JP2002108399A - Voice editing system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To edit audio signals at a high speed and to display a voice level in the state of being coded by AAC(MPEG2 Advanced Audio Coding). SOLUTION: For the audio signals, an adts(audio data transport stream) file 4 is generated which consists of adts frames encoded by an encoder for AAC from an uncompressed state. An adts fixed header and an adts variable header in an adts frame as a minimum unit are analyzed to generate and store time management data different from the adts file 4 in a management table. When editing processing for reproducing the adts file 4 from an arbitrary position is performed, time management data stored in the management table 5 are used to calculate the position of a reproduction point by the processing, etc., of an application program running on a CPU, thereby reads the corresponding adts frame based upon the position. A decoder 7 for AAC performs decoding to output an uncompressed audio data.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to AAC (MPEG2 Advanced Audio C) standardized as ISO 13818-7.
audio editing system that performs an audio signal recording / playback editing process on an audio signal, which is digital audio data compressed and encoded according to oding, without being restored to an uncompressed audio signal. It is about.

[0002]

2. Description of the Related Art In recent years, a method of encoding an audio signal by AAC has been studied from the viewpoint of good compression efficiency and good sound quality. However, in the AAC, since the 1adts frame is not of a fixed length and is used in a broadcasting station, in order to instantly skip a reproduction position or perform an editing process, an adts fixed header or an adts which is a header of encoded data is used. It was necessary to perform reproduction and editing processing while reading the variable header each time.

[0003]

However, in such a conventional audio editing system which performs editing processing such as reproduction of an audio signal encoded by AAC, a conventional audio editing system employs an adts fixed header for all encoded data. adts
Since it is necessary to read out and process the variable header in real time, hardware and software having a high-speed processing capability for decoding are required. In order to jump to an arbitrary reproduction point in the encoded audio signal, the reproduction point is determined after analyzing the information recorded in the adts frame in real time. There was a problem.

[0004] Further, finding an edit point in an audio signal encoded by AAC is not practical because it takes a lot of time to process the data in a conventional system, and it is generally impractical to use once uncompressed data. Editing work is performed after restoring the state, resulting in deterioration of sound quality and delay of processing time.

An object of the present invention is to solve the above-mentioned problems of the prior art, and to provide an audio editing system capable of high-speed editing processing and displaying an audio level in a state encoded by AAC. Aim.

[0006]

In order to achieve this object, an audio editing system according to the present invention encodes an audio signal by AAC and generates a plurality of adt based on AAC.
In an audio editing system that generates an s frame and performs a recording / playback editing process, a management table storing time management data for each adts frame for performing an editing process in an encoded state of an audio signal. Adts frame to be edited based on the time management data
a dts frame at high speed, and a means for generating audio level data obtained by extracting an audio signal peak level in the adts frame at the time of encoding the audio signal. , And / or audio level data is displayed to perform the editing process.

According to the above arrangement, the audio signal AA
By generating time management data from each adts frame in a state encoded by C, the adts file can be edited at a high speed, and the audio level data extracted and generated from the audio signal in the adts frame enables Audio level data can be displayed.

[0008]

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

FIG. 1 is a diagram showing a signal flow for encoding an audio signal of the audio editing system according to the first embodiment of the present invention, wherein digital data from a WAV file player 1 'of a CD (Compact Disc) 1 is shown. Or AES / EBU (Audio Engineering Societ) from playback device 2
An audio signal such as digital data of a y / European Broadcasting Union) is an adt composed of an adts frame encoded by the AAC encoder 3 from an uncompressed state.
An s-file 4 is generated.

In the first embodiment, before editing an audio signal, an adts file obtained by encoding the audio signal by AAC has a minimum unit of adt.
The adts fixed header and the adts variable header in the s frame are analyzed, and time management data different from the adts file is generated and stored in the management table.

FIG. 2 is a schematic diagram showing a search for an editing position (playback point) of an adts file using time management data in a management table according to the first embodiment. For example, when performing an editing process of reproducing from an arbitrary position in the adts file 4, using the time management data stored in the management table 5 generated earlier,
The position of the reproduction point is calculated by the processing of the application program 6 running on the CPU, etc.
read the ts frame, and further add the
The data is written to the AC decoder 7, decoded, and output as an uncompressed audio signal.

FIG. 3 shows a compressed a by AAC encoding.
FIG. 3 is a diagram illustrating the structure of a dts file and an adts frame. As shown in FIG. 3, the adts file is composed of a plurality of adts frames, and one adts frame is:
adts fixed header (adts_fixed_header), adts
Variable header (adts_variable_header), CRC (adts_e
rror_check), byte_alignment, actual data (raw_data_b
lock) and byte_alignment.

[0013] The adsts variable header is copyright_
identification_bit, copyright_identification_sta
t, frame_length, adts_buffer_fullness, number_of_r
It consists of each parameter of aw_data_block_in_frame, in which frame_length contains actual data (raw_data_block) + by
The data size of te_alignment is recorded.

The adts frame is composed of actual data (raw_frame) in one adts frame due to the relationship between the levels before and after the sound.
Although the length of data_block is different, the standard is that the data time reproduced in one adts frame is constant, so that it can be managed by time management data generated from each adts frame.

[0015] Specifically, a encoded by AAC
In order to expand the length of the dts frame as time management data, the contents of the adts fixed header and the adts variable header in the adts frame are analyzed, and the length (number of bits) of one adts frame is set to 2 bytes per one adts frame. The management data is generated and stored in the management table (see FIG. 4).

Length of 1adts frame (number of bits)
Is: adts fixed header (adts_fixed_header) size adts variable header (adts_variable_header) size CRC (The Cyclic Redundancy Check to verify th
e correctness of data; adts_error_check) size byte_alignment frame_length recorded in the adts variable header
The total of (actual data + byte_alignment) can be calculated by (Equation 1).

[0017]

## EQU1 ## The calculation result of 1adts frame length = ++++ (Equation 1) is obtained by sequentially accumulating time management data in the management table from the beginning of the adts file as binary data in units of 2 bytes.
When the data time determined by the sampling frequency of the ts frame is Xmsec and the reproduction point is Ymsec,
The Z-th bit from the top of the ts file can be calculated by (Equation 2).

[0018]

Long Z = 0; For (i = 0; 0 <= i <= Y / X-1; i ++) {Z + = adts frame data (i);} From the playback request time, it is possible to calculate the number of the adts frame from the beginning of the adts file that is the requested adts frame.

From the above, when changing the playback point in the conventional adts file, the fixed header and the variable header in the adts frame are calculated in real time by hardware or software, or the data is uncompressed in advance. However, according to the first embodiment, any processing can be performed without causing the above-described problem. The editing process for reproducing from the position can be performed at high speed.

FIG. 5 is a diagram schematically showing generation of audio level data performed when encoding an audio signal in the audio editing system according to the second embodiment of the present invention. In FIG. 5, input of an audio signal 8 such as a WAV file or AES / EBU is performed in the same manner as in the first embodiment.
Adts encoded by the AAC encoder 3 by the processing of the application program 6 'running on the PU
Immediately before the file 4 is generated, a peak level is extracted as audio level data from the audio signal 8 corresponding to one adts frame, generated for each adts frame, and stored in the level table 9.

The adts frame encoded by the encoder 3 is stored in an adts file 4. The time management data is generated by analyzing the adts frame output from the encoder 3 and performing the same processing as described in the first embodiment, and is stored in the management table 5.

The actual data encoded by the conventional AAC cannot be displayed on a screen or the like because there is no correlation with the audio level data, and the encoded data is uncompressed to be displayed. Needed to be restored to a state. Therefore, by referring to the audio level data generated in the second embodiment, the adts file 4 necessary for the editing process and the audio level corresponding to the adts frame arbitrarily selected from the files are displayed. Becomes possible.

Here, an example in which audio level data is generated by extraction from the audio signal 8 will be described. For example, if the audio data to be extracted is 16 bits,
The record of “it” is expressed between 0 and 7FFF, with the most significant bit being determined as positive or negative. If the minimum unit that can be edited for the data in units of 2 bytes is 0.1 second and sampling is 48 kHz, the maximum data between every 4800 data is stored in the level table 9 as one audio level data. Also, even if only the upper 1 byte of 2-byte data is stored in order to reduce the storage area,
Audio levels can be displayed correctly with considerable accuracy.

FIG. 6 is a diagram showing an outline of an editing process of the audio editing system according to the third embodiment of the present invention.
According to the third embodiment, the management table 5 and the level table 9 in the first and second embodiments allow the editing process to be performed while the adts file 4 is encoded.

Here, the general operation will be described. Since the audio editing system performs the editing processing while encoding the data by the AAC, the adts file 4, the audio level data generated from the audio signal immediately before generating the adts file 4, and the adts frame 4 before the editing processing are performed. adts
The time management data generated from the fixed header and the adts variable header is used. First, the position of the adts frame to be reproduced is calculated based on the time management data in the management table 5 from the reproduction request time. As a result, the corresponding adts frame is read from the adts file 4, and
The audio level data of the ts frame is read from the level table 9, and an editing process for reproducing an audio signal at an arbitrary position is performed. By this processing, the read out adts
In addition to writing and restoring the frame to the decoder 7 and outputting it, the corresponding audio level data is sent to the CRT 10 for display, so that editing can be performed at high speed.

As described above, according to the third embodiment, high-speed editing can be performed without deteriorating the quality by the restoration processing while the encoding is performed by the AAC. And the editing process can be performed more easily.

[0027]

As described above, according to the present invention,
The corresponding adts frame is searched by the time management data generated from the adts fixed header and the adts variable header of one adts frame, and the adts file can be edited at high speed while being encoded by AAC, and the time management can be performed. Simultaneously with the data, the peak level of the audio signal in each adts frame is generated as audio level data, and a silent portion of the audio level, the maximum level, and the like can be obtained without restoring an uncompressed audio signal from a state encoded by AAC. Can be displayed, and the editing process can be simplified.

[Brief description of the drawings]

FIG. 1 is a diagram showing a signal flow for encoding an audio signal of an audio editing system according to Embodiment 1 of the present invention.

FIG. 2 is a diagram showing an outline of an editing position search of an adts file using time management data of a management table according to the first embodiment of the present invention;

FIG. 3 is a diagram illustrating a structure of an adts file and an adts frame compressed by AAC encoding according to the first embodiment of the present invention.

FIG. 4 is a diagram illustrating a structure of a management table according to the first embodiment of the present invention.

FIG. 5 is a diagram schematically illustrating generation of audio level data performed when encoding an audio signal of the audio editing system according to the second embodiment of the present invention.

FIG. 6 is a diagram showing an outline of an editing process of the audio editing system according to the third embodiment of the present invention.

[Explanation of symbols]

 Reference Signs List 1 CD 1 'WAV file reproducer 2 Reproducing device 3 Encoder 4 adts file 5 Management table 6, 6', 6 "Application program 7 Decoder 8 Audio signal 9 Level table 10 CRT

Claims (4)

[Claims] An audio signal is converted to AAC (MPEG2).
Advanced Audio Coding) and the AA
C based on multiple ads (audio data transport str
eam) an audio editing system for generating frames and performing recording / playback editing processing, comprising: a management table storing time management data for each of the adts frames, wherein the audio signal is encoded in the encoded state. An audio editing system characterized in that an editing process is performed by specifying time using time management data. 2. A system comprising: means for quickly searching for an adts frame to be edited from each adts frame based on the time management data;
2. The audio editing system according to claim 1, wherein a 1-adts frame encoded by C is configured as a 2-byte unit and stored in a management table. 3. An audio editing system that encodes an audio signal by AAC, generates a plurality of adts frames based on the AAC, and performs recording / playback editing processing. Means for generating audio level data obtained by extracting the audio signal peak level in the audio signal, and displaying the audio level data in an encoded state of the audio signal to perform editing processing. system. 4. An audio editing system that encodes an audio signal by AAC, generates a plurality of adts frames based on the AAC, and performs recording / playback editing processing, wherein time management data for each of the adts frames is stored. Means for quickly searching for an adts frame to be edited from each of the adts frames based on the management table and the time management data, and encoding a peak level of the audio signal when encoding the audio signal. Means for generating audio level data, and performing editing processing by specifying time using the time management data and / or displaying the audio level data in an encoded state of the audio signal. Characteristic audio editing system.