-in memory, and during playback, each stored audio data is stored in one sample. They were extracted sequentially at each cycle.

[Problem that the invention seeks to solve]

As a result, for example, if one sampling period is 8 kilohertz, P
If the CM code is 8 bits, a storage capacity of 8 kilobytes will be required to store one second of audio data, which may impair the economic efficiency of the audio storage device.

[Means for solving problems]

FIG. 1 is a diagram showing the principle of the present invention.

In FIG. 1, a storage means 100 receives PCM-encoded audio data sampled at a predetermined period t, and stores it in a memory 200 every l audio data, and a storage means 100 that samples audio data stored in the memory 200 in two samples. There are provided an extraction means 300 for extracting data for each sampling period, and an interpolation means 400 for creating audio data at an intermediate point in each sampling period from the audio data extracted by the extraction means 300.

The interpolation means 400 uses the audio data extracted from the memory 200 immediately before as the audio data after one sampling period. Alternatively, the arithmetic mean value of two sets of audio data successively extracted from the memory 200 is determined, and the arithmetic mean value is determined as audio data at an intermediate point in time between the extraction times of the two sets of audio data.

[Effect]

That is, according to the present invention, audio data is stored in the memory built into the audio storage device every two sampling periods, and the required storage capacity is halved compared to the conventional method, making the audio storage device more economical. Improves sex.

Although the fidelity of the reproduced audio will deteriorate, there is no need to identify the speaker in usage forms such as voice guidance.
There is no problem in practical use.

〔Example〕

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

FIG. 2 is a diagram showing an audio data storage method according to an embodiment of the present invention, FIG. 3 is a diagram illustrating a reproduced waveform in FIG. 2, and FIG. 4 is a diagram showing an example of a reproduction waveform in FIG. 2. 5 is a diagram illustrating an audio data storage method, FIG. 5 is a diagram illustrating the operation process in FIG. 4, and FIG. 6 is a diagram illustrating a reproduced waveform in FIG. 4. Note that the same reference numerals indicate the same objects throughout the figures.

In FIG. 2, the receiving register 1 and the counting circuit 4 are provided as the storage means 100, the counting circuit 6 is provided as the extracting means 300, and the transmitting register 3 is provided as the interpolating means 400.

In Figures 2 and 3, the stored voice ■1 has a period t(
= 125 microseconds) and PCM encoded with a clock signal Ck of sampling period t, at time t6.
, t', . . . arrive at the receiving register 1. The receiving register 1 receives the arriving audio data. ,Vtl,...
is accumulated in synchronization with the clock signal ck.

On the other hand, the counting circuit 4 is stepped by a clock signal ck2 having a period 2t that is twice the sampling period, and inputs the counted value to the memory 2 via the selection circuit 5 as a write address al.

As a result, in the memory 2, audio data V is stored in the reception register l every sampling period. , Vtl, . . ., audio data ■2 (for example, V,
VtZ, . . . ) are sequentially accumulated at address at.

On the other hand, the counting circuit 6 is stepped by a clock signal Ck3 having a period 2t which is twice the sampling period, and inputs the counted value to the memory 2 via the selection circuit 5 as a successive address a2.

As a result, from the memory 2, the audio data V3 (for example, V,,Vi.(,V1
11+2,...) are read out sequentially, and the transmission register 3
is accumulated in

Audio data v3 (V, ,
V, , ,V,. 2, . . . ) are each extracted twice in synchronization with a clock signal ck4 having one sampling period t, and outputted as reproduced audio V4 (FIG. 3).

Next, in FIG. 4, in addition to the transmission register 3, a latch circuit 7, an arithmetic mean circuit 8, and a selection circuit 9 are provided as an interpolation means 400. In Figure 4, memory 2
In the same process as in Fig. 2, the second part of the accumulated sound ■,
Audio data V2 (=Vto,
Vtz,...) are accumulated.

During playback, in FIGS. 4 to 6, the clock signal c
Addresses a, (=m, m+L m+2,...) output from the counting circuit 6 incremented by k3 are input to the memory 2 via the selection circuit 5, and the audio stored in each address a2 is inputted to the memory 2 via the selection circuit 5. Data ■3 (=Vm, ■,, +1
% VSe2, . . . ) are sequentially extracted and transmitted to the launch circuit 7, the arithmetic mean circuit 8, and the selection circuit 9.

The latch circuit 7 receives audio data Vs to be transmitted in synchronization with a clock signal ck5 having a two-sampling period 2t.
(= V-, V-++, Vl, l"Z, ""),
Approximately 2 sampling periods 2 are delayed and held.

The arithmetic mean circuit 8 receives audio data Vz (=V-, V-, +, V□2, . . . ) transmitted from the memory 2, and
The arithmetic mean value Vb C"" Vm-o,s (= (
Vla-+ + Vm)' 2) -, Vm-+
, s (= (Vll, 2 +Vs-+)
'2), . . .] is calculated and transmitted to the selection circuit 9.

The selection circuit 9 selects the audio data V transmitted from the memory 2 using a clock signal Ck6 that is inverted every I sampling period.
3 (= ■@ , VSe1 % VSe2,...
), and the arithmetic mean value v, (
=v,,. +S, VSe1. S,...) are selected alternately, and the audio data ■Ma (=V, ,V,ya.,6
, ■m+I, VSe1.5.1m+2, "') are stored in the transmission register 3 in this order.

Audio data stored in the transmission register 3■.

(=V5, Vl11+.+5, Vm*t, Vls+
1. S, 1m+2, . . .) are sequentially extracted in synchronization with a clock signal ck4 having one sampling period t, and output as reproduced audio v8 (FIG. 6).

As is clear from the above description, according to this embodiment, the audio data 2 for every 2 sampling periods 2 are sequentially stored in the memory 2, and are extracted every 2 sampling periods 2 (V3), In FIG. 3, the same audio data is output twice in succession as reproduced audio ■4, and in FIG. 6, the arithmetic mean value is output as reproduced audio V in the middle.

Therefore, the storage capacity of the memory 2 is reduced by half compared to the conventional one. Furthermore, the reproduced voices ■4 and ■8 can also be used sufficiently for guidance purposes.

It should be noted that FIGS. 2 to 6 are only one embodiment of the present invention, and for example, stored audio ■1, reproduced audio ■4, and V
8 is not limited to what is shown in the drawings, and many other modifications may be considered, but the effects of the present invention will not change in any case. Furthermore, the interpolation means is not limited to playing back the same audio data twice in succession or outputting an arithmetic average value in the middle; many other variations can be considered, but in any case However, the effect of the present invention remains unchanged.

〔Effect of the invention〕

As described above, according to the present invention, the required storage capacity of audio data in the audio storage device is halved compared to the conventional one, and the economic efficiency of the audio storage device is improved without impairing the practical quality of reproduced audio. .

[Brief explanation of the drawing]

FIG. 1 is a diagram showing the principle of the present invention, FIG. 2 is a diagram showing an audio data storage method according to an embodiment of the present invention, and FIG.
FIG. 4 is a diagram illustrating an audio data storage method according to another embodiment of the present invention, FIG. 5 is a diagram illustrating the operation process in FIG. 4, and FIG. It is a figure which illustrates the reproduction|regeneration waveform in a figure. In the figure, 1 is a reception register, 2 is a memory, 3 is a transmission register, 4 and 6 are counting circuits, 5 and 9 are selection circuits, 7 is a latch circuit, 8 is an arithmetic mean circuit, ck to c
k& is the clock signal, t is the sampling period, V is the stored voice, V, ,V, ,V. , V? is audio data, V4 and v6 are playback audio, v6
indicates the arithmetic mean value. Quasi 1 Eye broom 2 Figure 1 b・Sound reproduction waveform stem 4 Figure $4 tL-I)-ke3 movement yh <= pm 5
Cabinet

Claims (3)

[Claims] (1) In an audio storage device that stores and plays back PCM-encoded audio data, it samples the PCM-encoded audio data at a predetermined period (t), receives the PCM-encoded audio data, and stores it in the memory (200) every other audio data. an extraction means (300) for extracting the audio data stored in the memory (200) every two sampling periods; an interpolation means (400) for creating audio data. (2) The interpolation means (400) immediately before the memory (
200) is used as audio data after one sampling cycle. (3) The interpolation means (400) includes the memory (200
), the arithmetic mean value of two sets of audio data successively extracted from 2 sets of audio data is obtained, and the arithmetic mean value is determined as audio data at an intermediate time point between the extraction times of the two sets of audio data. The audio data storage method described.