JPH07183854A - Sound compressed data editing device - Google Patents

Abstract

PURPOSE:To unnecessitate a compressor and an expandor by changing only normalization information and changing the sound volume at the time of a reproduction in sound compressed data having normalization information as information on the amplitude value of a sound signal. CONSTITUTION:In a compressed data reading means 11 for reading sound compressed data, compressed data D 11 is obtained. In a normalization information extracting means 12 for extracting normalization information in the compressed data D 11, normalization information D 12 is obtained. In a normalization information changing means 13 for changing the extracted normalization information D 12, changed normalization information D 13 is obtained. This device is provided with a normalization information writing means 14 for superscribing the changed normalization information D 13 in the compressed data D 11. Thus, by extracting the normalization information D 12 from the compressed data D 11, changing normalization information by adding a constant C to the normalization information D 12 and superscribing the changed normalization information D 13 on the normalization information of the original compressed data D 11, the sound volume at the time of a reproduction is changed.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an MPEG (Moving Picture Expert Group) when compressing an audio signal.
The present invention relates to an apparatus for editing compressed data that includes normalization information in compressed data, such as the sub-band coding method.

[0002]

2. Description of the Related Art MPEG audio signal compression, which is becoming the mainstream of audio compression systems at present, is sub-band coding.
One frame is composed of several milliseconds to several tens of milliseconds.
At this time, in order to increase the compression efficiency, the normalization coefficient is obtained from the maximum amplitude value in the frame to normalize the amplitude value of the audio signal. The normalization coefficient is multiplexed with other data in the compressed data as normalization information. The compressed data includes, in addition to the audio signal, synchronization information included in each frame, compression parameters, the normalization information, and the like.

An example of the prior art for editing the voice compression data thus configured will be described with reference to the block diagram of FIG. First, the compressed data reading means 61
The compressed data D1 is read from the recording medium, and the compressed data decompressing unit 62 decompresses the compressed data D1 to obtain PCM data D2. When the volume is edited, the volume changing means 63 is used to change the amplitude value of the PCM data D2. Then, the PCM data compression means 64 compresses the edited PCM data D2 to obtain compressed data D3. The compressed data D3 is stored in the compressed data writing means 65.
Is written on the recording medium.

[0004]

In the method described above, it is necessary to decompress the compressed data in order to restore it to the PCM data, and the compressed data decompression means is required. Also, since it is necessary to compress the PCM data after changing the volume,
PCM data compression means becomes essential. Further, there is a problem that sound quality is deteriorated because the compression process is additionally performed.

[0005]

In order to solve the above problems, in the present invention, normalization information is extracted from compressed data, a constant C is added to the normalization information, and the normalization information is changed. By overwriting the changed normalization information on the normalization information of the original compressed data, the volume at the time of reproduction is changed.

[0006]

According to the audio compressed data editing apparatus of the present invention, the compressor and the decompressor for changing the volume become unnecessary. Also,
Since editing can be done without going through the compression process again,
You can perform high-quality editing without deterioration of sound quality.

[0007]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is a block diagram for explaining an embodiment of an audio compressed data editing apparatus according to the present invention. In FIG. 1, reference numeral 11 denotes a compressed data reading means for reading the compressed audio data, which allows the compressed data D11 to be read.
To get Reference numeral 12 is a normalization information extraction means for extracting the normalization information in the compressed data, and obtains the normalization information D12. Reference numeral 13 is a normalization information changing unit for changing the extracted normalization information, and obtains the changed normalization information D13. 14 is the modified normalization information D13 which is compressed data D1.
It is a normalized information writing means for overwriting the data in 1.

The normalized information extracting means 12 will be described with reference to FIG. FIG. 2 is a frame diagram showing an example of the contents of audio compression data in the MPEG subband coding system. The compressed data is composed of frames of several milliseconds to several tens of milliseconds, and each frame has a header 201, bit allocation information 202, and normalization information 20 from the beginning.
3 and audio signal 204 are arranged in this order. Header 201
Is composed of 12-bit synchronization information 205 and 20-bit compression parameter 206. Also,
The bit allocation information 202 represents the quantized bits of each subband with 4 bits. The normalization information 203 is replaced with a scale factor index corresponding to the normalization coefficient of each subband, and this is designated as SCF, which is represented by 6 bits.
The larger the normalization coefficient, the smaller the SCF, and conversely, the smaller the normalization coefficient, the larger the SCF.

The header 201 is 32 bits, and if the number of subbands is N, the bit allocation information 202 is (4 × N) bits. Therefore, the synchronization information 205
Is detected, the normalized information is (32 + 4 × N) bits ahead of the header, which allows the normalized information extracting means 12 to extract the normalized information.
If the quantization bit of each subband is not 0, the size of the normalization information 203 is (6 × n), where SCF has the SCF corresponding to that subband and the number of subbands of the nonzero quantization bit is n. It can be seen that it is a bit, and the normalized information can be extracted.

A first embodiment of the normalization information changing means 13 will be described with reference to the flowchart of FIG. In step 301, a constant C and the number n of SCFs to be added to the SCF of each subband are set, and the upper limit and the adjustment of the value that the SCF can take are MAX and MIN, respectively.
In step 302, the counter i is set to 0, and in step 303, the constant C is added to the i-th SCF.
If SCF is smaller than MIN in step 304, SCF is clipped to MIN in step 305, and if SCF is larger than MAX in step 306, SCF is clipped to MAX in step 307. I is added in step 308, and step 309
Check if all SCFs have been changed. Step 309
If the condition is true, steps 303 to 308
If the condition of step 309 is false, the presence of the next frame is judged in step 310. If the condition is true, steps 302 to 309 are repeated, and if the condition is false, the process ends.

A second embodiment of the normalization information changing means 13 will be described below with reference to the flow chart of FIG. In step 401, the constant C is set to 0 and SC
The number n of F is set, and the upper limit and the addition / subtraction of the value that the SCF can take are MAX and MIN, respectively. Step 402
At 1, the constant C is incremented by 1 and the counter i is set at 0. In step 403, the constant C is added to the i-th SCF.
Add. If SCF is less than MIN in step 404, SCF is clipped to MIN in step 405, and if SCF is greater than MAX in step 406, SCF is MAX in step 407.
Clip to. In step 408, i is added, and in step 409, it is determined whether all SCFs have been changed. If the condition in step 409 is true, steps 403 to 408 are repeated, and if the condition in step 409 is false, it is determined in step 410 whether all SCFs have become MAX. If the condition of step 410 is false,
In step 411, the presence of the next frame is judged,
If the condition is true, steps 402 to 410 are repeated. If the condition of step 410 is true or the condition of step 411 is false, the process ends.

In the above two embodiments, the SCF is changed by adding the constant C, but the SCF can be similarly changed by the bit shift by the constant C.

FIG. 5B shows a waveform at the time of reproducing the compressed data when the constant C is set to 3 and the first embodiment is used.
(A) is the waveform without editing,
It can be seen that the amplitude value of (b) is half that of (a). (C) is a waveform at the time of reproducing the compressed data when the second embodiment is used. Since the constant C changes over time, the amplitude gradually decreases, and the fade-out effect is realized.

[0014]

As described above, according to the present invention, in the compressed data of voice having the normalization information as the information regarding the amplitude value of the voice signal, the volume at the time of reproduction is changed by changing only the normalization information. , A compressor and an expander are not required, and the circuit configuration is simplified. Further, since the process of expansion and compression is not performed, the volume can be changed without deterioration of sound quality.

[Brief description of drawings]

FIG. 1 is a block diagram of an embodiment of an audio compressed data editing device of the present invention.

FIG. 2 is a frame diagram for explaining an example of the content of compressed data.

FIG. 3 is a flowchart for explaining an embodiment of the audio compressed data editing device of the present invention.

FIG. 4 is a flowchart for explaining one embodiment of the audio compressed data editing device of the present invention.

FIG. 5 is a characteristic example diagram for explaining a change in waveform when the audio compressed data editing device of the present invention is used.

FIG. 6 is a block diagram showing a configuration example of a conventional method for editing compressed audio data.

[Explanation of symbols]

 11 Compressed Data Reading Means 12 Normalized Information Extracting Means 13 Normalized Information Changing Means 14 Normalized Information Writing Means D11 Compressed Data D12 Normalized Information D13 Modified Normalized Information 201 Header 202 Bit Allocation Information 203 Normalized Information 204 Voice Signals 205 synchronization information 206 compression parameter 207 first subband quantization bit 208 second subband quantization bit 209 Nth subband quantization bit 210 first subband normalization coefficient 211 second sub Band normalization factor 212 normalization factor for the nth subband

Claims (3)

[Claims] 1. In compressed audio data having normalization information as information on the amplitude value of a voice signal, compressed data reading means for reading the compressed data, normalization information extraction means for extracting the normalization information, and An audio compressed data editing apparatus having a normalization information changing unit for changing the normalization information by adding a constant C, and a normalization information writing unit for writing the changed normalization information in compressed data. 2. The audio compressed data editing apparatus according to claim 1, wherein the constant C added to the normalization information extracted from the audio compressed data is continuously changed. 3. The audio compressed data editing apparatus according to claim 1, wherein the normalization information changing means changes the normalization information by bit shifting with a constant C.