JPH0855428A - Sound recording signal processor - Google Patents

Abstract

PURPOSE:To make it possible to deal with large-sound volume input with a digital sound recorder without impairing sound quality. CONSTITUTION:Adequate digital data is obtd. even for the large-sound volume input by widening a dynamic range by an A/D conversion means 2 of >=n+1 bit. This digital data is delayed by a delaying means 6 and is supplied to an arithmetic means 5. On the other hand, the output of this A/D conversion means 2 is subjected to level detection by a level detecting means 4, by which the large-sound volume input is detected. An arithmetic control means 7 stepwise changes the limiter characteristic or attenuation characteristic of the arithmetic circuit 5 before the large-sound volume data is inputted with a delay into the arithmetic circuit 5 according thereto.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a recording signal processing apparatus for supplying digital audio data to a recording apparatus section which performs predetermined processing on quantized n-bit digital audio data and records it on a recording medium. Is.

[0002]

2. Description of the Related Art DAT recorder and MD (mini disk)
In recent years, so-called digital recorders, such as recorders, that record digital audio data have become popular. For such a digital recorder, for example, digital audio data is supplied from a CD player or the like to be recorded, and an analog audio signal is input from a CD player, an analog tape player, a microphone, etc. Then, A / D conversion is performed to record on a recording medium.

[0003]

By the way, in recent years, so-called live recording, in which music played by a user himself, live music of another person, various natural sounds and environmental sounds is recorded, has been actively performed. Digital recorders are widely used.

Here, there is a circumstance that it is difficult to set a recording level in a digital recorder as compared with an analog recorder. This is because an analog recorder has a so-called soft clip that gradually saturates even if a level over occurs due to a high volume input, so even if the recording level is set to a rough level to a certain degree It doesn't sound bad. However, in a digital recorder, a large volume that exceeds a dynamic range determined by the number of quantization bits, that is, an input voice of 0 dB or more cannot be expressed as digital data, and thus has an input / output characteristic called a hard clip, which results in a saturated voice. Is because the sound becomes unpleasant to the ear with many harmonic components. The input / output characteristics of such an analog recorder and a digital recorder are shown in FIG. 6 (a). 6B shows the distortion of the recorded voice waveform due to the soft clip and the hard clip of FIG. 6A.

In live performance recording and the like, the maximum volume must be predicted, and in such a case, if it is desired to avoid the occurrence of hard clips, the recording level must be set low. However, in this case, recording cannot be performed while making full use of the dynamic range of the digital recorder.

Further, in general, in order to handle a high level input voice, regardless of whether it is an analog recorder or a digital recorder,
Many devices are equipped with limiters. This limiter is
The input / output characteristic of the soft clip in FIG. 6A is to be realized by an electric circuit using an analog non-linear element. FIG. 7 shows the input / output characteristics when this limiter is used in a digital recorder. The solid line indicates the limiter off state, that is, the input / output characteristic of the hard clip, and the input / output characteristic of the soft clip as shown by the broken line is obtained by operating the limiter.

However, the use of such a limiter has the following problems. First, there is no analog non-linear element with high performance in terms of sound quality, and noise components increase, resulting in deterioration of sound quality. Therefore, even a digital recorder capable of recording with high sound quality cannot utilize its performance. Also, the limiter circuit distorts the area up to 0 dB due to its operating principle, so that there is a problem in that even an input voice that does not exceed 0 dB is affected.

[0008]

SUMMARY OF THE INVENTION The present invention has been made in view of these problems, and in a digital recorder,
An object of the present invention is to provide a recording signal processing device capable of handling a large volume input without impairing sound quality.

That is, an A / D conversion means is provided as a recording signal processing device for supplying digital audio data to a recording device section for performing predetermined processing on quantized n-bit digital audio data and recording it on a recording medium. A level detection means, a delay means, a calculation means, and a calculation control means are provided. The A / D conversion means outputs (n
It shall be converted into digital audio data with a quantization accuracy of +1) bits or more. The level detecting means detects the level of the digital audio data output from the A / D converting means. The delay means delays and outputs the digital audio data output from the A / D conversion means. The arithmetic means performs an arithmetic operation on the digital audio data output from the delay means with the arithmetic operation control value supplied, and sets the output level to a value within the dynamic range of quantized n bits, and the digital audio of quantized n bits. To be able to output data. The arithmetic control means enables the arithmetic control value supplied to the arithmetic means to be changed stepwise according to the detection value of the level detecting means.

Further, the arithmetic control means gradually changes the arithmetic control value, and when a certain signal attenuation amount is set as the input / output characteristic of the arithmetic means, the signal attenuation amount is continuously set for a required period. The operation control value to be supplied is fixed. Holds so-called attenuation characteristics.

Alternatively, the arithmetic control unit gradually changes the arithmetic control value, and after a certain signal attenuation amount is set as the input / output characteristic of the arithmetic unit, the arithmetic control unit changes in stages according to the detection value of the level detecting unit. The supplied calculation control value is gradually changed so as to reduce the signal attenuation amount by the calculation means and restore the normal input / output characteristic state. The so-called attenuation characteristic is recovered.

[0012]

First, the A / D conversion means of n + 1 bits or more widens the dynamic range to obtain proper digital data even for a large volume input. Then, this is delayed by the delay means and supplied to the arithmetic means. On the other hand, the level detection means detects the level of the output of the A / D conversion means to detect a large volume input. In response to this, the arithmetic control means gradually changes the limiter characteristic or the attenuation characteristic in the arithmetic circuit until the large volume data is delayed and input to the arithmetic circuit. Then, with respect to the large volume data, necessary limit processing or attenuation processing is executed in the arithmetic circuit, and the limit level or attenuation level is gradually changed until that point, so that sound quality deterioration due to so-called hard clip operation does not occur. It will not happen.

[0013]

Embodiments of the present invention will be described below with reference to FIGS. FIG. 1 is a block diagram of a recording signal processing device of the embodiment. In FIG. 1, reference numeral 10 denotes a sound recording circuit section which performs a predetermined process on the input 16-bit digital data and records the data on a recording medium (disk or tape). The recording signal processing device of this embodiment is a device for processing an input analog audio signal and supplying 16-bit digital data to the recording circuit section 10.

Reference numeral 1 is an input terminal, for example, a microphone is connected to input an analog audio signal. Reference numeral 2 denotes an A / D converter with a quantized 20-bit precision. The analog audio signal input from the input terminal 1 is output by the A / D converter 2 to 20
Bit digital data A ₂₀ . The 20-bit digital data A ₂₀ is supplied to the DSP (digital signal processor) 3.

In the DSP 3, reference numeral 4 is a level detecting section, and 5 is a calculating section. Further, 6 is a RAM formed of, for example, a D-RAM, and is connected to the DSP 3. The 20-bit digital data A ₂₀ is input to the level detector 4 and the level is detected. Here, the level detection unit 4 determines that the value of the 17th bit to the 20th bit of the 20-bit digital data A ₂₀ is “1”, so that the input audio data exceeds the dynamic range of 16 bits. Detect that you are doing. 17th bit ~ 2nd
Bit information up to the 0th bit, that is, level detection information C _DT
Is supplied to a controller 7 formed by a microcomputer.

The 20-bit digital data A ₂₀ is R
Written to AM6. Then, it is read out after a predetermined time and supplied to the arithmetic unit 5 as the delay signal A _20D . That is, the RAM 6 is used as a delay unit for the 20-bit digital data A ₂₀ . Writing / reading and address control in the RAM 6 are executed by the controller 7.

The arithmetic unit 5 is responsive to the arithmetic control value C _L supplied from the controller 7 to obtain 20-bit digital data A.
Performs arithmetic processing for ₂₀ . And after that calculation,
16 bits of data from the 1st bit to the 16th bit
The bit digital data A ₁₆ is output to the recording circuit section 10.

That is, the controller 7 is configured to set the operation control value C _L according to the level detection information C _DT and supply it to the operation circuit 5 to control the input / output characteristics.

The operation unit 20 that can be realized in this embodiment is provided with 20.
There are adaptive level attenuation processing and adaptive limiter processing as the bit → 16 bit conversion processing. First, the adaptive level attenuation process will be described.

Generally, when the input sound becomes over-level during recording on the recorder, the user manually lowers the recording level volume slowly, but the adaptive level attenuation process is It can be said that it is automatically performed by digital processing.

The arithmetic unit 5 multiplies the 20-bit digital data A ₂₀ by the arithmetic control value C _L , that is, the multiplication coefficient, given from the controller 7, so that AT ₁ , AT ₂ , AT ₃ ,. The input / output characteristics can be changed as shown by. That is, the attenuation amount is variably set by the controller 7. AT ₁ is the case where the attenuation is 0, and AT ₂ , AT ₃ ... Attenuates gradually.

When a level exceeding the 16-bit dynamic range is detected by the level detection information C _DT , the controller 7 gradually changes the multiplication coefficient for the arithmetic unit 5 according to the detected level. That is, the target of the attenuation characteristic is set according to the excess level, and the attenuation characteristic is changed until the large volume input data is delayed via the RAM 6 and input to the arithmetic unit 5. That is, when the level over of 6 dB is detected, if the arithmetic unit 5 attenuates by -6 dB, the overflow over the 16-bit dynamic range can be prevented. Then, by gradually increasing the amount of attenuation until the attenuation up to −6 dB is performed, there is almost no discomfort in terms of sound quality.

For example, if the 17th bit of the 20-bit digital data A ₂₀ is "1" and the 18th to 20th bits are "0", and the excess level is relatively small, for example, AT _{With 3} as the target attenuation characteristic, the multiplication coefficient for the arithmetic unit 5 is changed stepwise during the delay time by the RAM 6, and the attenuation characteristic by the arithmetic unit 5 is changed as AT ₁ → AT ₂ → AT ₃ .

On the other hand, when the 20th bit of the 20-bit digital data A ₂₀ is "1" and the excess level is large, for example, AT ₆ is set as the target attenuation characteristic and the delay time by the RAM ₆ is delayed. , The multiplication coefficient for the arithmetic unit 5 is changed stepwise, and the attenuation characteristic by the arithmetic unit 5 is changed to AT ₁ → AT ₂ → AT ₃ → AT ₄ → A
Change from T _{5 to} AT ₆ .

This operation is shown in FIG. 20 is shown in FIG.
The bit digital data A ₂₀ and the delayed signal A _20D are shown. 20-bit digital data A ₂₀ is 1 until time t ₀
Since the value is within the 6-bit dynamic range, the controller 7 supplies the multiplication coefficient 1 to the arithmetic unit 5. That is, input / output is executed with the characteristics of AT ₁ as shown in FIG. Then, the arithmetic unit 5 takes out the lower 16 bits of 20 bits and outputs 16-bit digital data A ₁₆ .

20-bit digital data A ₂₀ at time t ₀
Becomes a value that causes a level over with respect to the 16-bit dynamic range. The level-over 20-bit digital data A ₂₀ is input to the arithmetic unit 5 as the delay signal A _{20 D at} time t ₁ after the time Td. Td represents a delay time by the RAM 6.

Therefore, the controller 7 gradually changes the multiplication coefficient by the time t ₁ as shown in FIG. 3B, and changes the attenuation characteristic of the arithmetic unit 5 as AT ₁ → AT ₂ → AT _3. Go. As a result, the level-over 20-bit digital data A ₂₀ becomes a value that falls within the 16-bit dynamic range due to the attenuation processing of the arithmetic unit 5. Then, the arithmetic unit 5 extracts the lower 16 bits of the 20-bit digital data that has been subjected to the attenuation processing, and outputs 16-bit digital data A ₁₆ .

Further, it is assumed that the 20-bit digital data A ₂₀ has a value which causes a level over with respect to the 16-bit dynamic range at time t ₃ . Therefore, the controller 7
3B, the multiplication coefficient is gradually changed by the time t ₄ when the data is input to the arithmetic unit 5 as the delay signal A _20D , and the attenuation characteristic of the arithmetic unit 5 is AT ₁ →
Change it from AT _{2 to} AT ₃ . Further, in this case, since the input level is further increased after the time point t ₃ , a larger attenuation amount is set as the target of the attenuation characteristic. For example, as shown in the figure, the level detection information C _DT after the time point t _{3 is} also after the time point t _4.
In accordance with the above, the attenuation amount is further increased stepwise.

In any case, the level-over of the 20-bit digital data A ₂₀ becomes a value within the 16-bit dynamic range by the attenuation processing of the arithmetic unit 5.
Then, the arithmetic unit 5 extracts the lower 16 bits of the 20-bit digital data that has been subjected to the attenuation processing, and outputs 16-bit digital data A ₁₆ .

By performing such processing, the output 16-bit digital data A ₁₆ becomes as shown in FIG. 3C, that is, the 16-bit dynamic range does not overflow. Further, since the attenuation characteristic is gradually changed, the generation of abnormal noise due to abrupt attenuation and the discomfort in hearing are eliminated, and the sound quality is not impaired.

As shown by RC in FIG. 3B, 20-bit digital data A ₂₀ (delayed signal A
_{Although the} attenuation characteristic is gradually recovered from the timing when _20D ) becomes the value at which the level does not exceed the level, the attenuation characteristic may be held thereafter as indicated by the broken line HD in the figure. It is preferable to perform recovery in terms of dealing with sudden high volume input. On the other hand, if you want to make a rough manual recording level adjustment and then automatically make subsequent level adjustments, Therefore, it can be said that it is preferable to hold.

Next, the adaptive limiter processing will be described. This is a non-linear process with digital processing,
Level compression is performed. The adaptive type has no effect when the input level does not exceed the 16-bit dynamic range.

The arithmetic unit 5 uses the arithmetic control value C _L given by the controller 7 to obtain 20-bit digital data A.
Multiplying ₂₀ gives L ₁ , L ₂ , L in FIG.
_The input / output characteristics can be changed as shown as ₃ .... That is, the limiter characteristic is variably set by the controller 7. L ₁ is the case of limiter 0, and AT
₂ and AT ₃ ... The limit amount gradually increases.

Then, as in the case of the above attenuation processing, the controller 7 uses the level detection information C _DT to determine 16
When a level exceeding the bit dynamic range is detected, the multiplication coefficient for the arithmetic unit 5 is gradually changed according to the detected value. That is, the target of the limit characteristic is set according to the excess level, and the limit characteristic is changed until the large volume input data is delayed via the RAM 6 and input to the arithmetic unit 5.

For example, when the 17th bit of the 20-bit digital data A ₂₀ is "1" and the 18th to 20th bits are "0", and the excess level is relatively small, for example, L Set ₃ as the target limit characteristic and R
During the delay time by the AM 6, the multiplication coefficient for the arithmetic unit 5 is changed stepwise, and the limit characteristic by the arithmetic unit 5 is changed as L ₁ → L ₂ → L ₃ .

On the other hand, when the 20th bit of the 20-bit digital data A ₂₀ is “1” and the excess level is large, for example, L ₆ is set as the target limit characteristic, and the delay time by the RAM ₆ is delayed. , The multiplication coefficient for the calculation unit 5 is changed stepwise, and the limit characteristic by the calculation unit 5 is changed as L ₁ → L ₂ → L ₃ → L ₄ → L ₅ → L ₆ .

This operation is shown in FIG. 20 is shown in FIG.
The bit digital data A ₂₀ and the delayed signal A _20D are shown. 20-bit digital data A ₂₀ is 1 until t _10.
Since the value is within the 6-bit dynamic range, the controller 7 supplies the multiplication coefficient 1 to the arithmetic unit 5. That is, the input / output is executed with the characteristic of L ₁ as shown in FIG. Then, the arithmetic unit 5 takes out the lower 16 bits of 20 bits and outputs 16-bit digital data A ₁₆ .

20-bit digital data A ₂₀ at time t ₁₀
Becomes a value that causes a level over with respect to the 16-bit dynamic range. The level-over 20-bit digital data A ₂₀ is input to the arithmetic unit 5 as the delay signal A _{20 D at} time t ₁₁ after the time Td. Therefore, the controller 7 is t ₁₁ as shown in FIG.
By the time point, the multiplication coefficient is gradually changed, and the limit characteristic of the calculation unit 5 is changed as L ₁ → L ₂ → L ₃ . As a result, the level-over of the 20-bit digital data A ₂₀ is 16
The value is within the bit dynamic range. Then, the arithmetic unit 5 extracts the lower 16 bits of the 20-bit digital data that has been subjected to the attenuation processing, and outputs 16-bit digital data A ₁₆ .

Further, it is assumed that the 20-bit digital data A ₂₀ has a value which causes a level over with respect to the 16-bit dynamic range at time t ₁₃ . Therefore, the controller 7
5B, the multiplication coefficient is gradually changed by time t ₁₄ when the data is input to the arithmetic unit 5 as the delay signal A _20D , and the limit characteristic of the arithmetic unit 5 is changed from L _{1 to} L _2. →
L ₃ ... and so on. In this case t ₁₃ after the time point, since the increased further input level, as a target of the limit characteristic, greater attenuation is to be set. For example, as shown in the figure, after t ₁₄ as well, according to the level detection information C _DT after t ₁₃
Furthermore, the limit characteristics are changed step by step.

As a result, the level-over of the 20-bit digital data A ₂₀ becomes a value within the 16-bit dynamic range by the limit processing of the arithmetic unit 5.
Then, the arithmetic unit 5 extracts the lower 16 bits of the 20-bit digital data subjected to the limit processing and outputs the 16-bit digital data A ₁₆ .

By performing such processing, the output 16-bit digital data A ₁₆ becomes as shown in FIG. 5C, that is, the 16-bit dynamic range does not overflow. Further, since the limit characteristic is gradually changed, there is no discomfort in the sense of hearing and sound quality is not impaired.

Also in this case, in FIG. 5 (b),
The limit characteristic may be recovered as indicated by RC, or the limit characteristic may be held as indicated by a broken line HD in the figure.

Although the embodiments have been described above, various examples of actual adaptive processing are conceivable and may be determined according to the convenience of setting the delay time and the setting of the calculation coefficient.

[0044]

As described above, the recording signal processing apparatus according to the present invention uses the A / D conversion means of n + 1 bits or more,
Wide dynamic range, even for high volume input,
Appropriate digital data is obtained, delayed by delay means and supplied to the arithmetic means. Then, the level detection means performs level detection on the output of the A / D conversion means to detect a large volume input, and the arithmetic control means accordingly delays the large volume data and inputs it to the arithmetic circuit. In between
The limiter characteristic or the attenuation characteristic in the arithmetic circuit is gradually changed.

For this reason, the limit processing or the attenuation processing required for the large volume data is executed in the arithmetic circuit, so that the output data of n bits does not overflow. Also, since the limit level or the attenuation level is changed in accordance with the input level and is performed stepwise, there is no influence of the limit / attenuation on the input data that does not become the excess level, and the data to be subjected to the limit / attenuation processing. Also, it is considered that there is no discomfort in terms of hearing.

Further, with regard to the limit / attenuation characteristic, by recovering to the original characteristic at a predetermined timing, it is possible to carry out a processing suitable for the case of a sudden large volume input. Further, by holding the variable characteristic, it is suitable for the case where the recording level adjustment is first manually performed roughly and the subsequent level adjustment is automatically performed.

From the above, without sacrificing sound quality,
The effect that the difficulty of level adjustment in a digital recorder can be eliminated is obtained.

[Brief description of drawings]

FIG. 1 is a block diagram of a recording signal processing device of the present invention.

FIG. 2 is an explanatory diagram of an attenuation characteristic of the example.

FIG. 3 is an explanatory diagram of an adaptive attenuation process according to the embodiment.

FIG. 4 is an explanatory diagram of a limiter characteristic of the embodiment.

FIG. 5 is an explanatory diagram of an adaptive limiter process of the embodiment.

FIG. 6 is an explanatory diagram of digital and analog input / output characteristics.

FIG. 7 is an explanatory diagram of a limiter operation.

[Explanation of symbols]

 1 Input Terminal 2 A / D Converter 3 DSP 4 Level Detection Section 5 Computing Section 6 RAM 7 Controller 10 Recording Circuit Section

[Procedure amendment]

[Submission date] September 30, 1994

[Procedure Amendment 1]

[Document name to be amended] Statement

[Name of item to be corrected] 0015

[Correction method] Change

[Correction content]

In the DSP 3, reference numeral 4 is a level detecting section, and 5 is a calculating section. Further, 6 is a RAM formed of, for example, a D-RAM, and is connected to the DSP 3. The 20-bit digital data A ₂₀ is input to the level detector 4 and the level is detected. Here, the level detection unit 4 checks the value of the bit from the 17th bit to the 20th bit of the 20-bit digital data A ₂₀ , and the input voice data is 1
It is detected that the dynamic range of 6 bits is exceeded. The bit information from the 17th bit to the 20th bit, that is, the level detection information _CDT is supplied to the controller 7 formed by a microcomputer.

[Procedure Amendment 2]

[Document name to be amended] Statement

[Name of item to be corrected] 0023

[Correction method] Change

[Correction content]

For example, if the 20-bit digital data A ₂₀
If the overlevel is relatively small, for example AT
₃ is the target attenuation characteristic, the multiplication coefficient for the arithmetic unit 5 is changed stepwise during the delay time by the RAM 6, and the attenuation characteristic by the arithmetic unit 5 is AT ₁ → A
Change from T _{2 to} AT ₃ .

[Procedure 3]

[Document name to be amended] Statement

[Name of item to be corrected] 0024

[Correction method] Change

[Correction content]

On the other hand, if the 20-bit digital data A ₂₀ exceeds
When the overlevel is large, for example, AT ₆ is set as the target attenuation characteristic, and the multiplication coefficient for the arithmetic unit 5 is changed stepwise during the delay time by the RAM ₆ .
The attenuation characteristic by the calculation unit 5 is AT ₁ → AT ₂ →
Change AT ₃ → AT ₄ → AT ₅ → AT ₆ .

[Procedure amendment 4]

[Document name to be amended] Statement

[Correction target item name] 0035

[Correction method] Change

[Correction content]

For example, if the 20-bit digital data A ₂₀
When the overlevel is relatively small, for example, L ₃
Is set as a target limit characteristic, and the multiplication coefficient for the arithmetic unit 5 is changed stepwise during the delay time by the RAM 6,
The limit characteristic by the calculation unit 5 is changed as L ₁ → L ₂ → L ₃ .

[Procedure Amendment 5]

[Document name to be amended] Statement

[Correction target item name] 0036

[Correction method] Change

[Correction content]

On the other hand, if the 20-bit digital data A ₂₀
When the overlevel is large, for example, L ₆ is set as the target limit characteristic, and during the delay time by the RAM ₆ ,
The multiplication coefficient for the arithmetic unit 5 is changed stepwise, and the limit characteristic by the arithmetic unit 5 is changed to L ₁ → L ₂ → L ₃ → L ₄ → L
Change from _{5 to} L ₆ .

Claims (3)

[Claims] 1. A recording signal processing device for supplying digital audio data to a recording device section for performing a predetermined process on quantized n-bit digital audio data and recording it on a recording medium. A / D conversion means for converting into digital audio data with a quantization precision of (n + 1) bits or more; level detection means for detecting the level of the digital audio data output from the A / D conversion means; Delay means for delaying the digital audio data output from the D / D conversion means, and operation for the digital audio data output from the delay means by the operation control value supplied to quantize the output level n bits A calculation unit capable of outputting quantized n-bit digital audio data with a value within the dynamic range of In accordance with the detected value of Le detecting means, the recording signal processing apparatus characterized by being configured to include an arithmetic control means for stepwise changing the operational control value supplied to said calculation means. 2. The arithmetic control means gradually changes the arithmetic control value, and when a certain signal attenuation amount is set as an input / output characteristic of the arithmetic means, the signal attenuation amount continues for a required period. 2. The recording signal processing device according to claim 1, wherein the arithmetic control value to be supplied is fixed so as to be set. 3. The calculation control means gradually changes the calculation control value, and after a certain signal attenuation amount is set as an input / output characteristic of the calculation means, the calculation control value is changed to a detection value of the level detection means. In accordance therewith, the arithmetic control value to be supplied is changed stepwise so as to gradually reduce the signal attenuation amount by the arithmetic means and restore the normal input / output characteristic state. Recording signal processor.