Method and device for eliminating unstable noise

Abstract

The present invention provides a method and device for eliminating unstable noise, which is applicable to a recording device and processed by a codec. The unstable noise elimination method includes: starting the recording device to start recording; setting a suppression time and a cutoff frequency switching time according to the unstable noise and DC offset value of the recording device; processing the front audio of the recording with a filter having a first cutoff frequency, so that the unstable noise in the front audio converges quickly and outputs the front audio signal; suppressing the front audio signal according to the suppression time to eliminate the unstable noise; and adjusting the first cutoff frequency of the filter to a second cutoff frequency according to the cutoff frequency switching time, wherein the first cutoff frequency is greater than the second cutoff frequency.

Description

Technical Field

[0001] The present invention relates to an audio processing technology, and more particularly to a method and device for eliminating unstable noise. Background Art

[0002] When recording, when a general digital microphone is just turned on or the mode is switched, the instantaneous changes produced by the microphone often produce unstable noise, such as popping noise, which leads to poor listening experience and reduces the user's experience. The existing method is to use software to process audio, using software to compensate the high-pass filter (high pass filter) to make the DC value converge faster, and cooperate with a volume mask (volume mask) to make the initial popping noise unnoticeable due to the small volume. However, this method cannot guarantee that the volume mask added when the microphone produces a large DC value can be unnoticeable in the auditory sense, and maintaining a high cutoff frequency (-3dB corner) The high-pass filter will also filter out part of the audio bandwidth (audio BW), causing audio distortion. Summary of the invention

[0003] In view of this, the present invention proposes a method for eliminating unstable noise, which is applicable to a recording device and processed by a codec. The unstable noise elimination method includes: starting the recording device to start recording; setting a suppression time and a cutoff frequency switching time according to the unstable noise and DC offset value of the recording device; processing the front audio of the recording with a filter having a first cutoff frequency, so that the unstable noise in the front audio converges quickly and outputs the front audio signal; suppressing the front audio signal according to the suppression time to eliminate the unstable noise; and adjusting the first cutoff frequency of the filter to a second cutoff frequency according to the cutoff frequency switching time, wherein the first cutoff frequency is greater than the second cutoff frequency.

[0004] The present case also proposes a device for eliminating unstable noise, which includes a recording device and a codec, and the codec is electrically connected to the recording device. The codec includes a filter, an audio channel and a suppression circuit. The filter is electrically connected to the recording device and the audio channel. When the recording device starts recording, the filter processes the front audio of the recording with a first cutoff frequency and outputs a front audio signal. Then, the filter adjusts the first cutoff frequency to a second cutoff frequency according to a cutoff frequency switching time, wherein the first cutoff frequency is greater than the second cutoff frequency. The audio channel is electrically connected to the suppression circuit to transmit the front audio signal to the suppression circuit. The suppression circuit suppresses the front audio signal according to a suppression time to eliminate the unstable noise in the front audio.

[0005] According to some embodiments, the step of starting the recording device is when the recording device is powered on or when the recording device switches modes.

[0006] According to some embodiments, before the step of setting the suppression time and the cutoff frequency switching time according to the unstable noise and DC offset value of the recording device, the method further includes: using a detection circuit to detect the unstable noise and DC offset value of the recording device.

[0007] According to some embodiments, the detection circuit further transmits the unstable noise and the DC offset value to a decision circuit, so that the decision circuit sets the suppression time and the cut-off frequency switching time according to the unstable noise and the DC offset value.

[0008] According to some embodiments, in the step of setting the suppression time and the cutoff frequency switching time according to the unstable noise and the DC offset value of the recording device, the decision circuit can also set the first cutoff frequency according to the DC offset value.

[0009] According to some embodiments, the second cut-off frequency is a cut-off frequency when the recording device operates normally.

[0010] According to some embodiments, the recording device is a digital microphone.

[0011] In summary, the present invention uses hardware control to make the unstable noise generated by the recording device during recording converge and be suppressed faster, so as to remove the unstable noise (instantaneous changes) in the audio before the recording device is recorded. In addition, the present invention can dynamically adjust the suppression time, cutoff frequency switching time and first cutoff frequency according to the unstable noise and DC offset value of the recording device, so as to adapt to the instantaneous changes of the recording device during recording and reduce the duration of recording distortion, thereby achieving a smaller distortion effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 FIG. 4 is a block diagram of a device for eliminating unstable-state noise according to an embodiment of the present invention.

[0013] Figure 2 FIG. 4 is a flow chart of a method for eliminating unstable-state noise according to an embodiment of the present invention.

[0014] Figure 3 FIG. 4 is a block diagram of a device for eliminating unstable-state noise according to another embodiment of the present invention.

[0015] Figure 4 FIG. 4 is a flow chart of a method for eliminating unstable-state noise according to another embodiment of the present invention.

[0016] Figure 5 It is a schematic diagram of an audio signal without using the method of eliminating unstable noise in this case.

[0017] Figure 6 FIG. 4 is a schematic diagram of an audio signal using the method of eliminating unstable noise of the present invention. DETAILED DESCRIPTION

[0018] Figure 1 FIG. 1 is a block diagram of a device for eliminating unstable noise according to an embodiment of the present invention. Figure 1 As shown, the device 1 for eliminating unstable noise includes a codec 10 and a recording device 12. The codec 10 is electrically connected to the recording device 12. The codec 10 includes a register control circuit 14, a filter 16, an audio path 18, and a suppression circuit 20. The filter 16 is electrically connected to the recording device 12, the register control circuit 14, and the audio path 18, and the suppression circuit 20 is electrically connected to the register control circuit 14 and the audio path 18, so as to utilize the filter 16 to filter the audio signal, and utilize the suppression circuit 20 to suppress the audio signal within a specific time.

[0019] In one embodiment, the recording device 12 is any sound-converting device, such as a digital microphone. In another embodiment, the recording device 12 can also be composed of an analog microphone and an analog-to-digital converter.

[0020] In one embodiment, the filter 16 is a high pass filter.

[0021] Please also see Figure 1 and Figure 2 As shown, the method for eliminating unstable noise is applicable to the recording device 12 and is processed by the codec 10. The method for eliminating unstable noise includes: as shown in step S10, starting the recording device 12 to start recording, wherein the recording device 12 can be started when the recording device 12 is powered on or when the recording device 12 switches modes, so as to generate different instantaneous changes according to the recording device 12. Since the characteristics of the recording device 12 are known, the unstable noise and DC offset value (DC offset value) corresponding to the recording device 12 can be known. As shown in step S12, software or firmware can be used to set a suppression time T according to the unstable noise and DC offset value of the recording device 12. m and a cut-off frequency switching time T c , where this inhibition time T m And the cut-off frequency switching time T c can be stored in the temporary storage control circuit 14, and the temporary storage control circuit 14 will suppress the time T m is transmitted to the suppression circuit 20 and the cut-off frequency switching time T cThe recording device 12 transmits the recorded audio data to the filter 16. As shown in step S14, when the recording device 12 transmits the recorded audio data to the filter 16, the filter 16 filters the front-end audio of the recording at a higher first cutoff frequency, so that the unstable noise in the front-end audio can converge quickly, and then the filter 16 outputs a front-end audio signal. The front-end audio signal processed by the filter 16 is transmitted to the suppression circuit 20 through the audio channel 18. As shown in step S16, the suppression circuit 20 will suppress the front-end audio signal according to the suppression time T. m , suppress the previous audio signal and maintain the suppression time T m Finally, as shown in step S18, the temporary storage control circuit 14 switches the time T according to the cut-off frequency. c , the cutoff frequency of the filter 16 is adjusted from the first cutoff frequency to a second cutoff frequency, wherein the first cutoff frequency is greater than the second cutoff frequency, and the second cutoff frequency is the cutoff frequency when the recording device 12 is working normally, so the higher first cutoff frequency can be used to allow the unstable noise in the previous audio to converge quickly. In other words, when recording starts, the filter 16 is processed with the higher first cutoff frequency, and after a cutoff frequency switching time T c After that, the filter 16 will adjust from the initial higher first cutoff frequency back to the normal second cutoff frequency, and filter the subsequent audio data at the normal second cutoff frequency. The processed audio signal continues to pass through the audio channel 18 and the suppression circuit 20, so that the codec 10 outputs an audio stream. When the audio signal passes through the suppression circuit 20, the suppression time T has passed. m , the suppression circuit 20 will cancel the suppression and stop working, so the audio signal will not be suppressed, so the audio stream can be directly output normally through the suppression circuit 20.

[0022] In one embodiment, the aforementioned unstable noise may be direct current pop noise, but is not limited thereto.

[0023] In one embodiment, the inhibition time T is set in step S12. m And the cut-off frequency switching time T c When the filter 16 is initially set to the first cutoff frequency according to the DC offset value of the recording device 12, the higher first cutoff frequency can be used to quickly converge the unsteady-state noise (instantaneous change) of the previous audio.

[0024] In one embodiment, in step S16, the suppression step may use a mute method to eliminate the unstable noise. That is, the suppression circuit 20 may suppress the noise according to the suppression time T. m , mute the previous audio signal and maintain the suppression time T m , so as to eliminate the unsteady-state noise.

[0025] Figure 3 FIG. 1 is a block diagram of a device for eliminating unstable noise according to another embodiment of the present invention. Figure 3 As shown, a device 1 for eliminating unstable noise includes a codec 10 and a recording device 12. The codec 10 is electrically connected to the recording device 12. The codec 10 includes a detection circuit 22, a decision circuit 24, a filter 16, an audio channel 18, and a suppression circuit 20. The detection circuit 22 is electrically connected to the recording device 12 and the decision circuit 24. The filter 16 is electrically connected to the recording device 12, the decision circuit 24, and the audio channel 18. The suppression circuit 20 is electrically connected to the decision circuit 24 and the audio channel 18. In this codec 10, the detection circuit 22 detects the unstable noise and DC offset value corresponding to the recording device 12, and transmits the unstable noise and DC offset value to the decision circuit 24, so that the decision circuit 24 determines the suppression time T according to the unstable noise and DC offset value. m And the cut-off frequency switching time T c , to be transmitted to the suppression circuit 20 and the filter 16 respectively, so that the suppression circuit 20 can suppress the filter 16 according to the suppression time T m To suppress, the filter 16 switches the time T according to the cut-off frequency c to switch the cutoff frequency.

[0026] Please also see Figure 3 and Figure 4 As shown, the method for eliminating unstable noise is applicable to the recording device 12 and is processed by the codec 10. The method includes: as shown in step S20, when the recording device 12 is powered on or switches modes, the recording device 12 is started to start recording. As shown in step S22, the detection circuit 22 detects the unstable noise and DC offset value of the recording device 12, and transmits the detected unstable noise and DC offset value to the decision circuit 24. As shown in step S24, the decision circuit 24 sets the suppression time T according to the unstable noise and DC offset value. m And the cut-off frequency switching time T c , and the suppression time T m is transmitted to the suppression circuit 20 and the cut-off frequency switching time T c The filter 16 transmits the recorded audio data to the filter 16. As shown in step S26, when the recording device 12 transmits the recorded audio data to the filter 16, the filter 16 filters the front-end audio of the recording at a higher first cutoff frequency, so that the unstable noise in the front-end audio can converge quickly, and the filter 16 outputs a front-end audio signal. The front-end audio signal processed by the filter 16 is transmitted to the suppression circuit 20 through the audio channel 18. As shown in step S28, the suppression circuit 20 will receive the suppression time T according to the received suppression time T. m, suppress the previous audio signal and maintain the suppression time T m , to eliminate unstable noise. Finally, as shown in step S30, the decision circuit 24 switches the time T according to the cut-off frequency. c , the cutoff frequency of the filter 16 is adjusted from the higher first cutoff frequency to the normal second cutoff frequency. After the cutoff frequency switching time T of the filter 16 c After that, the first higher cutoff frequency is adjusted back to the normal second cutoff frequency, and the subsequent audio data is filtered at the normal second cutoff frequency. The processed audio signal passes through the audio channel 18 and the suppression circuit 20 in sequence to output the audio stream. When the audio signal passes through the suppression circuit 20, the suppression time T has passed. m , the suppression circuit 20 will cancel the suppression and stop working, so the audio signal will not be suppressed, so the audio stream can be directly output normally through the suppression circuit 20.

[0027] In one embodiment, in step S24, the decision circuit 24 sets the inhibition time T m And the cut-off frequency switching time T c When the filter 16 is set to the initial first cut-off frequency according to the DC offset value of the recording device 12, the decision circuit 24 can use the higher first cut-off frequency to quickly converge the unstable noise of the previous audio.

[0028] In one embodiment, in step S28, the suppression step may use a muting method to eliminate the unstable noise, that is, the suppression circuit 20 may suppress the noise according to the suppression time T m , mute the previous audio signal and maintain the suppression time T m , so as to eliminate the unsteady-state noise.

[0029] In Figure 3 and Figure 4 In the operation mode shown, the detection circuit 22 can be used to automatically detect the unstable noise and DC offset value contributed by the recording device 12 and dynamically adjust the suppression time T m And the cut-off frequency switching time T c , so that the system can be more adaptable to various recording devices 12, so as to remove the unstable noise (instantaneous changes) generated by the recording device 12 during recording.

[0030] In one embodiment, the codec 10 may be built into a host (not shown), and the recording device 12 may be plugged into the host to form an electrical connection with the codec 10, so as to eliminate unstable noise through the codec 10. The host may be a laptop computer, a tablet computer, a personal computer, etc., but is not limited thereto.

[0031] Please also see Figure 5 and Figure 6 , before using the method of eliminating unstable noise in this case, such as Figure 5 As shown in FIG. 1 , after the recording device is started, the audio signal starts at 0 seconds. At this time, unstable noise is generated in the front part of the audio signal. Then, after 400 milliseconds, the DC signal (audio signal) stabilizes. Figure 6 As shown, after the recording device is started, the audio signal will also start from 0 seconds, but the unstable noise appearing in the front audio of the audio signal has been suppressed, and the DC signal (audio signal) has stabilized at 150 milliseconds, so the entire stabilization time is much faster.

[0032] Therefore, the present invention uses hardware control to make the unstable noise generated by the recording device during recording converge faster and be suppressed, so as to remove the unstable noise (instantaneous changes) in the audio before the recording device is recorded. In addition, the present invention can dynamically adjust the suppression time, cutoff frequency switching time and first cutoff frequency according to the unstable noise and DC offset value of the recording device to adapt to the instantaneous changes of the recording device during recording and reduce the duration of recording distortion, thereby achieving a smaller distortion effect.

[0033] The embodiments described above are only for illustrating the technical ideas and features of the present invention, and their purpose is to enable those skilled in the art to understand the content of the present invention and implement it accordingly. They cannot be used to limit the scope of the claims of the present invention. That is, any equivalent changes or modifications made according to the spirit disclosed in the present invention should still be included in the scope of the claims of the present invention.

[0034]

Explanation of symbols

[0035] 1: Device for eliminating unstable noise

[0036] 10: Codec

[0037] 12: Recording device

[0038] 14: Temporary storage control circuit

[0039] 16: Filter

[0040] 18: Audio Channel

[0041] 20: Suppression circuit

[0042] 22: Detection circuit

[0043] 24: Decision circuit

[0044] S10 to S18: Steps

[0045] S20 to S30: Steps

[0046] T c : Cut-off frequency switching time

[0047] T m : Suppression time.

Claims

1. A method for eliminating unstable noise, applicable to a recording device and processed by a codec, the method for eliminating unstable noise comprising: activating the recording device to start recording; Setting a suppression time and a cutoff frequency switching time according to an unstable noise and a DC offset value of the recording device; Processing the front audio of the recording with a filter having a first cutoff frequency so that the unstable noise in the front audio converges quickly, and outputting a front audio signal; According to the suppression time, suppress the front audio signal to eliminate the unstable noise; as well as According to the cut-off frequency switching time, the first cut-off frequency of the filter is adjusted to a second cut-off frequency, wherein the first cut-off frequency is greater than the second cut-off frequency, wherein the cutoff frequency switching time is greater than the suppression time, and Among them, the filter is a high-pass filter. Before the step of setting the suppression time and the cut-off frequency switching time according to the unstable noise and the DC offset value of the recording device, the method further includes: using a detection circuit to detect the unstable noise and the DC offset value of the recording device, And wherein, the detection circuit transmits the unstable noise and the DC offset value to a decision circuit, so that the decision circuit sets the suppression time and the cut-off frequency switching time according to the unstable noise and the DC offset value.

2. The method for eliminating unstable-state noise as claimed in claim 1, wherein the step of starting the recording device is when the recording device is powered on or when the recording device switches modes.

3. The method for eliminating unstable noise as described in claim 1, wherein in the step of setting the suppression time and the cutoff frequency switching time according to the unstable noise of the recording device and the DC offset value, the decision circuit can also set the first cutoff frequency according to the DC offset value.

4. The method for eliminating unstable noise as claimed in claim 1, wherein the step of setting the suppression time and the cutoff frequency switching time according to the unstable noise of the recording device and the DC offset value further includes setting the first cutoff frequency according to the DC offset value.

5. A device for eliminating unstable-state noise, the device for eliminating unstable-state noise implementing the method for eliminating unstable-state noise according to any one of claims 1 to 4, and the device for eliminating unstable-state noise comprising: a recording device; and A codec, electrically connected to the recording device, the codec comprising: a filter electrically connected to the recording device, and when the recording device starts recording, the filter processes the front audio of the recording with a first cutoff frequency and outputs a front audio signal, and the filter adjusts the first cutoff frequency to a second cutoff frequency according to a cutoff frequency switching time, wherein the first cutoff frequency is greater than the second cutoff frequency; an audio channel electrically connected to the filter to transmit the front-end audio signal; and a suppression circuit electrically connected to the audio channel and receiving the front audio signal, wherein the suppression circuit suppresses the front audio signal according to a suppression time, in, The cutoff frequency switching time is greater than the suppression time, and Among them, the filter is a high-pass filter.

6. The device for eliminating unstable noise as claimed in claim 5, wherein the codec further comprises a temporary storage controller electrically connected to the suppression circuit and the filter, for storing the suppression time and the cut-off frequency switching time set in advance, and transmitting the suppression time to the suppression circuit and the cut-off frequency switching time to the filter respectively.

7. The device for eliminating unstable noise as claimed in claim 5, wherein the codec further comprises: a detection circuit electrically connected to the recording device for detecting an unstable noise and a DC offset value of the recording device; and a decision circuit electrically connected to the detection circuit, the suppression circuit and the filter, the decision circuit setting the suppression time and the cut-off frequency switching time according to the unstable noise and the DC offset value, and transmitting the suppression time to the suppression circuit and the cut-off frequency switching time to the filter respectively. in, Before the step of setting the suppression time and the cut-off frequency switching time according to the unstable noise and the DC offset value by the decision circuit, the method further includes: using the detection circuit to detect the unstable noise and the DC offset value of the recording device, And wherein, the detection circuit transmits the unstable noise and the DC offset value to the decision circuit, so that the decision circuit sets the suppression time and the cut-off frequency switching time according to the unstable noise and the DC offset value.

8. The device for eliminating unstable-state noise as claimed in claim 5, wherein the second cut-off frequency is a cut-off frequency when the recording device works normally.

Images