JP2003005798A - Recorder and reproducing device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a recorder and a reproducing device for recording or reproducing of an audio signal, which perform a muting processing without requiring extra constitution or processing. SOLUTION: Since the recording device is provided with an encoder 41 which generates encoded data by encoding data, a no-signal bit stream forming and CRC(cyclic redundancy check) check adding part 53 which generates no-sound data having the same compression rate as encoding in the encoder 41, and a data selector 55 which adds no-sound data to encoded data or substitutes encoded data with no-sound data at a prescribed time of encoding in the encoder 41, it is unnecessary to perform continuous processing of fade-in and fade-out in encoding, and processing for decoding is simplified.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a recording device and a reproducing device for recording or reproducing an encoded audio signal, for example.

[0002]

2. Description of the Related Art Conventionally, when starting and ending an audio frame, and when decoding an audio frame that was error data, after the audio data that is once ended is faded out and muted,
A sequence was used that fades in the audio data that starts next.

[0003]

However, the above-described conventional sequence has a disadvantage that extra processing such as fade-out, mute and fade-in is required.

Further, since the audio frame is faded out, there is an inconvenience that an extra memory is required to hold the data corresponding to the fade out.

Therefore, the present invention has been made in view of the above points, and a recording device and a reproducing device for recording or reproducing an audio signal capable of performing a mute process without requiring an extra configuration or process. The challenge is to provide.

[0006]

A recording apparatus of the present invention generates coding data by coding data, and silent data having the same compression ratio as the coding in the coding means. And a selection unit for adding silence data to the encoded data or replacing the encoded data with the silence data at a predetermined encoding time in the encoding unit.

Further, the reproducing apparatus of the present invention has a decoding means for generating decoded data by decoding the encoded data, a silent data generating means for generating silent data, and a predetermined input of the encoded data. Sometimes, it is provided with a selecting means for adding silence data to the encoded data or replacing the encoded data with the silence data.

Therefore, according to the present invention, the following operations are performed. First, at the start of encoding, silent data is added before the head data. At the start of encoding, the head data is replaced with silent data. At the end of encoding, silent data is added after the final data.
At the end of encoding, the final data is replaced with silent data. Further, at the time of encoding, the error data is replaced with silent data.

Next, at the start of decoding, silent data is added before the head data. At the start of decoding, the head data is replaced with silent data. At the end of decoding, silent data is added after the final data. At the end of decoding, the final data is replaced with silent data. At the time of decoding, error data is replaced with silent data.

[0010]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below. The recording device and the reproducing device of the present embodiment are M
PEG (Moving Picture Expert)
s Group) When an encoding error is detected in audio signal processing, no-signal data is added to the encoded bitstream and joined together, thereby removing annoying noise due to a DC step at the joint of the bitstream during decoding. It is possible to obtain the same effect as mute, and when a bitstream that causes a decoding error is input, the same error frame is switched to a bitstream compressed with no signal data and decoded. It is an effect.

FIG. 1 is a block diagram of a voice recording process of the recording / reproducing system. In FIG. 1, an audio signal from a microphone and an external device indicated by 1 is an MPEG composed of a bit stream compressed through a process of sampling 2, sub-band division 3, scaling 4, quantization 5, and framing 6.
Converted to audio frames.

This compressed bit stream is subjected to frame check 7, PES (Packetized El).
Elementary Stream) 8, TS (Tra)
nSport Stream) 9, MUX (Multi
Then, the data is recorded on the tape 13 in a predetermined format through the Iplexer 10, the parity generation 11, and the recording coding 12.

FIG. 2 is a block diagram of an audio reproduction process of the recording / reproduction system. When playing back in reverse, in FIG.
The transport stream reproduced from the tape 33 includes a reproduction decoding 32, an error correction 31, a DeM.
UX (DeMultiplexer) 30, PID analysis 29, TS analysis 28, PES analysis 27, ES (Elem
Entry Stream) processing 26 is performed.

After that, the frame check 25, the dequantization 24, the descaling 23, and the subband synthesis 22 are performed to output the PCM data, which is reproduced by the speaker 21 or the like.

In the present embodiment, when the framing 6 or the frame check 7 in FIG.
When the frame check 25 in (1) is implemented, an audio frame of no signal data is added or the audio frame is replaced with an audio frame of no signal data. As a result, the mute process can be performed on the audio data.

First, the first embodiment will be described. First, a case where no signal data is added before the first audio bitstream at the start of encoding will be described. FIG. 3 is a diagram showing the MPEG1 audio encoder layer 1/2. In FIG. 3, a microphone and an audio signal input from the outside are digitally converted by an A / D conversion circuit (not shown),
It is input to the encoder 41 as 16-bit digital audio input data Ai.

The 16-bit digital audio input data Ai input to the encoder 41 is temporarily stored in the PCM buffer 42. The PCM data stored in the PCM buffer 42 is encoded with the number of samples for one frame as one block. The encoded bitstream is encoded error information Er1.
After receiving the frame check 54 and confirming that the frame is normal, the data is output from the data select 55 as the data output Do.

Detailed processing of the encoder 41 will be described with reference to FIGS. FIG. 5 is a diagram showing a subband analysis filter. FIG. 5 is a diagram showing the operation of the subband analysis filter bank 43 in FIG. In FIG. 5, the full band input signal I is a band pass filter (BPF0 to BP0) formed by a polyphase filter bank (PFB).
F31) 71 is used to divide the entire band into 32 subbands (B0 to B31) 72 at equal intervals. Note that one frame has 12 samples × 32 subbands = 384 samples.

Next, the operations of the scale factor selection information 44 and the scale factor extraction 46 in FIG. 3 will be described. As the scale factor, the maximum level value in 12 samples is selected for each band, and the larger value and the smallest value among them are selected from the conversion table. It should be noted that it is 0 when no bit is assigned. In layer 2, 1152 samples (3 × 12 × 32) have three scale factors in one subband. Two difference data are calculated from the three scale factors, and the result is classified into five states. Since the transmission pattern is related to three elements, the relational expression has four levels (0, 1, 2,
It is divided into 3) and the number and position of scale factors are shown as scale factor selection information. The sample is normalized by being divided by the scale factor.

FIG. 6 is a diagram showing a psychoacoustic model.
FIG. 6 shows an FFT (Fast Furie) shown in FIG.
The operations of the Transfer) 45, the psychoacoustic model 47, and the dynamic bit allocation 48 will be described. FFT analysis of the input signal is performed in parallel with the subband filter. The sound pressure is calculated for each subband. The minimum audible threshold characteristic 83 is obtained by calculating the absolute audible threshold in the quiet section. Extract pure tone components (tones) and non-pure tone components (non-tones). The pure tone component and the non-pure tone component are thinned out, and the masking threshold is calculated according to the tone property and the non-tone property. The minimum masking level 82 is determined by calculating the global masking threshold. Signal-to-mask ratio SMR
By calculating (Signal to Mask Ratio), the bit allocation is SM as shown by 81.
It is performed based on R.

The masking effect is as follows. A weak sound (Makee) in the vicinity of a strong sound (Masker)
I can't hear. Forward masking is that you can't hear a loud sound immediately after a strong one. The temporal masking effect works just before a strong sound, and the fact that you cannot hear a small sound is called backward masking. The minimum audibility limit is the minimum level of sound that can be detected by the auditory sense of silence, and sounds below the minimum audible threshold cannot be heard.

FIG. 7 is a diagram showing masking levels. In FIG. 7, the masking level 93 is a combination of the masking effect and the minimum audible threshold. By calculating the signal level 91 to mask ratio SMR 92, bit allocation is done based on SMR.

FIG. 8 is a diagram showing bit allocation.
In FIG. 8, the number of quantization bits to be assigned to each subband 102 is determined within the range that can be used in one frame from SMR. For a signal having a level higher than the minimum audible characteristic and higher than the masking level, bits are allocated as indicated by 104, and the small band of the SMR 103 is 101.
Reduce the bit allocation as shown in.

Next, the operations of the linear quantizer 49, bit compression 50 and side information coding 51 in FIG. 3 will be described. As shown in Equation 1 below, Sqi = quantized sample for each subband, A, B = quantization constant, sc
If f = scale factor and SDi = sample of each sub-band, the sample Si of each sub-band is divided by the scale factor scf and normalized, and then the quantization constant A corresponding to the number of bits assigned to each sub-band. ，
The quantized sample Sqi for each subband is calculated from the parameters such as B.

[0025]

## EQU1 ## Sqi = [A (Si / scf) + B]

FIG. 9 is a diagram showing bitstream formation. FIG. 9 shows the bitstream formation and CRC (Cyclic Redundancy Ch) in FIG.
FIG. 7 is a diagram showing an operation of an eck) check addition 52. Figure 9
In addition, header information is added to coded data (sample) and auxiliary data (scale factor, bit allocation, etc.), and 32 bands / 2 channels worth of data is multiplexed in frame order in frame units. Layer 1 includes a header (32 bits) 111, an error check 112, a bit allocation (4 bits) 113, a scale factor (6 bits) 114, a sample (38 bits × 32) 1.
And 15. Layer 2 has a header (32
(Bit) 116, error check 117, allocation 118, scale factor selection information 119, scale factor 120, and sample 121.

Here, when the bit stream of the no signal data is added to the head frame in the no signal bit stream formation and CRC check addition 53 at the time of encoding disclosure, the timing is one frame earlier than the original encoding start time. The encoding is started, the encoding process is performed with the digital data 41 that clears the PCM buffer 42, and the data output Do is output from the bit stream of the no signal data generated by the no signal bit stream formation and the CRC check addition 53.

From the next frame onward, the PCM data continuously input is used to output the encoded bitstream from the bitstream formation and CRC check addition 52.

Here, the method of generating the bit stream of the no-signal data is as follows: when the PCM buffer 42 is cleared and then the encoding processing is performed,
There are two cases in which encoding processing is performed when silent digital data is given to the input PCM.

Further, if the compression rate is fixed, it is possible to write a bit stream of no-signal data prepared in advance in a ROM or the like for use.

Next, a second embodiment will be described. The operation of replacing the first audio bitstream with the bitstream of no signal data at the start of encoding will be described. In the data select 55 shown in FIG. 3, when outputting the encoded first audio bitstream from the bitstream formation and CRC check addition 52, no signal data generated by the no signal bitstream formation and CRC check addition 53 is output. Replace and output.

Next, a third embodiment will be described. An operation of adding a bit stream of no signal data after the final audio bit stream at the end of encoding will be described. In the data select 55 shown in FIG. 3, bitstream formation and CRC are performed.
After outputting the final audio bit stream encoded from the check addition 52, the no signal data generated by the no signal bit stream formation and CRC check addition 53 is added and output.

Next, a fourth embodiment will be described. The operation of replacing the final audio bitstream with a bitstream of no signal data at the end of encoding will be described. In the data select 55 shown in FIG. 3, when outputting the final audio bitstream encoded from the bitstream formation and CRC check addition 52, it is placed in the no signal data generated by the no signal bitstream formation and CRC check addition 53. Replace and output.

Next, a fifth embodiment will be described. The operation of replacing the error bitstream with the bitstream of no signal data at the time of encoding will be described. Encoding error information Er in the encoding process in the frame check 54 shown in FIG.
When outputting the encoded error stream from the bitstream formation and CRC check addition 52 in the data select 55 for the stream in which the encoding error is confirmed by 1, no signal generated by the bitstream formation and CRC check addition 53 Replace with data and output.

Next, a sixth embodiment will be described. An operation of adding a bit stream of no signal data before the first audio bit stream at the start of decoding will be described. FIG. 4 is a diagram showing an MPEG audio decoder. In FIG. 4, the data input Di of the audio bit stream extracted from the signal processing system is stored in the input buffer 67 of the decoder 59 after the frame check in the frame check 69.

The audio bit stream stored in the input buffer 67 is supplied to the separator 66, and is converted into a data portion in which the information necessary for data decoding by the scale factor extraction 65 and the bit allocation by the dynamic bit allocation control 64 are performed. Be divided.

Data is decoded from the above information by the data decoding 63, converted into PCM through the sub-band analysis filter inverse bank 62 and stored in the PCM output buffer 61,
It is output as a 16-bit audio output Ao.

Here, when a bit stream of no signal data is added by the no signal bit stream addition 70 at the start of decoding, the decoder 59 is at a timing one frame earlier than the original decoding start time.
The decoding is started, and the no signal bit stream is added to the input buffer 67 via the data select 68 to set the no signal bit stream.

In the following frames and thereafter, decoding is performed using the audio bit stream continuously input.

Here, regarding the bit stream of no signal data, if the compression rate (layer, bit rate, sampling frequency) is fixed or if the input compression rate is limited to some types, the bit of no signal data is The stream is stored in ROM and read from there for use. If the compression rate of the input bitstream is not known, the decoding processing by the decoder 59 in FIG. 4 and the encoding processing by the encoder 41 in FIG. 3 are performed at the same time to generate no signal data that matches the compression rate of the bitstream to be decoded. Can be produced and replaces all data with no signal.

If the decoding start is a bit stream of no signal data, this process is not necessary.

Next, a seventh embodiment will be described. The operation of replacing the first audio bitstream with the bitstream of no signal data at the start of decoding will be described.

In the data select 68 shown in FIG. 4, when the first audio bit stream is input,
It is replaced with the no signal data generated by the no signal bit stream formation 70 and decoded.

Next, an eighth embodiment will be described. An operation of adding a bit stream of no signal data after the final audio bit stream at the end of decoding will be described.

In the data select 68 shown in FIG. 4, after inputting the final audio bit stream,
The no signal data generated by the no signal bit stream formation 70 is input to the input buffer 67 and decoded.

Next, a ninth embodiment will be described. The operation of replacing the final audio bitstream with a bitstream of no signal data at the end of decoding will be described.

In the data select 68 shown in FIG. 4, when the final audio bit stream is input,
It is replaced with the no signal data generated by the no signal bit stream formation 70 and decoded.

Next, a tenth embodiment will be described. The operation of replacing the error bit stream with the bit stream of no signal data at the time of decoding will be described.

In the frame check 54 shown in FIG. 4, the stream in which the decoding error is confirmed by the error information Er3 or the decoding error Er2 in the decoder 59 is converted into the no signal data generated by the no signal bit stream formation 70 in the data select 68. Replace and enter.

The recording / reproducing system described above can be realized by both hardware and software.

Needless to say, the recording / reproducing system described above can be applied not only to the magnetic tape recording / reproducing apparatus but also to another recording / reproducing apparatus having an MPEG audio encoding or decoding function.

According to the present embodiment described above, it is not necessary to perform continuous fade-in and fade-out processing, control can be simplified, and reproduction can be performed without performing these processing at the data joint. You can

[0053]

According to the recording apparatus of the present invention, the encoding means for generating encoded data by encoding the data and the silent data generation for producing the silent data having the same compression ratio as the encoding in the encoding means. Means and adding silent data to the encoded data at a predetermined encoding time in the encoding means,
Alternatively, since it is provided with a selection means for replacing encoded data with silent data, there is no need to perform continuous fade-in and fade-out processing at the time of encoding, the processing at the time of decoding can be simplified, and DC steps It is possible to suppress the occurrence and improve the recording quality.

Further, the recording apparatus of the present invention has the effect that the recording quality at the start of encoding can be improved because the above-mentioned predetermined time is the start of encoding by the encoding means.

Further, the recording apparatus of the present invention has an effect that the recording quality at the end of encoding can be improved since the predetermined time is the end of encoding by the encoding means.

Further, in the above-described recording apparatus of the present invention, when the coding error is detected in the coding means, the selecting means replaces the coded data with the silent data. Therefore, the frame in which the error is detected is replaced with the silent data. Thus, it is possible to improve the recording quality when an error is detected.

Further, the reproducing apparatus of the present invention has a decoding means for generating decoded data by decoding the encoded data, a silent data generating means for generating silent data, and a predetermined input of the encoded data. Since it is equipped with a selection means for adding silent data to encoded data or replacing encoded data with silent data at times, there is no need to perform continuous processing of fade-in and fade-out during decoding,
It is possible to simplify the processing at the time of decoding, suppress the occurrence of a DC step, and improve the reproduction quality.

Further, in the above-mentioned reproducing apparatus of the present invention, since the predetermined time is the beginning input of the encoded data,
It is possible to improve the reproduction quality when the head of the encoded data is input.

Further, in the reproducing apparatus of the present invention, in the above description, since the predetermined time is the final input of the encoded data,
It is possible to improve the reproduction quality at the final input of encoded data.

Further, in the above-described reproducing apparatus of the present invention, when the decoding means detects the decoding error, the selecting means replaces the coded data with the silent data. Therefore, the frame in which the error is detected is replaced with the silent data. Thus, it is possible to improve the reproduction quality when an error is detected.

[Brief description of drawings]

FIG. 1 is a block diagram of an audio recording process of a recording / reproducing system applied to this embodiment.

FIG. 2 is a block diagram of audio reproduction processing of the recording / reproduction system applied to the present embodiment.

FIG. 3 MPEG1 audio encoder layer 1/2
FIG.

FIG. 4 is a diagram showing an MPEG audio decoder.

FIG. 5 is a diagram showing a subband analysis filter.

FIG. 6 is a diagram showing a psychoacoustic model.

FIG. 7 is a diagram showing masking levels.

FIG. 8 is a diagram showing bit allocation.

FIG. 9 is a diagram showing bitstream formation.

[Explanation of symbols]

1 ... Microphone, 2 ... Sampling, 3 ... Subband division, 4 ... Scaling, 5 ... Quantization, 6 ... Frame conversion, 7 ... Frame check, 8 ... PES conversion, 9
...... TS conversion, 10 MUX, 11 parity generation, 12 recording coding, 13 tape, 21
...... Speaker, 22 ...... Subband synthesis, 23 …… Inverse scaling, 24 …… Inverse quantization, 25 …… Frame check, 26 …… ES conversion, 27 …… PES analysis, 28 ……
TS analysis, 29 ... PID analysis, 30 ... DeMUX,
31 ... Error correction, 32 ... Playback decoding, 3
3 ... tape, 41 ... encoder, 53 ... no signal bitstream formation and CRC check addition,
54 ... frame check, 55 ... data select,
59 ... Decoder, 70 ... No signal bit stream formation, 69 ... Frame check, 68 ... Data select,

Front page continuation (51) Int.Cl. ⁷ Identification code FI theme code (reference) G11B 20/18 572 G11B 20/18 574D G10L 7/04 G 574 9/18 HM

Claims (8)

[Claims] 1. A coding means for generating coded data by coding data, a silent data generation means for generating silent data having the same compression ratio as the coding in the coding means, and the above coding. A recording device comprising a selecting means for adding the silent data to the encoded data or replacing the encoded data with the silent data at a predetermined time of encoding by the means. 2. The recording apparatus according to claim 1, wherein the predetermined time is a start time of encoding by the encoding means. 3. The recording apparatus according to claim 1, wherein the predetermined time is at the end of encoding by the encoding means. 4. The recording apparatus according to claim 1, wherein when the encoding error is detected by the encoding means, the selecting means replaces the encoded data with the silent data. 5. Decoding means for generating decoded data by decoding the coded data, silence data generating means for generating silence data, and converting the encoded data into the encoded data at a predetermined input time. A reproducing device provided with a selecting unit for adding the silent data or replacing the encoded data with the silent data. 6. The reproducing apparatus according to claim 5, wherein the predetermined time is when the head of the encoded data is input. 7. The reproducing apparatus according to claim 5, wherein the predetermined time is a final input of the encoded data. 8. The reproducing apparatus according to claim 5, wherein the selecting means replaces the encoded data with the silent data when a decoding error is detected in the decoding means.