KR20170084054A - Aoustic processor having low latency - Google Patents

Abstract

An audio system with low latency includes a digital audio processor and a sensor input coupled to the processor. The sensor input may be a microphone input. The audio processor operates at the same frequency as the sensor input, which is typically much higher than the audio signal provided to the audio processor. In some aspects, the audio processor is operated as a noise reduction processor and does not include an audio input.

Description

AOUSTIC PROCESSOR HAVING LOW LATENCY [0002]

The present invention relates to a sound processing, and more particularly, to a reconfigurable sound processor capable of operating in close proximity to real time or real time.

Generally, noise that is almost always present in a listening environment compromises the experience of listening to audio through headphones all the time. For example, in an aircraft cabin, the noise from an aircraft causes unwanted sound waves, such as noise, which are passed on to the listener's ear, in addition to the audio program. Other embodiments include office or house computers and air conditioning noise, vehicles and passenger noise in public or private traffic, or other noise environments.

In an effort to reduce the amount of noise received by a listener, two major styles of noise reduction have been improved, passive noise reduction and active noise reduction. Passive noise reduction refers to the reduction of noise caused by placing physical barriers, and is generally a headphone or ear plug between the ear cavity and the noise external environment. The amount of reduced noise depends on the quality of the barrier. In general, a larger mass noise-reducing headphone provides higher passive noise reduction. However, large, heavy headphones may not be comfortable to wear for extended periods of time. For a given headphone, passive noise reduction works better to reduce greater frequency noise, while low frequencies still pass through a passive noise reduction system.

An active noise reduction system, also referred to as active noise cancellation (ANC), refers to the reduction of noise achieved by playing noise suppression signals through headphone speakers. The anti-noise signal is generated as a negative approximation of the noise signal that the ANC may have in the ear canal when negative. The noise signal is then canceled in combination with the noise suppression signal.

In a typical noise reduction process, one or more sensors (e. G., A microphone) monitors ambient noise or noise in the earcup of the headphones in real time, after which the system generates a noise suppression signal from ambient or residual noise. The noise suppression signal can be used to determine the physical shape and size of the ANC system (e.g., headphones, etc.), the frequency response of the sensor and transducer, such as the speaker, the latency of the transducer at various frequencies, the sensitivity of the sensor, Can be generated by relying on different factors. Changes in these factors between different sensors and transducers (e. G., Headphones) and even between two ear cups of the same system means that the optimum filter design must also change to produce noise protection.

Latency during anti-noise signal processing prevents the active noise cancellation system from operating effectively. For example, digitizing a sensor signal and processing the signal at a normal rate, e.g., 44.1 KHz or 48 KHz, in audio processing causes a large latency. Since the performance of an acoustic processor, such as ANC, depends on its ability to detect noise and generate noise protection signals fast enough to remove noise, large latency hampers the acoustic noise removal process.

Embodiments of the present application describe these and other limitations to the prior art.

1 is a circuit diagram showing a prior art of feed-forward active noise rejection.
2 is a circuit diagram showing a prior art of feed-back active noise rejection.
3 is a circuit diagram showing a prior art of combined feed-forward and feed-back active noise rejection.
4 is a block diagram of an audio system including a reconstruction acoustic processor in accordance with an embodiment of the present invention.
FIG. 5 is a functional block diagram of an exemplary reconfigurable acoustic processor of FIG. 4;
Figure 6 is a block diagram illustrating the reconstruction acoustic processor of Figure 4 implementing a combined feed-forward and feed-back active noise canceling operation.
FIG. 7 is a functional block diagram of a component of an exemplary reconfigurable acoustic processor of FIG. 4 in accordance with an embodiment of the present disclosure.

Embodiments of the present disclosure relate to a digital acoustic processor, such as a Reconfigurable Acoustic Processor (RAP), for use in an audio system using digitized sensor inputs.

There are three main types of active noise cancellation (ANC), which are identified based on the microphone position or sensor in the system. In a feed-forward ANC, the sensor senses ambient noise, but does not noticeably detect the signal generated by the transducer, such as a speaker. Such a system is shown in Fig. 1, the feed-forward ANC system 10 includes a sensor 12 for sensing ambient noise, but does not monitor the signal directly from the transducer 14. The output from the sensor 12 is filtered at a filter output that is combined with the feed-forward filter 16 and the feed-forward mixer 18, and the filtered signal is mixed with the input audio signal. The filtered signal from the filter 16 is a noise suppression signal generated from the output of the sensor 12. The output of the transducer 14, which is a combination of the filtered noise suppression signal and the mixed input signal when the noise suppression signal is mixed with the desired signal in the mixer 18, is smaller if no noise suppression signal is generated It has noise.

In the feedback ANC, the sensor is placed in a position to sense the total audio signal present in the ear canal. In other words, the sensor senses the sum of both the ambient noise as well as the audio played back by the transducer. Such a system is shown in Fig. 2, in the feedback ANC system 20, the sensor 32 monitors the output directly from the transducer 24. [ The output from the sensor 32 is mixed with the audio input signal in the feedback mixer 30 and then the combined signal is sent to the feedback filter 34 and the combined signal is filtered to produce a noise suppression signal. This anti-noise signal from the filter 34 is mixed with the original audio signal in the mixer 28, and the combined output thereof is then fed to the transducer 24. The feedback ANC system 20 also reduces the noise heard by the listener of the speaker 24.

The combined feed-forward and feedback ANC system uses more than one sensor, the first position is for sensors in the feed-forward path shown in FIG. 1, and the second position of the sensors is feedback It is in the path. A combined feed-forward and feedback ANC system 40 is shown in FIG. 3 and includes sensor locations 42 and 52, and the position of one or more transducers is shown in FIG. The sensed signal from the feedback sensor (s) in position 52 is mixed in a feedback mixer 50 and the combined signal is filtered by a feedback filter 54. Similarly, the signal sensed from the feed-forward sensor (s) at position 42 is filtered at feed-forward filter 46 and the filtered signal is combined with the incoming audio signal at feed-forward mixer 48 . The output of the transducer (s) in position 44 will reduce noise by filtering and mixing operations.

While existing systems use fixed topologies and filters, embodiments of the present application use a selectable system for a plurality of different applications as described in detail below.

A typical audio processing rate is 44.1 kHz or 48 kHz, which is based on the frequency range of typical human auditory sense. At this sample rate, the sampling period is about 20 μs. In an ANC system, digitizing and filtering always take a large number of samples. At this rate, a delay occurs in an alignment of 100 microseconds. ANC performance is significantly lower because any delay in processing degrades the generation of noise suppression signals. This is usually maximized by limiting the maximum noise frequency that can be removed.

4 is a block diagram of an audio system 100 that includes a low latency or ultra-low latency sound processor. In some embodiments, the sound processor can be re-calibrated and is referred to as a reconfigurable audio processor (RAP) 150. The audio system of FIG. 4 includes a digital part 104 that operates at a rate of three part-analog parts 102, an analog to digital converter (ADC), and a standard audio sample rate such as 44.1 or 48 KHz And a digital unit 106 that operates. This portion is also referred to as a domain.

The analog portion 102 does not require a clock, and typically the signal in this portion is typically continuous and analog. For example, transducer or speaker 110 may generate analog audio signals such as headphones or other speakers. A sensor such as a digital microphone 112 automatically generates a digital output from an analog input signal while a standard analog sensor such as a microphone 114 is coupled to an ADC 124 to provide a digital signal from an analogue sensor 114 . The sensor 116, e.g., a microphone, is placed in a feedback position and coupled to the ADC 126. The ADCs 124 and 126 may use sigma-delta processing. In other embodiments, the ADCs 124 and 126 may be of the Pulse Code Modulation (PCM) or Successive Approximation Register (SAR) type. The signal sensors 112, 114, and 116 may be used to sample ambient noise while providing them as input microphones for various purposes, e.g., telephones. One or more filters 128 are present to filter the output from ADCs 124 and 126, but are not required in all embodiments.

A digital signal processor (DSP) 130 or other audio source is operated on the digital section 106 at a frequency of a standard audio sample rate. The operating frequency of the digital portion 106 of the audio system 100 is typically 44.1 or 48 KHz.

Conversely, the operating frequency of the digital section 104 operates from a low rate of about 50 KHz to a rate of about 100 MHz, preferably within a range of 2-100 MHz. In some embodiments, the digital portion 104 is capable of operating at frequencies of 10 KHz to 10 KHz, such as at frequencies of 50 KHz, 96 KHz, and several tens MHz, e.g., up to about 100 MHz, in the range of hundreds of KHz, Lt; / RTI > In the present embodiment, each component of a particular domain operates at the frequency of the domain. For example, in FIG. 4, the ADCs 124 and 126 operate at the same frequency as the audio processor or RAP 150. This is somewhat different from previous systems that typically use a decimation filter to downsample the sensor signal before processing by the audio processor.

Interpolator 140 converts an audio signal from DSP 130 operating at 48 KHz to an audio signal operating at 3 MHz or 6 MHz as an input signal to RAP 150, for example. Conversely, decimator 144 need not be present in all audio systems 100, which translates the signal from RAP 150, for example, at 3 or 6 MHz, to the operating frequency of digital section 106. The resulting latency of the RAP 150 is extremely low, e.g., less than 2.5 μs, preferably less than 0.5 μs, because the sensor is a digital microphone or the sensor signal is digitized by the ADCs 124 and 126 Or the RAP 150 processes the signal at the same rate as the rate generated by the sensor or microphone 112, 114, 116.

As described in further detail below, the RAP 150 controls the acoustic signal from the transducer 110 in real time. As described above, the RAP 150 is configured to operate on a raw sensor sample from the microphone 112, 114, and / or 116 without any intermediate processing, such as a decimation filter or other sample rate converter. This causes the RAP 150 to respond to the microphone signal with a zero or near zero computation delay and enables implementation of a real time audio processing algorithm. The effect of using real-time sensor sampling is to eliminate the delay from the decimation filter of the previous system, which in turn increases the reactivity of the control loop dramatically.

The sample rate of the digital portion 104 may be varied according to the sample rate of the ADC 124 or the digital sensor 112 coupled to the analog sensor 114. [ There is a linear tradeoff between the amount of processing that can be processed per sample and the sample rate.

FIG. 5 is a functional block diagram of an exemplary reconfigurable sound processor (RAP) 250, which may be an embodiment of the RAP 150 of FIG. The RAP 250 of FIG. 5 includes bi-quadratic, or bi-quad filters (BQ0-BQ6) of six chains, functions to be described below. Bi-quad filters are known as electrical processing, especially audio processing. A bi-quad filter typically includes two zeros and two poles. The bi-quad chains (BQ0-BQ6) each include a cascade of bi-quad filters. In some embodiments, the chains BQ0-BQ6 may comprise 4, 6, 8, 12, or 16 cascaded bi-quad filters, and this range is preferred. The bi-quad filter chain (BQ0-BQ6) is programmable so that the filtering values can be varied according to the desired implementation. The values may also be set to pass-through or integrated settings, which means that the values do not noticeably affect the signal passing through them.

Each of the bi-quad filter chains BQ0-BQ6 is connected to a gain unit for the purpose described below, and has an additional gain unit M7. The gain units (M0-M7) are programmable, and the amount of gain provided between their inputs and outputs is controllable. The output of a particular bi-quad filter chain (BQ0-BQ6) can be controlled by its combined gain units (M0-M6). Setting the gain of any gain unit (M0-M6) to zero can effectively turn off a particular circuit branch. While this is not entirely necessary to maintain a one-to-one relationship between the bi-quad filter chain and the gain unit, the relationship maintains the flexibility to configure the RAP. The RAP 250 of FIG. 5 shows a single audio channel. For one or more channels, for example, for stereo processing, additional hardware may be used.

By programming the specific filter coefficients in the bi-quad filter chain (BQ0-BQ6) and the specific gain values in the gain units (M0-M6), different audio applications, such as audio noise reduction, 250). ≪ / RTI >

It may also include an input from digital sensors 212 and 214 coupled to RAP 250 that may be a microphone, decimator 218 and interpolator 220. One or both of the sensor inputs 212 and 214 may be generated by including an analog microphone coupled to the ADC. Decimator 218 and interpolator 220 operate as described in connection with FIG.

In operation, when the RAP 250 receives an input from the sensor 212 in the bi-quad filter chains BQ0 and BQ3 and the input from the sensor 214 is received in the bi-quad filter chains BQ1 and BQ5 do. The audio signal is received in the bi-quad filter chains BQ2 and BQ6. In some embodiments, the audio signal is not entirely necessary. For example, in a hunter or an industrial noise canceling headphone, the audio signal may not be present.

The gain unit M7 can be used as a controllable gain for the processed audio signal before final combination in the combiner A2 with the processed audio signal from the interpolator 220 to the raw audio signal. The gain unit M7 can be controlled to gradually increase its gain so that noise cancellation or other processing can be added to the raw audio signal progressively to make the pop or fast The change can be eliminated.

Adders or combiners A0, A1, and A2 combine the intermediate signal outputs from the bi-quad filter chain, as shown in FIG.

In one embodiment, RAP 350 operates at 49.152 MHz, which is the standard rate for audio processing. The input sample rate is typically 3.072 Msps, and the filter portion also operates at the same rate.

A simple embodiment of RAP 250 operation is a simple audio processor that does not use inputs from any of the sensors 212, In this embodiment, the gain unit M7 is set to zero and turned off, for example, while the audio signal from the interpolator is filtered by the bi-quad filter chain BQ6. Control of the gain unit M6 controls the output signal level of the filtered audio signal and is transmitted to the transducer 210, which may be a speaker or other transducer output.

In a more complex embodiment, the RAP 250 may be configured as a feed-forward / feedback ANC, which has the same functionality as the feed-forward and feed-back ANC circuitry shown in FIG. Figure 6 shows how RAP 250 is set up for such a configuration. In this configuration, the gain units M0 and M5 are set to zero and are shown in Figure 6 as having "x ", indicating that there is no contribution to processing. The gain units M2, M4, M6, and M7 are set to one. The gain units M1 and M3 are set to -1, and their outputs are subtracted. The bi-quad filter chain (BQ1, BQ2 and BQ6) is set to pass-through setting. 3 and 6, the bi-quad filter chain BQ3 serves as the feed-forward filter 46 and the bi-quad filter chain BQ4 serves as the feedback filter 54. [

By setting the RAP 250, particularly the gain units M0-M7 and the bi-quad filter chains BQ1-BQ6, the RAP can be configured to perform any type of audio processing to a maximum. For example, the RAP 250 may be configured as an ANC processor for active noise canceling headphones, either in feedback, feed-forward, or in a combined feed-forward feedback configuration. The RAP 250 can be used for active noise removal in a phone headset by using the input from the headset microphone and providing the audio output for one or more speakers in the headset. The RAP 250 can further enhance the input audio signal while simultaneously performing noise cancellation. RAP 250 receives an ambient sound at one of the microphone inputs, modifies it through one or more bi-quad filter chains, sets a suitable gain level, and then sends the modified ambient signal And can be used to raise ambient tone.

Actually, the RAP 250 of FIG. 6 or the RAP 150 of FIG. 5 includes functions, processes or operations for modifying the audio signal input. Indeed, as the programmed function is operated on a general purpose or special purpose processor, such as a digital signal processor (DSP), this function may be implemented by specially formed hardware circuitry or may be implemented by a Field Programmable Gate Arrays ) Or a programmable logic device (PLD). Other variations are also possible.

Figure 7 is a functional block diagram of components of an embodiment of the reconstruction acoustic processor of Figure 4 in accordance with an embodiment of the present disclosure. In FIG. 7, the RAP 350 includes a bi-quad engine 310 and a multiplier accumulator 320. The multiplier accumulator 320 implements all of the multipliers and adders in the functional block diagrams of Figures 5 and 6. In one embodiment, there are 7 multi-add operations per sample. The bi-quad engine 310 includes one or more sensors, e.g., an input from a microphone, and an input of an audio signal to be processed. The bi-quad engine also receives input from the multiplication adder output. The input from the sensor is clocked at the same rate as the biphasic engine. In other words, the sensor input can be processed without any decimation or rate reduction. The bi-quad engine 310 may be sized to operate on a 16 bi-quad filter. Quad descriptor section 330 includes filter values to implement a bi-quad filter chain and bi-quad state memory 332 is a memory for storing intermediate values during bi-quad processing. A gain table 322 stores values for the gain unit and feathering control is provided separately by the feathering control 334, as provided by the gain unit M7 of FIG. The RAP 350 is programmed and configured by writing a specific value into the bi-quad descriptor 330 and the gain table 322 of FIG.

By using programmable techniques, the filter is selected to enhance, rather than eliminate, certain sounds or noises. For example, instead of the bi-quad chain filter parameter selected for an effort to reduce the sound sensed by a particular microphone, a parameter that enhances a particular sound may be selected. For example, a person may use noise canceling headphones in a noisy working environment with a variety of noisy machines, but still want to be able to talk to their peers without removing the noise canceling headphones. When the microphone detects noise in the vocal band, using adaptive filter coefficients, different parameters can be automatically loaded into the RAP system, which enhances the voice of the peer. Thus, the listener has a noise-canceling headphone that adaptively amplifies a specific sound. Sounds such as voice, audio television signals, and traffic can be enhanced. If such a sound is distracted, for example, if a co-worker stops talking, the standard filtering coefficients can again be dynamically loaded into the filter of the RAP system.

Embodiments of the present application may be integrated into an integrated circuit, such as a sound processing circuit, or other audio circuitry. In turn, integrated circuits can be used in audio devices such as headphones, cell phones, portable computing systems, sound bars, audio docks, amplifiers, speakers, and the like.

It should be appreciated that the principles herein have been illustrated and described herein with reference to illustrated embodiments, and that the illustrated embodiments may be varied in arrangement and content, and combined in any desired manner, without departing from such principles. Also, although the foregoing is focused on particular embodiments, other configurations may be contemplated.

In particular, even though expressions such as "according to embodiments of the present disclosure" are described, this expression generally refers to the feasibility of the embodiments and is not intended to limit the scope of the specific embodiments. As used herein, such terms refer to the same or different embodiments that may be combined in other embodiments.

As a result, with respect to a wide variety of orders for the embodiments described herein, such description and accompanying material are merely exemplary in nature and are not intended to limit the scope of the disclosure herein.

Claims (38)

As an audio system,
A sensor for generating a digital sensor signal at a first rate higher than 50 KHz; And
A digital audio processor having a first input for receiving the input audio signal at the first rate and a second input for receiving the sensor signal at the first rate,
/ RTI >
Audio system. The method according to claim 1,
Wherein the sensor is a microphone. 3. The method of claim 2,
Wherein the microphone is part of a system comprising a microphone coupled to an analog to digital converter (ADC), the output of the ADC being the sensor signal. The method of claim 3,
Wherein the ADC is configured to perform sigma-delta processing. The method of claim 3,
Wherein the ADC is configured to perform SAR processing. The method according to claim 1,
And wherein the output is coupled to a transducer. The method according to claim 6,
Wherein the transducer is a speaker. The method according to claim 1,
Wherein the first rate is a rate of at least about 50 KHz. The method according to claim 1,
Wherein the first rate is a rate of at least about 96 KHz. The method according to claim 1,
Wherein the first rate is a rate of at least about 200 KHz. The method according to claim 1,
Wherein the first rate is a rate of at least about 350 KHz. The method according to claim 1,
Wherein the first rate is a rate of at least about 750 KHz. The method according to claim 1,
Wherein the first rate is a rate of at least about 1 MHz. The method according to claim 1,
Wherein the first rate is a rate greater than or equal to about 3 MHz. The method according to claim 1,
Wherein the first rate is a rate of at least about 6 MHz. The method according to claim 1,
Further comprising an interpolator having an input for receiving a digital audio signal at a second rate and converting the digital audio signal to an audio signal having the first rate. 17. The method of claim 16,
Wherein the second rate is lower than the first rate. 17. The method of claim 16,
And wherein the second rate is a rate of about 100 KHz or less. The method according to claim 1,
The audio processor
A plurality of programmable filters, and
A plurality of controllable gain stages, wherein at least some of the plurality of controllable gain stages are each coupled to at least a portion of the plurality of programmable filters. 1. A reconstruction noise reduction system,
An input for receiving a digital audio signal at a first rate;
An interpolator for changing the sample rate of the digital audio signal from the first rate to a second rate higher than the first rate;
At least one sensor for generating a sensor signal at the second rate; And
A digital audio processor coupled to the interpolator and the sensor,
Lt; / RTI >
Wherein the reconfigurable audio processor is operative at the second rate,
A plurality of programmable filters,
A plurality of controllable gain stages, wherein at least some of the plurality of controllable gain stages are each coupled to at least a portion of the plurality of programmable filters,
Adders configured to combine outputs of one or more gain stages, and
And an audio output coupled to at least one of the adders to deliver an output audio signal that is modified from the input audio signal.
Reconstruction noise reduction system. 21. The method of claim 20,
Wherein the at least one sensor is a digital sampling microphone operating at the second rate. 21. The method of claim 20,
Wherein the at least one sensor is an analog microphone coupled to an analog to digital converter (ADC). 23. The method of claim 22,
Wherein the ADC is configured to perform sigma-delta processing. 23. The method of claim 22,
Wherein the ADC is configured to perform SAR processing. 21. The method of claim 20,
Wherein the programmable filter is configured to be programmed during operation of the noise control system. 21. The method of claim 20,
Wherein at least some of the plurality of controllable gain stages are configured to be updated during operation of the noise reduction system. CLAIMS 1. A method of operating an audio system,
Operating the digital audio processor at a first rate that is greater than or equal to about 50 KHz;
Receiving at the digital audio processor a digital input audio signal having the first rate;
Receiving at the digital audio processor a digital sensor signal having the first rate;
Processing the digital input audio signal in a digital audio processor operating at the first rate by combining a digital input audio signal with a signal derived from the digital sensor signal; And
And outputting the processed digital input audio signal to the output unit
/ RTI >
How to operate the audio system. 28. The method of claim 27,
Wherein receiving the digital sensor signal having the first rate at the digital audio processor comprises receiving a microphone signal at the first rate. 28. The method of claim 27,
Wherein outputting the processed digital input audio signal to the output includes outputting a digital input audio signal to the transducer. 28. The method of claim 27,
Wherein the first rate is a rate of at least about 96 KHz. 28. The method of claim 27,
Wherein the first rate is a rate of about 1 MHz or more. 28. The method of claim 27,
Wherein the first rate is a rate of at least about 3 MHz. 28. The method of claim 27,
Wherein the first rate is a rate of at least about 6 MHz. 28. The method of claim 27,
Receiving a digital audio signal at a second rate;
Converting the digital audio signal from the second rate to the first rate; And
And transmitting the digital audio signal to the digital audio processor at the first rate. 35. The method of claim 34,
Wherein the second rate is lower than the first rate. 35. The method of claim 34,
Wherein the second rate is a rate of about 100 KHz or less. CLAIMS What is claimed is: 1. A method of operating a reconstruction noise reduction processor,
Receiving an audio signal through an audio input at a first frequency;
Receiving one or more sensor signals of the monitored environment at the first frequency through one or more sensor inputs;
Configuring a filter parameter section of the plurality of programmable filters in the reconfiguration noise removal processor;
Configuring a plurality of controllable gain stages in the reconfiguration noise removal processor, wherein at least some of the plurality of controllable gain stages are each coupled to at least a portion of the plurality of programmable filters, Comprising; And
Mixing the selected outputs of the plurality of controllable gain stages with the audio signal to produce a modified audio signal output
/ RTI >
How to operate the reconstruction noise reduction processor. 39. The method of claim 37,
Further comprising changing a filter parameter section of the plurality of programmable filters during operation of the reconfiguration noise removal processor.

Images